0% found this document useful (0 votes)

108 views232 pages

Arml 2014-2019

Uploaded by

Anthony C

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

108 views232 pages

Arml 2014-2019

Uploaded by

Anthony C

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 232

ARML Competition 2014

Paul J. Karafiol, Head Writer

Paul Dreyer
Edward Early
Zuming Feng
Benji Fisher
Zachary Franco
Chris Jeuell
Winston Luo
Andy Niedermaier
George Reuter
Andy Soffer
Eric Wepsic

May 30-31, 2014

1 Team Problems

Problem 1. There exists a digit Y such that, for any digit X, the seven-digit number 1 2 3 X 5 Y 7 is
not a multiple of 11. Compute Y .

√
Problem 2. A point is selected at random from the interior of a right triangle with legs of length 2 3
and 4. Let p be the probability that the distance
√ between the point and the nearest vertex is less
than 2. Then p can be written in the form a + bπ, where a and b are rational numbers. Compute
(a, b).

Problem 3. The square ARM L is contained in the xy-plane with A = (0, 0) and M = (1, 1). Compute
the length of the shortest path from the point (2/7, 3/7) to itself that touches three of the four sides
of square ARM L.

Problem 4. For each positive integer k, let Sk denote the infinite arithmetic sequence of integers with
first term k and common difference k 2 . For example, S3 is the sequence 3, 12, 21, . . . . Compute the
sum of all k such that 306 is an element of Sk .

Problem 5. Compute the sum of all values of k for which there exist positive real numbers x and y
satisfying the following system of equations.

logx y 2 + logy x5 = 2k − 1

logx2 y 5 − logy2 x3 = k − 3

Problem 6. Let W = (0, 0), A = (7, 0), S = (7, 1), and H = (0, 1). Compute the number of ways to
tile rectangle WASH with triangles of area 1/2 and vertices at lattice points on the boundary of
WASH.

Problem 7. Compute sin2 4◦ + sin2 8◦ + sin2 12◦ + · · · + sin2 176◦ .

Problem 8. Compute the area of the region defined by x2 + y 2 ≤ |x| + |y|.

Problem 9. The arithmetic sequences a1 , a2 , a3 , . . . , a20 and b1 , b2 , b3 , . . . , b20 consist of 40 distinct pos-
itive integers, and a20 + b14 = 1000. Compute the least possible value for b20 + a14 .

Problem 10. Compute the ordered triple (x, y, z) representing the farthest lattice point from the origin
that satisfies xy − z 2 = y 2 z − x = 14.

1
2 Answers to Team Problems

Answer 1. 4

1 1

Answer 2. 4 , 27

2
√
Answer 3. 7 53

Answer 4. 326

43
Answer 5. 48

Answer 6. 3432

45
Answer 7. 2

Answer 8. 2 + π

Answer 9. 10

Answer 10. (−266, −3, −28)

2
3 Solutions to Team Problems

Problem 1. There exists a digit Y such that, for any digit X, the seven-digit number 1 2 3 X 5 Y 7 is
not a multiple of 11. Compute Y .

Solution 1. Consider the ordered pairs of digits (X, Y ) for which 1 2 3 X 5 Y 7 is a multiple of 11. Recall
that a number is a multiple of 11 if and only if the alternating sum of the digits is a multiple of 11.
Because 1 + 3 + 5 + 7 = 16, the sum of the remaining digits, namely 2 + X + Y , must equal 5 or 16.
Thus X + Y must be either 3 or 14, making X = 3 − Y (if Y = 0, 1, 2, or 3) or 14 − Y (if Y = 5, 6,
7, 8, or 9). Thus a solution (X, Y ) exists unless Y = 4.

√
Solution 2. Label the triangle as 4ABC, with AB = 2 3 and BC = 4. Let D and E lie on AB such
that DB = AE = 2. Let F be the midpoint of BC, so that BF = F C = 2. Let G and H lie on AC,
with AG = HC = 2. Now draw the arcs of radius 2 between E and G, D and F , and F and H. Let
the intersection of arc DF and arc EG be J. Finally, let M be the midpoint of AB. The completed
diagram is shown below.

G
D
M
E J H

B F C

The region R consisting of all points within 4ABC that lie within 2 units of any vertex is the union
of the three sectors EAG, DBF , and F CH. The angles of these sectors, being the angles ∠A, ∠B,
and ∠C, sum to 180◦ , so the sum of their areas is 2π. Computing the area of R requires subtracting
the areas of all intersections of the three sectors that make up R.
The only sectors that intersect are EAG and DBF . Half this area of intersection, the part above
M J, equals the difference between the areas of sector DBJ and of 4M BJ. Triangle M BJ is a
√ √ √
3
1 : 3 : 2 right triangle because BM = 3 and BJ = 2, so the area of 4M BJ is 2 . Sector DBJ
has area 12 1
(4π) = π3 , because m∠DBJ = 30◦ . Therefore the area of intersection of the sectors is
√ √ √ √
2( π3 − 23 ) = 2π 2π 4π
3 − 3. Hence the total area of R is 2π − ( 3 − 3) = 3 + 3.

3
√
√ 4π
+ 3 π
The total area of 4ABC is 4 3, therefore the desired probability is 3 √
4 3
= √
3 3
+ 41 . Then a = 1
4
2
1 1 1 1

and b = 3√ 3
= 27 , hence the answer is 4
, 27
.

Solution 3. Consider repeatedly reflecting square ARM L over its sides so that the entire plane is cov-
ered by copies of ARM L. A path starting at (2/7, 3/7) that touches one or more sides and returns to
(2/7, 3/7) corresponds to a straight line starting at (2/7, 3/7) and ending at the image of (2/7, 3/7)
in one of the copies of ARM L. To touch three sides, the path must cross three lines, at least one of
which must be vertical and at least one of which must be horizontal.

• • •

If the path crosses two horizontal lines and the line x = 0, p

it will have traveled a distance of 2 units
vertically and 4/7 units vertically for a total distance of 22 + (4/7)
p
2 units. Similarly, the total

distance traveled when crossing two horizontal lines and 2 2

p x = 1 is 2 + (10/7) , the total distance
traveled when crossing two vertical lines and p 2 2
y = 0 is 2 + (6/7) , and the total distance traveled
when crossing two vertical lines and y = 1 is 22 + (8/7)2 . The least of these is
p 2√
22 + (4/7)2 = 53.
7

Solution 4. If 306 is an element of Sk , then there exists an integer m ≥ 0 such that 306 = k + mk 2 .
Thus k | 306 2
√ and k | 306 − k. The second relation can be rewritten as2 k | 306/k − 1, which implies
that k ≤ 306 unless k = 306. The prime factorization of 306 is 2 · 3 · 17, so the set of factors of

4
√
306 less than 306 is {1, 2, 3, 6, 9, 17}. Check each in turn:

306 − 1 = 305, 12 | 305

306 − 2 = 304, 22 | 304
306 − 3 = 303, 32 - 303
306 − 6 = 300, 62 - 300
306 − 9 = 297, 92 - 297
306 − 17 = 289, 172 | 289.

Thus the set of possible k is {1, 2, 17, 306}, and the sum is 1 + 2 + 17 + 306 = 326.

Problem 5. Compute the sum of all values of k for which there exist positive real numbers x and y
satisfying the following system of equations.

logx y 2 + logy x5 = 2k − 1

logx2 y 5 − logy2 x3 = k − 3

Solution 5. Let logx y = a. Then the first equation is equivalent to 2a + a5 = 2k − 1, and the second
equation is equivalent to 5a 3
2 − 2a = k − 3. Solving this system by eliminating k yields the quadratic
equation 3a2 + 5a − 8 = 0, hence a = 1 or a = − 83 . Substituting each of these values of a into either
of the original equations and solving for k yields (a, k) = (1, 4) or − 83 , − 149
48 . Adding the values of
k yields the answer of 43/48.

Alternate Solution: In terms of a = logx y, the two equations become 2a + a5 = 2k − 1 and

5a 3 1
2 − 2a = k − 3. Eliminate a to obtain 31a = 16k − 33; substitute this into either of the origi-
nal equations and clear denominators to get 96k 2 − 86k − 1192 = 0. The sum of the two roots is
86/96 = 43/48.

Problem 6. Let W = (0, 0), A = (7, 0), S = (7, 1), and H = (0, 1). Compute the number of ways to
tile rectangle WASH with triangles of area 1/2 and vertices at lattice points on the boundary of
WASH.

Solution 6. Define a fault line to be a side of a tile other than its base. Any tiling of W ASH can be
represented as a sequence of tiles t1 , t2 , . . . , t14 , where t1 has a fault line of W H, t14 has a fault line
of AS, and where tk and tk+1 share a fault line for 1 ≤ k ≤ 13. Also note that to determine the
position of tile tk+1 , it is necessary and sufficient to know the fault line that tk+1 shares with tk , as
well as whether the base of tk+1 lies on W A (abbreviated “B” for “bottom”) or on SH (abbreviated
“T” for “top”). Because rectangle W ASH has width 7, precisely 7 of the 14 tiles must have their
bases on W A. Thus any permutation of 7 B’s and 7 T’s determines a unique tiling t1 , t2 , . . . , t14 , and
conversely, any
tiling t1 , t2 , . . . , t14 corresponds to a unique permutation of 7 B’s and 7 T’s. Thus the
answer is 147 = 3432.

Alternate Solution: Let T (a, b) denote the number of ways to triangulate the polygon with ver-
tices at (0, 0), (b, 0), (a, 1), (0, 1), where each triangle has area 1/2 and vertices at lattice points. The
problem is to compute T (7, 7). It is easy to see that T (a, 0) = T (0, b) = 1 for all a and b. If a and b

5
are both positive, then either one of the triangles includes the edge from (a − 1, 1) to (b, 0) or one of
the triangles includes the edge from (a, 1) to (b − 1, 0), but not both. (In fact, as soon as there is an
edge from (a, 1) to (x, 0) with x < b, there must be edges from (a, 1) to (x0 , 0) for all x ≤ x0 < b.)
If there is an edge from (a − 1, 1) to (b, 0), then the number of ways to complete the triangulation
is T (a − 1, b); if there is an edge from (a, 1) to (b − 1, 0), then the number of ways to complete the
triangulation is T (a, b − 1); thus T (a, b) = T (a − 1, b) + T (a, b − 1). The recursion and the initial
conditions describe Pascal’s triangle, so T (a, b) = a+b 14
a . In particular, T (7, 7) = 7 = 3432.

Problem 7. Compute sin2 4◦ + sin2 8◦ + sin2 12◦ + · · · + sin2 176◦ .

1−cos 2x
Solution 7. Because cos 2x = 1 − 2 sin2 x, sin2 x = 2 . Thus the desired sum can be rewritten as
1 − cos 8◦ 1 − cos 16◦ 1 − cos 352◦ 44 1
+ + ··· + = − (cos 8◦ + cos 16◦ + · · · + cos 352◦ ) .
2 2 2 2 2
If α = cos 8◦ + i sin 8◦ , then α is a primitive 45th root of unity, and 1 + α + α2 + α3 + · · · + α44 = 0.
Hence α + α2 + · · · + α44 = −1, and because the real part of αn is simply cos 8n◦ ,

cos 8◦ + cos 16◦ + · · · + cos 352◦ = −1.

Thus the desired sum is 22 − (1/2)(−1) = 45/2.

Alternate Solution: The problem asks to simplify the sum

sin2 a + sin2 2a + sin2 3a + · · · + sin2 na

where a = 4◦ and n = 44. Because cos 2x = 1 − 2 sin2 x, sin2 x = 1−cos

2
2x
. Thus the desired sum can
be rewritten as
1 − cos 2a 1 − cos 4a 1 − cos 2na n 1
+ + ··· + = − (cos 2a + cos 4a + · · · + cos 2na) .
2 2 2 2 2
Let Q = cos 2a + cos 4a + · · · + cos 2na. By the sum-to-product identity,

sin 3a − sin a = 2 cos 2a sin a,

sin 5a − sin 3a = 2 cos 4a sin a,
..
.
sin(2n + 1)a − sin(2n − 1)a = 2 cos 2na sin a.

Thus

Q · 2 sin a = (sin 3a − sin a) + (sin 5a − sin 3a) + · · · + (sin(2n + 1)a − sin(2n − 1)a)
= sin(2n + 1)a − sin a.

With a = 4◦ and n = 44, the difference on the right side becomes sin 356◦ − sin 4◦ ; note that the
terms in this difference are opposites, because of the symmetry of the unit circle. Hence

Q · 2 sin 4◦ = −2 sin 4◦ , and

Q = −1.

Thus the original sum becomes 44/2 − (1/2)(−1) = 45/2.

6
Problem 8. Compute the area of the region defined by x2 + y 2 ≤ |x| + |y|.

Solution 8. Call the region R, and let Rq be the portion of R in the q th quadrant. Noting that the
point (x, y) is in R if and only if (±x, ±y) is in R, it follows that [R1 ] = [R2 ] = [R3 ] = [R4 ], and so
[R] = 4[R1 ]. So it suffices to determine [R1 ].

In the first quadrant, the boundary equation is just x2 + y 2 = x + y ⇒ (x − 12 )2 + (y − 12 )2 = 12 . This

√
equation describes a circle of radius 22 centered at ( 12 , 12 ). The portion of the circle’s interior which
is inside the first quadrant
√ can be decomposed into a right isosceles triangle with side length 1 and
half a circle of radius 22 . Thus [R1 ] = 12 + π4 , hence [R] = 2 + π.

Solution 9. Write an = a1 + r(n − 1) and bn = b1 + s(n − 1). Then a20 + b14 = a1 + b1 + 19r + 13s,
while b20 + a14 = a1 + b1 + 13r + 19s = a20 + b14 + 6(s − r). Because both sequences consist
only of integers, r and s must be integers, so b20 + a14 ≡ a20 + b14 mod 6. Thus the least possible
value of b20 + a14 is 4. If b20 = 3 and a14 = 1, then {an } must be a decreasing sequence (else a13
would not be positive) and a20 ≤ −5, which is impossible. The case b20 = a14 = 2 violates the
requirement that the terms be distinct, and by reasoning analogous to the first case, b20 = 1, a14 = 3
is also impossible. Hence the sum b20 + a14 is at least 10. To show that 10 is attainable, make {an }
decreasing and b20 as small as possible: set b20 = 1, a14 = 9, and an = 23 − n. Then a20 = 3,
997−1
yielding b14 = 997. Hence s = 14−20 = 996
−6 = −166 and b1 = 997 − (13)(−166) = 3155, yielding
bn = 3155 − 166(n − 1). Because b20 = 1 ≤ a20 and b19 = 167 ≥ a1 , the sequences {bn } and {an } are
distinct for 1 ≤ n ≤ 20, completing the proof. Hence the minimum possible value of b20 + a14 is 10.
[Note: This solution, which improves on the authors’ original solution, is due to Ravi Jagadeesan of
Phillips Exeter Academy.]

Problem 10. Compute the ordered triple (x, y, z) representing the farthest lattice point from the origin
that satisfies xy − z 2 = y 2 z − x = 14.

Solution 10. First, eliminate x: y(y 2 z − x) + (xy − z 2√

) = 14(y + 1) ⇒ z 2 − y 3 z + 14(y + 1) = 0. Viewed
y3 ± y 6 −56(y+1)
as a quadratic in z, this equation implies z = 2 . In order for z to be an integer, the
discriminant must be a perfect square. Because y = (y ) and (y 3 − 1)2 = y 6 − 2y 3 + 1, it follows
6 3 2

that |56(y + 1)| ≥ 2|y 3 | − 1. This inequality only holds for |y| ≤ 5. Within that range, the only
values of y for which y 6 − 56y − 56 is a perfect square are −1 and −3. If y = −1, then z = −1 or
z = 0. If y = −3, then z = 1 or z = −28. After solving for the respective values of x in the various
cases, the four lattice points satisfying the system are (−15, −1, −1), (−14, −1, 0), (−5, −3, 1), and
(−266, −3, −28). The farthest solution point from the origin is therefore (−266, −3, −28).

7
4 Power Question 2014: Power of Potlucks
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next
to the part. To receive full credit, the presentation must be legible, orderly, clear, and concise. If a prob-
lem says “list” or “compute,” you need not justify your answer. If a problem says “determine,” “find,” or
“show,” then you must show your work or explain your reasoning to receive full credit, although such expla-
nations do not have to be lengthy. If a problem says “justify” or “prove,” then you must prove your answer
rigorously. Even if not proved, earlier numbered items may be used in solutions to later numbered items,
but not vice versa. Pages submitted for credit should be NUMBERED IN CONSECUTIVE ORDER AT
THE TOP OF EACH PAGE in what your team considers to be proper sequential order. PLEASE WRITE
ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM NUMBER (not the team name)
on the cover sheet used as the first page of the papers submitted. Do not identify the team in any other way.

In each town in ARMLandia, the residents have formed groups, which meet each week to share math prob-
lems and enjoy each others’ company over a potluck-style dinner. Each town resident belongs to exactly
one group. Every week, each resident is required to make one dish and to bring it to his/her group.
It so happens that each resident knows how to make precisely two dishes. Moreover, no two residents of
a town know how to make the same pair of dishes. Shown below are two example towns. In the left column
are the names of the town’s residents. Adjacent to each name is the list of dishes that the corresponding
resident knows how to make.

ARMLton ARMLville
Resident Dishes Resident Dishes
Paul pie, turkey Sally steak, calzones
Arnold pie, salad Ross calzones, pancakes
Kelly salad, broth David steak, pancakes
The population of a town T , denoted pop(T ), is the number of residents of T . Formally, the town
itself is simply the set of its residents, denoted by {r1 , . . . , rpop(T ) } unless otherwise specified. The set of
dishes that the residents of T collectively know how to make is denoted dish(T ). For example, in the town
of ARMLton described above, pop(ARMLton) = 3, and dish(ARMLton) = {pie, turkey, salad, broth}.
A town T is called full if for every pair of dishes in dish(T ), there is exactly one resident in T who
knows how to make those two dishes. In the examples above, ARMLville is a full town, but ARMLton is
not, because (for example) nobody in ARMLton knows how to make both turkey and salad.
Denote by Fd a full town in which collectively the residents know how to make d dishes. That is,
|dish(Fd )| = d.

1a. Compute pop(F17 ). [1 pt]

1b. Let n = pop(Fd ). In terms of n, compute d. [1 pt]
1c. Let T be a full town and let D ∈ dish(T ). Let T 0 be the town consisting of all residents of T who do
not know how to make D. Prove that T 0 is full. [2 pts]

In order to avoid the embarrassing situation where two people bring the same dish to a group dinner, if
two people know how to make a common dish, they are forbidden from participating in the same group
meeting. Formally, a group assignment on T is a function f : T → {1, 2, . . . , k}, satisfying the condition
that if f (ri ) = f (rj ) for i 6= j, then ri and rj do not know any of the same recipes. The group number of
a town T , denoted gr(T ), is the least positive integer k for which there exists a group assignment on T .

8
For example, consider once again the town of ARMLton. A valid group assignment would be f (Paul) =
f (Kelly) = 1 and f (Arnold) = 2. The function which gives the value 1 to each resident of ARMLton is
not a group assignment, because Paul and Arnold must be assigned to different groups.

2a. Show that gr(ARMLton) = 2. [1 pt]

2b. Show that gr(ARMLville) = 3. [1 pt]

3a. Show that gr(F4 ) = 3. [1 pt]

3b. Show that gr(F5 ) = 5. [2 pts]
3c. Show that gr(F6 ) = 5. [2 pts]

4. Prove that the sequence gr(F2 ), gr(F3 ), gr(F4 ), . . . is a non-decreasing sequence. [2 pts]

For a dish D, a resident is called a D-chef if he or she knows how to make the dish D. Define chef T (D)
to be the set of residents in T who are D-chefs. For example, in ARMLville, David is a steak-chef and a
pancakes-chef. Further, chef ARMLville (steak) = {Sally, David}.

5. Prove that X
|chef T (D)| = 2 pop(T ). [2 pts]
D∈dish(T )

6. Show that for any town T and any D ∈ dish(T ), gr(T ) ≥ |chef T (D)|. [2 pts]

If gr(T ) = |chef T (D)| for some D ∈ dish(T ), then T is called homogeneous. If gr(T ) > |chef T (D)| for
each dish D ∈ dish(T ), then T is called heterogeneous. For example, ARMLton is homogeneous, because
gr(ARMLton) = 2 and exactly two chefs make pie, but ARMLville is heterogeneous, because even though
each dish is only cooked by two chefs, gr(ARMLville) = 3.

7. For n = 5, 6, and 7, find a heterogeneous town T of population 5 for which |dish(T )| = n. [3 pts]

A resident cycle is a sequence of distinct residents r1 , . . . , rn such that for each 1 ≤ i ≤ n − 1, the residents
ri and ri+1 know how to make a common dish, residents rn and r1 know how to make a common dish, and
no other pair of residents ri and rj , 1 ≤ i, j ≤ n know how to make a common dish. Two resident cycles
are indistinguishable if they contain the same residents (in any order), and distinguishable otherwise. For
example, if r1 , r2 , r3 , r4 is a resident cycle, then r2 , r1 , r4 , r3 and r3 , r2 , r1 , r4 are indistinguishable resident
cycles.

8a. Compute the number of distinguishable resident cycles of length 6 in F8 . [1 pt]

8b. In terms of k and d, find the number of distinguishable resident cycles of length k in Fd . [1 pt]

9. Let T be a town with at least two residents that has a single resident cycle that contains every
resident. Prove that T is homogeneous if and only if pop(T ) is even. [3 pts]

9
10. Let T be a town such that, for each D ∈ dish(T ), |chef T (D)| = 2.

a. Prove that there are finitely many resident cycles C1 , C2 , . . . , Cj in T so that each resident
belongs to exactly one of the Ci . [3 pts]
b. Prove that if pop(T ) is odd, then T is heterogeneous. [3 pts]

11. Let T be a town such that, for each D ∈ dish(T ), |chef T (D)| = 3.

a. Either find such a town T for which |dish(T )| is odd, or show that no such town exists. [2 pts]
b. Prove that if T contains a resident cycle such that for every dish D ∈ dish(T ), there exists a
chef in the cycle that can prepare D, then gr(T ) = 3. [3 pts]

12. Let k be a positive integer, and let T be a town in which |chef T (D)| = k for every dish D ∈ dish(T ).
Suppose further that |dish(T )| is odd.

a. Show that k is even. [1 pt]

b. Prove the following: for every group in T , there is some dish D ∈ dish(T ) such that no one in
the group is a D-chef. [3 pts]
c. Prove that gr(T ) > k. [3 pts]

13a. For each odd positive integer d ≥ 3, prove that gr(Fd ) = d. [3 pts]
13b. For each even positive integer d, prove that gr(Fd ) = d − 1. [4 pts]

10
5 Solutions to Power Question
17

1. a. There are = 136 possible pairs of dishes, so F17 must have 136 people.
2
2 2
b. With d dishes there are d2 = d 2−d possible pairs, so n = d 2−d . Then 2n = d2 − d, or

√
d2 − d − 2n = 0. Using the quadratic formula yields d = 1+ 21+8n (ignoring the negative value).
c. The town T 0 consists of all residents of T who do not know how to make D. Because T is full,
every pair of dishes {di , dj } in dish(T ) can be made by some resident rij in T . If di 6= D and
dj 6= D, then rij ∈ T 0 . So every pair of dishes in dish(T ) \ {D} can be made by some resident
of T 0 . Hence T 0 is full.

2. a. Paul and Arnold cannot be in the same group, because they both make pie, and Arnold and
Kelly cannot be in the same group, because they both make salad. Hence there must be at
least two groups. But Paul and Kelly make none of the same dishes, so they can be in the same
group. Thus a valid group assignment is

Paul 7→ 1
Kelly 7→ 1
Arnold 7→ 2.

Hence gr(ARMLton) = 2.
b. Sally and Ross both make calzones, Ross and David both make pancakes, and Sally and David
both make steak. So no two of these people can be in the same group, and gr(ARMLville) = 3.

3. a. Let the dishes be d1 , d2 , d3 , d4 and let resident rij make dishes di and dj , where i < j. There
are six pairs of dishes, which can be divided into nonoverlapping pairs in three ways: {1, 2} and
{3, 4}, {1, 3} and {2, 4}, and {1, 4} and {2, 3}. Hence the assignment r12 , r34 7→ 1, r13 , r24 7→ 2,
and r14 , r23 7→ 3 is valid, hence gr(F4 ) = 3.
b. First, gr(F5 ) ≥ 5: there are 52 = 10 people in F5 , and because each person cooks two different

dishes, any valid group of three people would require there to be six different dishes—yet there
are only five. So each group can have at most two people. A valid assignment using five groups
is shown below.
Residents Group
r12 , r35 1
r13 , r45 2
r14 , r23 3
r15 , r24 4
r25 , r34 5

c. Now there are 62 = 15 people, but there are six different dishes, so it is possible (if done

carefully) to place three people in a group. Because four people in a single group would require
there to be eight different dishes, no group can have more than three people, and so 15/3 = 5
groups is minimal. (Alternatively, there are five different residents who can cook dish d1 , and no
two of these can be in the same group, so there must be at least five groups.) The assignment

11
below attains that minimum.
Residents Group
r12 , r34 , r56 1
r13 , r25 , r46 2
r14 , r26 , r35 3
r15 , r24 , r36 4
r16 , r23 , r45 5

4. Pick some n ≥ 2 and a full town Fn whose residents prepare dishes d1 , . . . , dn , and let gr(Fn ) = k.
Suppose that fn : Fn → {1, 2, . . . , k} is a valid group assignment for Fn . Then remove from Fn
all residents who prepare dish dn ; by problem 1c, this operation yields the full town Fn−1 . Define
fn−1 (r) = fn (r) for each remaining resident r in Fn . If r and s are two (remaining) residents
who prepare a common dish, then fn (r) 6= fn (s), because fn was a valid group assignment. Hence
fn−1 (r) 6= fn−1 (s) by construction of fn−1 . Therefore fn−1 is a valid group assignment on Fn−1 , and
the set of groups to which the residents of Fn−1 are assigned is a (not necessarily proper) subset of
{1, 2, . . . , k}. Thus gr(Fn−1 ) is at most k, which implies the desired result.

5. Because each chef knows how to prepare exactly two dishes, and no two chefs know how to prepare
the same two dishes, each chef is counted exactly twice in the sum Σ |chef T (D)|. More formally,
consider the set of “resident-dish pairs”:

S = {(r, D) ∈ T × dish(T ) | r makes D}.

Second, each resident knows how to make exactly two dishes, so

X
|S| = 2 = 2 pop(T ).
r∈T

6. Let D ∈ dish(T ). Suppose that f is a valid group assignment on T . Then for r, s ∈ chef T (D), if
r 6= s, it follows that f (r) 6= f (s). Hence there must be at least |chef T (D)| distinct groups in the
range of f , i.e., gr(T ) ≥ |chef T (D)|.

7. For n = 5, this result is attained as follows:

Resident Dishes
Amy d1 , d2
Benton d2 , d3
Carol d3 , d4
Devin d4 , d5
Emma d5 , d1

12
For each dish D, note that chef T (D) = 2. But gr(T ) > 2, because if T had at most two groups, at
least one of them would contain three people, and choosing any three people will result in a common
dish that two of them can cook. Hence T is heterogeneous.

For n ≥ 6, it suffices to assign dishes to residents so that there are three people who must be
in different groups and that no dish is cooked by more than two people, which guarantees that
gr(T ) ≥ 3 and chef T (D) ≤ 2 for each dish D.

Resident Dishes
Amy d1 , d2
Benton d1 , d3
Carol d2 , d3
Devin d4 , d5
Emma d5 , d6

Note that Devin’s and Emma’s dishes are actually irrelevant to the situation, so long as they do not
cook any of d1 , d2 , d3 , which already have two chefs each. Thus we can adjust this setup for n = 7
by setting Devin’s dishes as d4 , d5 and Emma’s dishes as d6 , d7 . (In this last case, Devin and Emma
are extremely compatible: they can both be put in a group with anyone else in the town!)

8. a. Because the town is full, each pair of dishes is cooked by exactly one resident, so it is simplest
to identify residents by the pairs of dishes they cook. Suppose the first resident cooks (d1 , d2 ),
the second resident (d2 , d3 ), the third resident (d3 , d4 ), and so on, until the sixth resident, who
cooks (d6 , d1 ). Then there are 8 choices for d1 and 7 choices for d2 . There are only 6 choices for
d3 , because d3 6= d1 (otherwise two residents would cook the same pair of dishes). For k > 3, the
requirement that no two intermediate residents cook the same dishes implies that dk+1 cannot
equal any of d1 , . . . , dk−1 , and of course dk and dk+1 must be distinct dishes. Hence there are
8 · 7 · 6 · 5 · 4 · 3 = 20,160 six-person resident cycles, not accounting for different starting points
in the cycle and the two different directions to go around the cycle. Taking these into account,
there are 20,160/(6 · 2) = 1,680 distinguishable resident cycles.
b. Using the logic from 8a, there are d(d − 1) · · · (d − k + 1) choices for d1 , d2 , . . . , dk . To account for
indistinguishable cycles, divide by k possible starting points and 2 possible directions, yielding
d(d−1)···(d−k+1) d!
2k or 2k(d−k)! distinguishable resident cycles.

9. Note that for every D ∈ dish(T ), chef T (D) ≤ 2, because otherwise, r1 , r2 , . . . , rn could not be a
resident cycle. Without loss of generality, assume the cycle is r1 , r2 , . . . , rn . If n is even, assign
resident ri to group 1 if i is odd, and to group 2 if i is even. This is a valid group assignment, because
the only pairs of residents who cook the same dish are (ri , ri+1 ) for i = 1, 2, . . . , n − 1 and (rn , r1 ). In
each case, the residents are assigned to different groups. This proves gr(T ) = 2, so T is homogeneous.
On the other hand, if n is odd, suppose for the sake of contradiction that there are only two groups.
Then either r1 and rn are in the same group, or for some i, ri and ri+1 are in the same group.
In either case, two residents in the same group share a dish, contradicting the requirement that no
members of a group have a common dish. Hence gr(T ) ≥ 3 when n is odd, making T heterogeneous.

10. a. First note that the condition |chef T (D)| = 2 for all D implies that pop(T ) = |dish(T )|, using
the equation from problem 5. So for the town in question, the population of the town equals the

13
number of dishes in the town. Because no two chefs cook the same pair of dishes, it is impossible
for such a town to have exactly two residents, and because each dish is cooked by exactly two
chefs, it is impossible for such a town to have only one resident.
The claim is true for towns of three residents satisfying the conditions: such towns must have
one resident who cooks dishes d1 and d2 , one resident who cooks dishes d2 and d3 , and one
resident who cooks dishes d3 and d1 , and those three residents form a cycle. So proceed by
(modified) strong induction: assume that for some n > 3 and for all positive integers k such
that 3 ≤ k < n, every town T with k residents and |chef T (D)| = 2 for all D ∈ dish(T ) can be
divided into a finite number of resident cycles such that each resident belongs to exactly one
of the cycles. Let Tn be a town of n residents, and arbitrarily pick resident r1 and dishes d1
and d2 cooked by r1 . Then there is exactly one other resident r2 who also cooks d2 (because
|chef Tn (d2 )| = 2). But r2 also cooks another dish, d3 , which is cooked by another resident, r3 .
Continuing in this fashion, there can be only two outcomes: either the process exhausts all the
residents of Tn , or there exists some resident rm , m < n, who cooks the same dishes as rm−1
and r` for ` < m − 1.
In the former case, rn cooks another dish; but every dish besides d1 is already cooked by two
chefs in Tn , so rn must also cook d1 , closing the cycle. Because every resident is in this cycle,
the statement to be proven is also true for Tn .
In the latter case, the same logic shows that rm cooks d1 , also closing the cycle, but there are
other residents of Tn who have yet to be accounted for. Let C1 = {r1 , . . . , rm }, and consider the
town T 0 whose residents are Tn \ C1 . Each of dishes d1 , . . . , dm is cooked by two people in C1 , so
no chef in T 0 cooks any of these dishes, and no dish in T 0 is cooked by any of the people in C1
(because each person in C1 already cooks two dishes in the set dish(C1 )). Thus |chef T 0 (D)| = 2
for each D in dish(T 0 ). It follows that pop(T 0 ) < pop(T ) but pop(T 0 ) > 0, so by the inductive
hypothesis, the residents of T 0 can be divided into disjoint resident cycles.
Thus the statement is proved by strong induction.
b. In order for T to be homogeneous, it must be possible to partition the residents into exactly
two dining groups. First apply 10a to divide the town into finitely many resident cycles Ci , and
assume towards a contradiction that such a group assignment f : T → {1, 2} exists. If pop(T )
is odd, then at least one of the cycles Ci must contain an odd number of residents; without loss
of generality, suppose this cycle to be C1 , with residents r1 , r2 , . . . , r2k+1 . (By the restrictions
noted in part a, k ≥ 1.) Now because ri and ri+1 cook a dish in common, f (ri ) 6= f (ri+1 ) for
all i. Thus if f (r1 ) = 1, it follows that f (r2 ) = 2, and that f (r3 ) = 1, etc. So f (ri ) = f (r1 ) if
i is odd and f (ri ) = f (r2 ) if i is even; in particular, f (r2k+1 ) = f (1). But that equation would
imply that r1 and r2k+1 cook no dishes in common, which is impossible if they are the first and
last residents in a resident cycle. So no such group assignment can exist, and gr(T ) ≥ 3. Hence
T is heterogeneous.

11. a. In problem 5, it was shown that

X
2 pop(T ) = |chef T (D)| .
D∈dish(T )
P
Therefore D∈dish(T ) |chef T (D)| is even. But if |chef T (D)| = 3 for all D ∈ dish(T ), then the
sum is simply 3 |dish(T )|, so |dish(T )| must be even.
b. By problem 6, it must be the case that gr(T ) ≥ 3. Let C = {r1 , r2 , . . . , rn } denote a resident
cycle such that for every dish D ∈ dish(T ), there exists a chef in C that can prepare D. Each
resident is a chef for two dishes, and every dish can be made by two residents in C (although by

14
three in T ). Thus the number of residents in the resident cycle C is equal to |dish(T )|, which
was proved to be even in the previous part.
Define a group assignment by setting

1 if r 6∈ C

f (r) = 2 if r = ri , i is even

3 if r = ri , i is odd.


For any D ∈ dish(T ), there are exactly three D-chefs, and exactly two of them belong to the
resident cycle C. Hence exactly one of the D-chefs r will have f (r) = 1. The remaining two
D-chefs will be ri and ri+1 for some i, or r1 and rn . In either case, the group assignment f will
assign one of them to 2 and the other to 3. Thus any two residents who make a common dish
will be assigned different groups by f , so f is a valid group assignment, proving that gr(T ) = 3.

12. a. From problem 5, X

2 pop(T ) = |chef T (D)| .
D∈dish(T )

Because |chef T (D)| = k for all D ∈ dish(T ), the sum is k · dish(T ). Thus 2 pop T = k · dish(T ),
and so k · dish(T ) must be even. By assumption, |dish(T )| is odd, so k must be even.
b. Suppose for the sake of contradiction that there is some n for which the group R = {r ∈ T |
f (r) = n} has a D-chef for every dish D. Because f is a group assignment and f assigns every
resident of R to group n, no two residents of R make the same dish. Thus for every D ∈ dish(T ),
exactly one resident of R is a D-chef; and each D-chef cooks exactly one other dish, which itself
is not cooked by anyone else in R. Thus the dishes come in pairs: for each dish D, there is
another dish D0 cooked by the D-chef in R and no one else in R. However, if the dishes can be
paired off, there must be an even number of dishes, contradicting the assumption that |dish(T )|
is odd. Thus for every n, the set {r ∈ T | f (r) = n} must be missing a D-chef for some dish D.
c. Let f be a group assignment for T , and let R = {r ∈ T | f (r) = 1}. From problem 12b, there
must be some D ∈ dish(T ) with no D-chefs in R. Moreover, f cannot assign two D-chefs to the
same group, so there must be at least k other groups besides R. Hence there are at least 1 + k
different groups, so gr(T ) > k.

Label the dishes D1 , D2 , . . . , Dd , and label the residents ri,j for 1 ≤ i < j ≤ d so that ri,j
is a Di -chef and a Dj -chef. Define f : Fd → {0, 1, . . . , d − 1} by letting f (ri,j ) ≡ i + j mod d.

Suppose that f (ri,j ) = f (rk,` ), so i + j ≡ k + ` mod d. Then ri,j and rk,` are assigned to
the same group, which is a problem if they are different residents but are chefs for the same
dish. This overlap occurs if and only if one of i and j is equal to one of k and `. If i = k,
then j ≡ ` mod d. As j and ` are both between 1 and d, the only way they could be congruent
modulo d is if they were in fact equal. That is, ri,j is the same resident as rk,` . The other three
cases (i = `, j = k, and j = `) are analogous. Thus f is a valid group assignment, proving that
gr(Fd ) ≤ d. Therefore gr(Fd ) = d.

15
b. In problem 4, it was shown that the sequence gr(F2 ), gr(F3 ), . . . is nondecreasing. If d is even,
gr(Fd ) ≥ gr(Fd−1 ), and because d − 1 is odd, problem 13a applies: gr(Fd−1 ) = d − 1. Hence
gr(Fd ) ≥ d − 1. Now it suffices to show that gr(Fd ) ≤ d − 1 by exhibiting a valid group
assignment f : Fd → {1, 2, . . . , d − 1}.
Label the dishes D1 , . . . , Dd , and label the residents ri,j for 1 ≤ i < j ≤ d so that ri,j is a
Di -chef and a Dj -chef. Let R = {ri,j | i, j 6= d}. That is, R is the set of residents who are not
Dd -chefs. Using 1c, R is a full town with d − 1 dishes, so from 12a, it has a group assignment
f : R → {1, 2, . . . , d − 1}. For each Di ∈ dish(Fd ), i 6= d, | chef R (Di )| = d − 2. Because there
are d − 1 groups and |chef R (Di )| = d − 2, exactly one group ni must not contain a Di -chef for
each dish Di .
It cannot be the case that ni = nj for i 6= j. Indeed, suppose for the sake of contradiction that
ni = nj . Without loss of generality, assume that ni = nj = 1 (by perhaps relabeling the dishes).
Then any resident r ∈ R assigned to group 1 (that is, f (r) = 1) would be neither a Di -chef nor
a Dj -chef. The residents in R who are assigned to group 1 must all be chefs for the remaining
d − 3 dishes. Because each resident cooks two dishes, and no two residents of group 1 can make
a common dish,
d−3
|{r ∈ R | f (r) = 1}| ≤ .
2
For each of the other groups 2, 3, . . . , d − 1, the number of residents of R in that group is no
more than (d − 1)/2, because there are d − 1 dishes in R, each resident cooks two dishes, and
no two residents in the same group can make a common dish. However, because d − 1 is odd,
the size of any group is actually no more than (d − 2)/2. Therefore
d−1
X
|R| = |{r ∈ R | f (r) = k}|
k=1
d−1
X
= |{r ∈ R | f (r) = 1}| + |{r ∈ R | f (r) = k}|
k=2
d−1
d−3 Xd−2
≤ +
2 2
k=2
d − 3 (d − 2)2
= +
2 2
d2 − 3d + 1 d2 − 3d + 2
= < = |R| .
2 2
This is a contradiction, so it must be that ni 6= nj for all i 6= j, making f a valid group
assignment on Fd . Hence gr(Fd ) = d − 1.

16
6 Individual Problems

Problem 1. Charlie was born in the twentieth century. On his birthday in the present year (2014), he
notices that his current age is twice the number formed by the rightmost two digits of the year in
which he was born. Compute the four-digit year in which Charlie was born.

Problem 2. Let A, B, and C be randomly chosen (not necessarily distinct) integers between 0 and 4
inclusive. Pat and Chris compute the value of A + B · C by two different methods. Pat follows the
proper order of operations, computing A+(B ·C). Chris ignores order of operations, choosing instead
to compute (A + B) · C. Compute the probability that Pat and Chris get the same answer.

Problem 3. Bobby, Peter, Greg, Cindy, Jan, and Marcia line up for ice cream. In an acceptable lineup,
Greg is ahead of Peter, Peter is ahead of Bobby, Marcia is ahead of Jan, and Jan is ahead of Cindy.
For example, the lineup with Greg in front, followed by Peter, Marcia, Jan, Cindy, and Bobby, in
that order, is an acceptable lineup. Compute the number of acceptable lineups.

Problem 4. In triangle ABC, a = 12, b = 17, and c = 13. Compute b cos C − c cos B.

Problem 5. The sequence of words {an } is defined as follows: a1 = X, a2 = O, and for n ≥ 3, an

is an−1 followed by the reverse of an−2 . For example, a3 = OX, a4 = OXO, a5 = OXOXO, and
a6 = OXOXOOXO. Compute the number of palindromes in the first 1000 terms of this sequence.

Problem 6. Compute the smallest positive integer n such that 214·n and 2014·n have the same number
of divisors.

Problem 7. Let N be the least integer greater than 20 that is a palindrome in both base 20 and base 14.
For example, the three-digit base-14 numeral (13)5(13)14 (representing 13 · 142 + 5 · 141 + 13 · 140 ) is
a palindrome in base 14, but not in base 20, and the three-digit base-14 numeral (13)3114 is not a
palindrome in base 14. Compute the base-10 representation of N .

Problem 8. In triangle ABC, BC = 2. Point D is on AC such that AD = 1 and CD = 2. If

m∠BDC = 2m∠A, compute sin A.

C D A

1000

Problem 9. Compute the greatest integer k ≤ 1000 such that k is a multiple of 7.

17
Problem 10. An integer-valued function f is called tenuous if f (x) + f (y) > x2 for all positive integers
x and y. Let g be a tenuous function such that g(1) + g(2) + · · · + g(20) is as small as possible.
Compute the minimum possible value for g(14).

18
7 Answers to Individual Problems

Answer 1. 1938

9
Answer 2. 25

Answer 3. 20

Answer 4. 10

Answer 5. 667

Answer 6. 19133

Answer 7. 105

√
6
Answer 8. 4

Answer 9. 979

Answer 10. 136

19
8 Solutions to Individual Problems

Solution 1. Let N be the number formed by the rightmost two digits of the year in which Charlie was
born. Then his current age is 100 − N + 14 = 114 − N . Setting this equal to 2N and solving yields
N = 38, hence the answer is 1938.

Alternate Solution: Let N be the number formed by the rightmost two digits of the year in
which Charlie was born. The number of years from 1900 to 2014 can be thought of as the number of
years before Charlie was born plus the number of years since he was born, or N plus Charlie’s age.
Thus N + 2N = 114, which leads to N = 38, so the answer is 1938.

Solution 2. If Pat and Chris get the same answer, then A+(B ·C) = (A+B)·C, or A+BC = AC +BC,
or A = AC. This equation is true if A = 0 or C = 1; the equation places no restrictions on B. There
are 25 triples (A, B, C) where A = 0, 25 triples where C = 1, and 5 triples where A = 0 and C = 1.
As all triples are equally likely, the answer is 25+25−5
53
45
= 125 9
= 25 .

Solution 3. There are 6 people, so there are 6! = 720 permutations. However, for each arrangement of
the boys, there are 3! = 6 permutations of the girls, of which only one yields an acceptable lineup.
The same logic holds for the boys. Thus the total number of permutations must be divided by
3! · 3! = 36, yielding 6!/(3! · 3!) = 20 acceptable lineups.

Alternate Solution: Once the positions of Greg, Peter, and Bobby are determined, the entire
lineup is determined, because there is only one acceptable ordering of the three girls. Because the
boys occupy three of the six positions, there are 63 = 20 acceptable lineups.

Problem 4. In triangle ABC, a = 12, b = 17, and c = 13. Compute b cos C − c cos B.

Solution 4. Using the Law of Cosines, a2 + b2 − 2ab cos C = c2 implies

a2 + b2 − c2
b cos C = .
2a

20
Similarly,
a2 − b2 + c2
c cos B = .
2a
Thus
a2 + b2 − c2 a2 − b2 + c2
b cos C − c cos B = −
2a 2a
2b2 − 2c2
=
2a
b2 − c2
= .
a
With the given values, the result is (172 − 132 )/12 = 120/12 = 10.

Alternate Solution: Let H be the foot of the altitude from A to BC; let BH = x, CH = y,
and AH = h. Then b cos C = y, c cos B = x, and the desired quantity is Q = y − x. However,
y + x = a, so y 2 − x2 = aQ. By the Pythagorean Theorem, y 2 = b2 − h2 and x2 = c2 − h2 , so
2 2
y 2 − x2 = (b2 − h2 ) − (c2 − h2 ) = b2 − c2 . Thus aQ = b2 − c2 , and Q = b −c
a as in the first solution.

Problem 5. The sequence of words {an } is defined as follows: a1 = X, a2 = O, and for n ≥ 3, an

is an−1 followed by the reverse of an−2 . For example, a3 = OX, a4 = OXO, a5 = OXOXO, and
a6 = OXOXOOXO. Compute the number of palindromes in the first 1000 terms of this sequence.

Solution 5. Let P denote a palindromic word, let Q denote any word, and let R denote the reverse
of word R. Note that if two consecutive terms of the sequence are an = P , an+1 = Q, then
an+2 = QP = QP and an+3 = QP Q. Thus if an is a palindrome, so is an+3 . Because a1 and a2
are both palindromes, then so must be all terms in the subsequences a4 , a7 , a10 , . . . and a5 , a8 , a11 , . . . .

To show that the other terms are not palindromes, note that if P 0 is not a palindrome, then QP 0 Q is
also not a palindrome. Thus if an is not a palindrome, then an+3 is not a palindrome either. Because
a3 = OX is not a palindrome, neither is any term of the subsequence a6 , a9 , a12 , . . .. (Alternatively,
counting the number of X’s in each word ai shows that the number of X’s in a3k is odd. So if a3k
were to be a palindrome, it would have to have an odd number of letters, with an X in the middle.
However, it can be shown that the length of a3k is even. Thus a3k cannot be a palindrome.)

In total there are 1000 − 333 = 667 palindromes among the first 1000 terms.

Problem 6. Compute the smallest positive integer n such that 214·n and 2014·n have the same number
of divisors.

Solution 6. Let D(n) be the number of divisors of the integer n. Note that if D(214n) = D(2014n)
and if some p divides n and is relatively prime to both 214 and 2014, then D( 214n p ) = D( p ).
2014n

Thus any prime divisor of the smallest possible positive n will be a divisor of 214 = 2 · 107 or
2014 = 2 · 19 · 53. For the sake of convenience, write n = 2a−1 19b−1 53c−1 107d−1 , where a, b, c, d ≥ 1.
Then D(214n) = (a + 1)bc(d + 1) and D(2014n) = (a + 1)(b + 1)(c + 1)d. Divide both sides by a + 1
and expand to get bcd + bc = bcd + bd + cd + d, or bc − bd − cd − d = 0.

21
Because the goal is to minimize n, try d = 1: bc−b−c−1 = 0 ⇒ (b−1)(c−1) = 2, which has solutions
(b, c) = (2, 3) and (3, 2). The latter gives the smaller value for n, namely 192 · 53 = 19133. The only
quadruples (a, b, c, d) that satisfy 2a−1 19b−1 53c−1 107d−1 < 19133 and d > 1 are (1, 1, 2, 2), (1, 2, 1, 2),
and (1, 1, 1, 3). None of these quadruples satisfies bc − bd − cd − d = 0, so the minimum value is
n = 19133.

Solution 7. Because N is greater than 20, the base-20 and base-14 representations of N must be at
least two digits long. The smallest possible case is that N is a two-digit palindrome in both bases.
Then N = 20a + a = 21a, where 1 ≤ a ≤ 19. Similarly, in order to be a two-digit palindrome in base
14, N = 14b + b = 15b, with 1 ≤ b ≤ 13. So N would have to be a multiple of both 21 and 15. The
least common multiple of 21 and 15 is 105, which has the base 20 representation of 105 = 5520 and
the base-14 representation of 105 = 7714 , both of which are palindromes. Thus the answer is 105.

Problem 8. In triangle ABC, BC = 2. Point D is on AC such that AD = 1 and CD = 2. If

m∠BDC = 2m∠A, compute sin A.

C D A

2
Solution 8. Let [ABC] = K. Then [BCD] = 3 · K. Let DE be the bisector of ∠BDC, as shown below.

C D A

Notice that m∠DBA = m∠BDC − m∠A = m∠A, so triangle ADB is isosceles, and BD = 1.
(Alternately, notice that DE k AB, and by similar triangles, [CDE] = 94 · K, which means [BDE] =
2 ∼ CD
9 · K. Because [CDE] : [BDE] = 2 and ∠BDE = ∠CDE, conclude that BD = 2, thus BD = 1.)
1 1
Because BCD is isosceles, it follows that cos ∠BDC = 2 BD/CD = 4 . By the half-angle formula,
r r √
1 − cos ∠BDC 3 6
sin A = = = .
2 8 4

22
1000

Problem 9. Compute the greatest integer k ≤ 1000 such that k is a multiple of 7.

Solution 9. The ratio of binomial coefficients 1000

1000 k+1
k / k+1 = 1000−k . Because 1000 is 1 less than a
multiple of 7, namely 1001 = 7 · 11 · 13, either 1000 − k and k + 1 are both multiples of 7 or neither is.
Hence whenever the numerator is divisible by 7, the denominator is also. Thus for the largest value of
k such that 1000
k
k+1
is a multiple of 7, 1000−k must equal 7· pq , where p and q are relatively prime integers
and 7 - q. The only way this can happen is when k + 1 is a multiple of 49, the greatest of which
less than 1000 is 980. Therefore the greatest value of k satisfying the given conditions is 980−1 = 979.

Alternate Solution:
Rewrite 1000 in base 7: 1000 = p2626 7 . Let k = a b c d7 . By Lucas’s Theorem,
1000 2 6 2 6

k ≡ a b c d mod 7. The binomial coefficient q = 0 only when q > p. Base 7 digits cannot
exceed 6, and k ≤ 1000, thus the greatest value of k that works is 25667 = 979. (Alternatively, the
least value of k that works is 307 = 21; because nk = n−k n
, the greatest such k is 1000 − 21 = 979.)

Problem 10. An integer-valued function f is called tenuous if f (x) + f (y) > x2 for all positive integers
x and y. Let g be a tenuous function such that g(1) + g(2) + · · · + g(20) is as small as possible.
Compute the minimum possible value for g(14).

Solution 10. For a tenuous function g, let Sg = g(1) + g(2) + · · · + g(20). Then:

Sg = (g(1) + g(20)) + (g(2) + g(19)) + · · · + (g(10) + g(11))

≥ (202 + 1) + (192 + 1) + · · · + (112 + 1)
20
X
= 10 + k2
k=11
= 2495.

The following argument shows that if a tenuous function g attains this sum, then g(1) = g(2) = · · · =
g(10). First, if the sum equals 2495, then g(1) + g(20) = 202 + 1, g(2) + g(19) = 192 + 1, . . . , g(10) +
g(11) = 112 + 1. If g(1) < g(2), then g(1) + g(19) < 192 + 1, which contradicts the tenuousness of g.
Similarly, if g(2) > g(1), then g(2) + g(20) < 202 + 1. Therefore g(1) = g(2). Analogously, comparing
g(1) and g(3), g(1) and g(4), etc. shows that g(1) = g(2) = g(3) = · · · = g(10).

Now consider all functions g for which g(1) = g(2) = · · · = g(10) = a for some integer a. Then
g(n) = n2 + 1 − a for n ≥ 11. Because g(11) + g(11) > 112 = 121, it is the case that g(11) ≥ 61.
Thus 112 + 1 − a ≥ 61 ⇒ a ≤ 61. Thus the smallest possible value for g(14) is 142 + 1 − 61 = 136.

23
9 Relay Problems

Relay 1-1 Let T = (0, 0), N = (2, 0), Y = (6, 6), W = (2, 6), and R = (0, 2). Compute the area of
pentagon T N Y W R.

Relay 1-2 Let T = T N Y W R. The lengths of the sides of a rectangle are the zeroes of the polynomial
x2 − 3T x + T 2 . Compute the length of the rectangle’s diagonal.

Relay 1-3 Let T = T N Y W R. Let w > 0 be a real number such that T is the area of the region above
the x-axis, below the graph of y = dxe2 , and between the lines x = 0 and x = w. Compute d2we.

Relay 2-1 Compute the least positive integer n such that gcd(n3 , n!) ≥ 100.

Relay 2-2 Let T = T N Y W R. At a party, everyone shakes hands with everyone else exactly once, except
Ed, who leaves early. A grand total of 20T handshakes take place. Compute the number of people
at the party who shook hands with Ed.

Relay 2-3 Let T = T N Y W R. Given the sequence un such that u3 = 5, u6 = 89, and un+2 = 3un+1 − un
for integers n ≥ 1, compute uT .

24
10 Relay Answers

Answer 1-1 20

√
Answer 1-2 20 7

Answer 1-3 10

Answer 2-1 8

Answer 2-2 7

Answer 2-3 233

25
11 Relay Solutions

Relay 1-1 Let T = (0, 0), N = (2, 0), Y = (6, 6), W = (2, 6), and R = (0, 2). Compute the area of
pentagon T N Y W R.

Solution 1-1 Pentagon T N Y W R fits inside square T AY B, where A = (6, 0) and B = (0, 6). The region
of T AY B not in T N Y W R consists of triangles 4N AY and 4W BR, as shown below.

B W Y

T N A

Thus

[T N Y W R] = [T AY B] − [N AY ] − [W BR]
1 1
= 62 − · 4 · 6 − · 2 · 4
2 2
= 20.

Relay 1-2 Let T = T N Y W R. The lengths of the sides of a rectangle are the zeroes of the polynomial
x2 − 3T x + T 2 . Compute the length of the rectangle’s diagonal.

Solution 1-2√ Let r and p s denote the zeros of the polynomial x2 − 3T x + T 2 . The rectangle’s diagonal has
length r + s = (r + s)2 − 2rs. Recall that for a quadratic polynomial ax2 + bx + c, the sum
2 2

of its zeros is −b/a, and the product of its zeros is c/a. In

√ this particular instance,
√ r + s = 3T and
2 2 2
rs = T . Thus the length √ of the rectangle’s diagonal is 9T − 2T = T · 7. With T = 20, the
rectangle’s diagonal is 20 7.

Relay 1-3 Let T = T N Y W R. Let w > 0 be a real number such that T is the area of the region above
the x-axis, below the graph of y = dxe2 , and between the lines x = 0 and x = w. Compute d2we.

Solution 1-3 Write w = k + α, where k is an integer, and 0 ≤ α < 1. Then

T = 12 + 22 + · · · + k 2 + (k + 1)2 · α.

Computing
√ d2we requires
√ computing w to the nearest half-integer. First obtain the integer k. As
7 > 2, with T = 20 7, one obtains T > 40. As 12 + 22 + √ 32 + 42 = 30, it follows
√ that k ≥ 4.
To obtain an upper bound for k, note that 700 < 729, so 10 7 < 27, and T = 20 7 < 54. As
12 + 22 + 32 + 42 + 52 = 55, it follows that 4 < w < 5, and hence k = 4.

26
It now suffices to determine whether or not α > 0.5. To this end, one must determine whether
T > 12 + 22 + 32 + 42 + 52 /2 = 42.5. Indeed, note that 2.52 = 6.25 < 7, so T > (20)(2.5) = 50. It
follows that α > 0.5, so 4.5 < w < 5. Thus 9 < 2w < 10, and d2we = 10.

Alternate Solution: Once it has been determined that 4 < w√ < 5, the√ formula for T yields
√ √
4 7−6 8 7−12
1 + 4 + 9 + 16 + 25 · α = 20 7, hence α = 5 . Thus 2α = 5 = 448−12
5 > 21−12
5 = 1.8.
Because 2w = 2k + 2α, it follows that d2we = d8 + 2αe = 10, because 1.8 < 2α < 2.

Relay 2-1 Compute the least positive integer n such that gcd(n3 , n!) ≥ 100.

Solution 2-1 Note that if p is prime, then gcd(p3 , p!) = p. A good strategy is to look for values of n with
several (not necessarily distinct) prime factors so that n3 and n! will have many factors in common.
For example, if n = 6, n3 = 216 = 23 · 33 and n! = 720 = 24 · 32 · 5, so gcd(216, 720) = 23 · 32 = 72.
Because 7 is prime, try n = 8. Notice that 83 = 29 while 8! = 27 · 32 · 5 · 7. Thus gcd(512, 8!) = 27 =
128 > 100, hence the smallest value of n is 8.

Solution 2-2 If there were n people at the party, including Ed, and if Ed had not left early, there would
n

have been 2 handshakes. Because Ed left early, the number of handshakes is strictly less than that,
but greater than n−1

2 (everyone besides Ed shook everyone else’s hand). So find the least number
n such that n2 ≥ 160. The least such n is 19, because 18 19

2 = 153 and 2 = 171. Therefore there
were 19 people at the party. However, 171 − 160 = 11 handshakes never took place. Therefore the
number of people who shook hands with Ed is 19 − 11 − 1 = 7.

Relay 2-3 Let T = T N Y W R. Given the sequence un such that u3 = 5, u6 = 89, and un+2 = 3un+1 − un
for integers n ≥ 1, compute uT .

Solution 2-3 By the recursive definition, notice that u6 = 89 = 3u5 − u4 and u5 = 3u4 − u3 = 3u4 − 5.
This is a linear system of equations. Write 3u5 − u4 = 89 and −3u5 + 9u4 = 15 and add to obtain
u4 = 13. Now apply the recursive definition to obtain u5 = 34 and u7 = 233.

Alternate Solution: Notice that the given values are both Fibonacci numbers, and that in the
Fibonacci sequence, f1 = f2 = 1, f5 = 5, and f11 = 89. That is, 5 and 89 are six terms apart in
the Fibonacci sequence, and only three terms apart in the given sequence. This relationship is not
a coincidence: alternating terms in the Fibonacci sequence satisfy the given recurrence relation for
the sequence {un }, that is, fn+4 = 3fn+2 − fn . Proof: if fn = a and fn+1 = b, then fn+2 = a + b,
fn+3 = a + 2b, and fn+4 = 2a + 3b = 3(a + b) − b = 3fn+2 − fn . To compute the final result, continue
out the Fibonacci sequence to obtain f12 = 144 and u7 = f13 = 233.

27
12 Super Relay
1. The sequence a1 , a2 , a3 , . . . is a geometric sequence with a20 = 8 and a14 = 221 . Compute a21 .

2. Let T = T N Y W R. Circles L and O are internally tangent and have radii T and 4T , respectively.
Point E lies on circle L such that OE is tangent to circle L. Compute OE.

3. Let T = T N Y W R. In a right triangle, one leg has length T 2 and the other leg is 2 less than the
hypotenuse. Compute the triangle’s perimeter.

4. Let T = T N Y W R. If x + 9y = 17 and T x + (T + 1)y = T + 2, compute 20x + 14y.

5. Let T = T N Y W R. Let f (x) = ax2 + bx + c. The product of the roots of f is T . If (−2, 20) and
(1, 14) lie on the graph of f , compute a.

6. Let T = T N Y W R. Let z1 = 15 + 5i and z2 = 1 + Ki. Compute the smallest positive integral value
of K such that |z1 − z2 | ≥ 15T .

7. Let T = T N Y W R. Suppose that T people are standing in a line, including three people named
Charlie, Chris, and Abby. If the people are assigned their positions in line at random, compute the
probability that Charlie is standing next to at least one of Chris or Abby.

15. Compute the smallest positive integer N such that 20N is a multiple of 14 and 14N is a multiple of
20.

14. Let T = T N Y W R. Chef Selma is preparing a burrito menu. A burrito consists of: (1) a choice of
chicken, beef, turkey, or no meat, (2) exactly one of three types of beans, (3) exactly one of two types
of rice, and (4) exactly one of K types of cheese. Compute the smallest value of K such that Chef
Selma can make at least T different burrito varieties.
√
13. Let T = T N Y W R. Regular hexagon SU P ERB has side length T . Compute the value of BE ·
SU · RE.
rq
p
T
12. Let T = T N Y W R. Compute 10T 2 −T .

11. Let T = T N Y W R. Nellie has a flight from Rome to Athens that is scheduled to last for T + 30
minutes. However, owing to a tailwind, her flight only lasts for T minutes. The plane’s speed is 1.5
miles per minute faster than what it would have been for the originally scheduled flight. Compute
the distance (in miles) that the plane travels.

10. Let T = T N Y W R. If log T = 2 − log 2 + log k, compute the value of k.

9. Let T = T N Y W R. If r is the radius of a right circular cone and the cone’s height is T − r2 , let V
be the maximum possible volume of the cone. Compute π/V .

8. Let A be the number you will receive from position 7 and let B be the number you will receive from
position 9. Let α = sin−1 A and let β = cos−1 B. Compute sin(α + β) + sin(α − β).

28
13 Super Relay Answers
1. 1
√
2. 2 2

3. 40

4. 8

5. 8/5

6. 25

7. 47/300
15. 70

14. 3

13. 9

12. 100

11. 650

10. 13

9. 12/169
8. 94/4225

29
14 Super Relay Solutions
Problem 1. The sequence a1 , a2 , a3 , . . . is a geometric sequence with a20 = 8 and a14 = 221 . Compute a21 .
23
Solution 1. Let r be the common ratio of the sequence. Then a20 = r20−14 · a14 , hence 8 = r6 · 221 ⇒ r6 = 221
=
2−18 , so r = 2−3 = 18 . Thus a21 = r · a20 = 81 · 8 = 1.

Problem 2. Let T = T N Y W R. Circles L and O are internally tangent and have radii T and 4T , respectively.
Point E lies on circle L such that OE is tangent to circle L. Compute OE.

p LE ⊥ OE. Also√note that LO = 4T − T = 3T . Hence, by

Solution 2. Because OE is tangent to circle L,
2 2
the Pythagorean Theorem, OE = √ (3T ) − T = 2T 2 (this also follows from the Tangent-Secant
Theorem). With T = 1, OE = 2 2.

Problem 3. Let T = T N Y W R. In a right triangle, one leg has length T 2 and the other leg is 2 less than the
hypotenuse. Compute the triangle’s perimeter.

Solution 3. Let c be the length of the hypotenuse. Then, by the Pythagorean Theorem, (T 2 )2 + (c − 2)2 = c2 ⇒
4 √
c = T4 + 1. With T = 2 2, T 4 = 64, and c = 17. So the triangle is a 8−15−17 triangle with
perimeter 40.

Problem 4. Let T = T N Y W R. If x + 9y = 17 and T x + (T + 1)y = T + 2, compute 20x + 14y.

Solution 4. Multiply each side of the first equation by T to obtain T x+9T y = 17T . Subtract the second equation
to yield 9T y − T y − y = 16T − 2 ⇒ y(8T − 1) = 2(8T − 1). Hence either T = 18 (in which case, the
value of y is not uniquely determined) or y = 2. Plug y = 2 into the first equation to obtain x = −1.
Hence 20x + 14y = −20 + 28 = 8.

Problem 5. Let T = T N Y W R. Let f (x) = ax2 + bx + c. The product of the roots of f is T . If (−2, 20) and
(1, 14) lie on the graph of f , compute a.

Solution 5. Using Vièta’s Formula, write f (x) = ax2 + bx + T a. Substituting the coordinates of the given points
yields the system of equations: 4a − 2b + T a = 20 and a + b + T a = 14. Multiply each side of the
latter equation by 2 and add the resulting equation to the former equation to eliminate b. Simplifying
yields a = T16
+2 . With T = 8, a = 8/5.

Problem 6. Let T = T N Y W R. Let z1 = 15 + 5i and z2 = 1 + Ki. Compute the smallest positive integral value
of K such that |z1 − z2 | ≥ 15T .
p
Solution 6. Note that z1 − z2 = 14 + (5 − K)i, hence |z1√− z2 | = 14√ 2 + (5 − K)2 . With T = 8/5, 15T = 24,
2 2 2 2 2
hence 14 + (5 − K) ≥ 24 . Thus |5 − K| ≥ 24 − 14 = 380. Because K is a positive integer, it
follows that K − 5 ≥ 20, hence the desired value of K is 25.

Problem 7. Let T = T N Y W R. Suppose that T people are standing in a line, including three people named
Charlie, Chris, and Abby. If the people are assigned their positions in line at random, compute the
probability that Charlie is standing next to at least one of Chris or Abby.

Solution 7. First count the number of arrangements in which Chris stands next to Charlie. This is (T − 1) · 2! ·
(T − 2)! = 2 · (T − 1)! because there are T − 1 possible leftmost positions for the pair {Charlie, Chris},
there are 2! orderings of this pair, and there are (T − 2)! ways to arrange the remaining people.
There are equally many arrangements in which Abby stands next to Charlie. However, adding these

30
overcounts the arrangements in which Abby, Charlie, and Chris are standing next to each other, with
Charlie in the middle. Using similar reasoning as above, there are (T − 2) · 2! · (T − 3)! = 2 · (T − 2)!
such arrangements. Hence the desired probability is 2·2·(T −1)!−2·(T
T!
−2)!
= 2·(T −2)!(2T
T!
−2−1)
= 2(2T −3)
T (T −1) .
47
With T = 25, the fraction simplifies to 300 .

Problem 15. Compute the smallest positive integer N such that 20N is a multiple of 14 and 14N is a multiple of
20.

Solution 15. Because gcd(14, 20) = 2, the problem is equivalent to computing the smallest positive integer N such
that 7 | 10N and 10 | 7N . Thus 7 | N and 10 | N , and the desired value of N is lcm(7, 10) = 70.

Problem 14. Let T = T N Y W R. Chef Selma is preparing a burrito menu. A burrito consists of: (1) a choice of
chicken, beef, turkey, or no meat, (2) exactly one of three types of beans, (3) exactly one of two types
of rice, and (4) exactly one of K types of cheese. Compute the smallest value of K such that Chef
Selma can make at least T different burrito varieties.

Solution 14. Using the Multiplication Principle, Chef Selma can make 4 · 3 · 2 · K = 24K
70 different burrito varieties.
With T = 70, the smallest integral value of K such that 24K ≥ 70 is 24 = 3.
√
Problem 13. Let T = T N Y W R. Regular hexagon SU P ERB has side length T . Compute the value of BE ·
SU · RE.
√
Solution 13. Because SU and RE are sides of the hexagon, SU = RE = T . Let H be the foot of the altitude
from R to BE in 4BRE √ and
√ note
that each interior angle of a regular hexagon is 120◦ . Thus
√ √ √ √ √
BE = BH + HE = 2 23 T = 3T . Thus BE · SU · RE = 3T · T · T = T 3T . With
T = 3, the answer is 9.
rq
pT
Problem 12. Let T = T N Y W R. Compute 10T 2 −T .
!12

2 −T
1 12
T
Solution 12. The given radical equals 10T = 10(T −1)/4 . With T = 9, this simplifies to 102 = 100.

Problem 11. Let T = T N Y W R. Nellie has a flight from Rome to Athens that is scheduled to last for T + 30
minutes. However, owing to a tailwind, her flight only lasts for T minutes. The plane’s speed is 1.5
miles per minute faster than what it would have been for the originally scheduled flight. Compute
the distance (in miles) that the plane travels.
D D
Solution 11. Let D be the distance in miles traveled by the plane. The given conditions imply that T − T +30 =
30D T (T +30)
1.5 ⇒ T (T +30) = 1.5 ⇒ D = 20 . With T = 100, D = 5 · 130 = 650.

Problem 10. Let T = T N Y W R. If log T = 2 − log 2 + log k, compute the value of k.

Solution 10. Write 2 = log 100 and use theT well-known properties for the sum/difference of two logs to obtain
100k
log T = log 2 , hence k = 50 . With T = 650, k = 13.

Problem 9. Let T = T N Y W R. If r is the radius of a right circular cone and the cone’s height is T − r2 , let V
be the maximum possible volume of the cone. Compute π/V .

31
Solution 9. The cone’s volume is 31 πr2 (T − r2 ). Maximizing this is equivalent to maximizing x(T − x), where
x = r2 . Using the formula for the vertex of a parabola (or the AM-GM inequality), the maximum
2
value occurs when x = T2 . Hence V = 13 π · T2 · T2 = πT 2 12
12 , and π/V = 12/T . With T = 13, V = 169 .

Problem 8. Let A be the number you will receive from position 7 and let B be the number you will receive from
position 9. Let α = sin−1 A and let β = cos−1 B. Compute sin(α + β) + sin(α − β).

Solution 8. The given conditions are equivalent to sin α = A and cos β = B. Using either the sum-to-product or
the sine of a sum/difference identities, the desired expression is equivalent to 2(sin α)(cos β) = 2·A·B.
47 12 2·47 94
With A = 300 and B = 169 , 2 · A · B = 25·169 = 4225 .

32
15 Tiebreaker Problems

Problem 1. A student computed the repeating decimal expansion of N1 for some integer N , but inserted
six extra digits into the repetend to get .00231846597. Compute the value of N .

Problem 2. Let n be a four-digit number whose square root is three times the sum of the digits of n.
Compute n.

Problem 3. Compute the sum of the reciprocals of the positive integer divisors of 24.

33
16 Tiebreaker Answers

Answer 1. 606

Answer 2. 2916

Answer 3. 5/2

34
17 Tiebreaker Solutions

Problem 1. A student computed the repeating decimal expansion of N1 for some integer N , but inserted
six extra digits into the repetend to get .00231846597. Compute the value of N .

1
Solution 1. Because the given repetend has ten digits, the original had four digits. If N = .0A B C D =
ABC D 2
99990 , then the numerator must divide 99990 = 10 · 99 · 101 = 2 · 3 · 5 · 11 · 101.

Note that all 3- and 4-digit multiples of 101 contain at least one digit which appears twice. Because
the 10-digit string under the vinculum (i.e., 0231846597) contains no repeated digits, A B C D cannot
be a multiple of 101. So A B C D divides 2·32 ·5·11 = 990. The only divisor of 990 that can be formed
from four of the given digits (taken in order) is 0165, that is, 165. Hence N1 = 99990
165 1
= 606 ⇒ N = 606.

Problem 2. Let n be a four-digit number whose square root is three times the sum of the digits of n.
Compute n.

√
Solution 2. Because n is a multiple of 3, n must be a multiple of 9. Therefore the sum of the digits
√
of n is a multiple of 9. Thus n must be a multiple of 27, which implies that n is a multiple of
272 . The only candidates to consider are 542 (= 2916) and 812 (= 6561), and only 2916 satisfies the
desired conditions.

Problem 3. Compute the sum of the reciprocals of the positive integer divisors of 24.

Solution 3. The map n 7→ 24/n establishes a one-to-one correspondence among the positive integer
divisors of 24. Thus
X1 X 1
=
n 24/n
n|24 n|24
n>0 n>0
1 X
= n.
24
n|24
n>0

Because 24 = 23 · 3, the sum of the positive divisors of 24 is (1 + 2 + 22 + 23 )(1 + 3) = 15 · 4 = 60.

Hence the sum is 60/24 = 5/2.

Alternate Solution: Because 24 = 23 · 3, any positive divisor of 24 is of the form 2a 3b where

a = 0, 1, 2, or 3, and b = 0 or 1. So the sum of the positive divisors of 24 can be represented as
the product (1 + 2 + 4 + 8)(1 +3). Similarly, the sum of their reciprocals can be represented as the
product 11 + 12 + 14 + 81 11 + 13 . The first sum is 15
8 and the second is 4
3 so the product is 5/2.
,

35
ARML Competition 2015

Paul J. Karafiol, Head Writer

Andy Soffer, Associate Head Writer
Chris Jeuell, Lead Editor
Paul Dreyer
Edward Early
Zuming Feng
Zachary Franco
Winston Luo
Andy Niedermaier
George Reuter
Andy Soffer
Eric Wepsic

May 29–30, 2015

1 Team Problems

p √ k
rj
Problem 1. Compute the greatest integer x such that b xc = 2.

1 540
Problem 2. Compute the number of ordered pairs of integers (x, y) such that + = 2.
x xy

Problem 3. A positive integer has the Kelly Property if it contains a zero in its base-17 representation. Compute
the number of positive integers less than 1000 (base 10) that have the Kelly Property.

Problem 4. Compute the smallest positive integer n such that n + i, (n + i)2 , and (n + i)3 are the vertices of a
triangle in the complex plane whose area is greater than 2015.

Problem 5. A loop is made by connecting rods of lengths 1, 2, . . . , 90 in that order. (The rod of length 90 is
connected to the rods of lengths 89 and 1.) The loop is laid in the shape of an equilateral triangle of perimeter
4095. Rods cannot be bent or broken. Compute the sum of the lengths of the shortest rods on each side of
the triangle. (For example, the loop with rods 1, 2, . . . , 9 can be arranged into an equilateral triangle because
4 + 5 + 6 = 7 + 8 = 9 + 1 + 2 + 3.)

Problem 6. Four spheres S1 , S2 , S3 , and S4 are mutually externally tangent and are tangent to a plane, on the
same side of the plane. Let S1 and S2 have radius r and let S3 and S4 have radius s. Given that r > s,
compute r/s.

Problem 7. In 4ABC, point D is on AB and point E is on AC. The measures of the nine angles in triangles
ADE, BCD, and CDE can be arranged to form an arithmetic sequence. Compute the greatest possible degree
measure for ∠A.

Problem 8. Let a1 = a2 = a3 = 1. For n > 3, let an be the number of real numbers x such that

x4 − 2an−1 x2 + an−2 an−3 = 0.

Compute the sum a1 + a2 + a3 + · · · + a1000 .

Problem 9. For any real number k, let region Rk consist of all points (x, y) such that x ≥ 0, y ≥ 0, and
bx + yc + {x} ≤ k, where {u} denotes the fractional part of u. Compute the value of k for which the area of
Rk is equal to 100.

−−→
Problem 10. Let ABCD be a parallelogram with m∠A > 90◦ . Point E lies on DA such that BE ⊥ AD. The
circumcircles of 4ABC and 4CDE intersect at points F and C. Given that AD = 35, DC = 48, and
CF = 50, compute AC.

1
2 Answers to Team Problems

Answer 1. 6560

Answer 2. 15

Answer 3. 106

Answer 4. 9

Answer 5. 72

√
Answer 6. 2 + 3

Answer 7. 84

Answer 8. 2329

√
Answer 9. 29 − 221

√
Answer 10. 5 23

2
3 Solutions to Team Problems

p √ k
rj
Problem 1. Compute the greatest integer x such that b xc = 2.

2 2 2
Solution 1. Ignore the floor functions and replace 2 with 3. Then x would
rjbe ((3 )k) = 6561. If x < 6561,
√ √ p √ jp √ k p √
then x < 81, so b xc ≤ 80, and b xc < 9, so b xc ≤ 8, and b xc ≤ 2. Thus the greatest
possible integral value of x is 6561 − 1 = 6560.

1 540
Problem 2. Compute the number of ordered pairs of integers (x, y) such that + = 2.
x xy

Solution 2. Multiply through by xy and rearrange terms to obtain 540 = 2xy − y = (2x − 1) · y. Because 2x − 1
is odd, the question reduces to counting the number of odd divisors (positive and negative) of 540. Because
540 = 22 · 33 · 5, there are (3 + 1)(1 + 1) = 8 positive odd divisors, hence 540 has 16 odd integer divisors and
16 potential integral values for x. However, the divisor −1 corresponds to x = 0, which is extraneous to the
original equation. Hence there are 15 solutions, given in the tables below.

2x − 1 −135 −45 −27 −15 −9 −5 −3

x −67 −22 −13 −7 −4 −2 −1
y −4 −12 −20 −36 −60 −108 −180

2x − 1 1 3 5 9 15 27 45 135
x 1 2 3 5 8 14 23 68
y 540 180 108 60 36 20 12 4

Solution 3. Compute the base-17 representation of 1000:

1000 = 3 · 172 + 7 · 171 + 14 · 170 .

Let yp denote the number of integers less than 1000 with the Kelly Property whose 172 -digit is p. If p = 1
or p = 2, then so long as one of the two rightmost digits is zero, the number will have the Kelly Property.
There are 172 − 162 such numbers because 162 of the 172 options have non-zero values in both of the two
rightmost digits. Thus y1 = y2 = 172 − 162 = 33. For p = 0, the only possibility for a number with the Kelly
Property is for its 1s digit to be zero. There are 16 such numbers, because the 17s digit can be any non-zero
value. Thus y0 = 16. Finally consider the case p = 3. If the 17s digit is zero, then any value for the 1s digit
will yield a number with the Kelly Property, contributing another 17 numbers with the Kelly Property. If the
17s digit is a number 1 through 7 inclusive, then to have the Kelly Property, the number must have a zero in
its 1s digit, contributing another 7 numbers with the Kelly Property. Thus y3 = 17 + 7 = 24.

The number of positive integers less than 1000 which have the Kelly Property is therefore

y0 + y1 + y2 + y3 = 16 + 33 + 33 + 24 = 106.

3
Problem 4. Compute the smallest positive integer n such that n + i, (n + i)2 , and (n + i)3 are the vertices of a
triangle in the complex plane whose area is greater than 2015.

Solution 4. The complex number (n + i)2 can be broken into real and imaginary parts as n2 + 2ni + i2 =
(n2 − 1) + 2ni. The complex number (n + i)3 can broken into real and imaginary parts as n3 + 3n2 i + 3ni2 + i3 =
(n2 − 3n) + (3n2 − 1)i. Therefore the triangle has the same area in the complex plane as the triangle in the
Cartesian plane with coordinates (n, 1), (n2 − 1, 2n), and (n3 − 3n, 3n2 − 1). The Shoelace Theorem gives that
the area of this triangle is the absolute value of
1 n 1 1 n2 − 1 2n 1 n3 − 3n 3n2 − 1
2 + 3 2 +
2 n − 1 2n 2 n − 3n 3n − 1 2 n 1
1
= (2n2 + 3n4 − n2 − 3n2 + 1 + n3 − 3n − n2 + 1 − 2n4 + 6n2 − 3n3 + n)
2
1
= (n4 − 2n3 + 3n2 − 2n + 2).
2
Therefore the problem reduces to finding the smallest positive n such that f (n) = n4 − 2n3 + 3n2 − 2n + 2 >
4030. Notice that the dominant term is n4 , and so n = 8 gives n4 = 4096, which is a good approximation to
4030. However, substituting n = 8 gives f (n) = |4096 − 2 · 512 + 3 · 64 − 2 · 8 + 2| = 3250, which is too small.
Trying n = 9 gives f (n) = |6561 − 2 · 729 + 3 · 81 − 2 · 9 + 2| = 5330 > 4030, so the answer is n = 9.

Alternate Solution: Let z = (n + i), and let K1 be the area of the triangle whose vertices are 0, z, and z 2 ,
let K2 be the area of the triangle whose vertices are 0, z 2 , and z 3 , and let K3 be the area of the triangle whose
vertices are 0, z, and z 3 . Then if K is the area of the triangle with vertices z, z 2 , and z 3 , K = K1 + K2 − K3 .

First compute K1 . Let θ be the angle that n + i makes with the x-axis. By DeMoivre’s Theorem, the
angle between the ray from 0 through z and the ray from 0 through z 2 is also θ, so using the triangle area
formula 21 ab sin C yields
1
K1 = · |n + i| · (n + i)2 · sin θ.
2
Because absolute values are multiplicative, the expression on the right-hand side simplifies to
1 3
K1 = · |n + i| · sin θ,
2
√
and because |n + i| = n2 + 1,
1
K1 = · (n2 + 1)3/2 sin θ.
2
Now sin θ = √ 1 , so
n2 +1
n2 + 1
K1 = .
2
(n2 +1)2
Similarly, K2 = 2 .

Computing K3 is not much more involved. The angle between z and z 3 is 2θ. Using the trigonometric
identity sin 2x = 2 sin x cos x and the observation that cos θ = √nn2 +1 yields

sin 2θ = 2
sin θ cos θ
1 n
= 2 √ √
n2 + 1 n2 + 1
2n
= .
n2 + 1
Thus
1 2n
K3 = · (n2 + 1)2 · 2 = n(n2 + 1).
2 n +1

4
Hence

K = K1 + K2 − K3
n2 + 1 (n2 + 1)2
= + − n(n2 + 1)
2 2
1 n2 + 1

= n2 + 1 + −n
2 2
2
2
n − 2n + 2
= n +1
2
1 2
n + 1 (n − 1)2 + 1 .

=
2
To compute
the smallest
positive integral value of n such that K > 2015, first multiply by 2 to obtain
n2 + 1 (n − 1)2 + 1 > 4030. The two factors are approximately n2 and (n − 1)2 , so look for n such
that n4 > 4030. Because 84 = 212 = 4096, try n = 8 to obtain a product of (64 + 1)(49 + 1) = 3250 which is
too small; n = 9 yields a product of (81 + 1)(64 + 1) = 5330. Thus n = 9.

Solution 5. The side length of the triangle is 31 · 91·90

2 = 91·15 = 3·5·7·13 = 1365. Call this value S. Start walking
around the triangle at the length-1 rod, going in the direction away from the length-90 rod. Let the last rods
of each of the three sides have lengths a, b, and c, with a < b < c. (In the 9-rod example, (a, b, c) = (3, 6, 8)).
Thus the three sides are composed of rods a + 1, . . . , b; b + 1, . . . c; and c + 1, . . . , 90, 1, . . . , a.

Focus on the side that ends in the length-c rod. It has c − b rods in total, and S = (c − b)( b+c+1 2 ). If
S
c − b is odd, then it divides S. If c − b is even, then it divides 2S. If c − b ≤ 15, then c−b — the average length
of the rods on that side — will be at least 91, which is too big.

S
If c − b = 21, then c−b = 65, so this side must have rods 55, . . . , 75. That would imply b = 54. Does there
exist an a such that S = (54 − a)( a+54+1
2 )? As it turns out, a = 15 is a solution. Thus (a, b, c) = (15, 54, 75)
produces an equilateral triangle.

S
If c − b = 26, then c−b = 52.5, so this side must have rods 40, . . . , 65. But note that this means the side
after this must contain rods 66 through 90, as well as the length-1 rod and possibly more after that. The total
length of those rods is at least 78 · 25 + 1 > S, which is too large.

Thus for all values of c − b that divide 2S except 21, there will not exist an equilateral triangle satisfying
the conditions of the problem. Thus the only possible value for the sum of the lengths of the shortest rods of
the three sides is 1 + 16 + 55 = 72.

Remark: This problem admits several interesting generalizations, three of which are discussed below.

No squares. If (A, B, C) = (2a + 1, 2b + 1, 2c + 1), then it must be the case that 8S = B 2 − A2 = C 2 − B 2 .

That is, A2 , B 2 , C 2 must form an arithmetic progression. In general, to create a regular k-gon with the n-loop
(rods of lengths 1, 2, . . . , n), then it is necessary to be able to create arithmetic progressions of k odd perfect
squares. It so happens that there are no such progressions for k ≥ 4, a result shown by Euler. Therefore it is
not possible make the loop into a square (or larger k-gon) for any integer n.

Equilateral triangles are hard to find. The values of n less than 1000 for which the n-loop can be made

5
into an equilateral triangle are 9, 90, 125, 153, 189, 440, 819, and 989. In the n = 125 case, there are two
different ways to create an equilateral triangle! Can you determine whether any other value n has this property?

There are infinitely many such equilateral triangles, though. One can use Diophantine equations to find an
infinite family of values of n that allow for equilateral triangles.

Solution 6. Let the centers of S1 and S2 be at (r, r, 0) and (−r, r, 0), so that they are tangent to the plane y = 0
at points (r, 0, 0) and (−r, 0, 0). Then the centers of S3 and S4 lie on the plane x = 0, at (0, s, −s) and (0, s, s).
Because all of the spheres are mutually tangent, the distance between the center of S1 and the center of S4
is r + s, so the 3-dimensional Pythagorean Theorem gives
(r + s)2 = r2 + (r − s)2 + s2
which simplifies to
r2 − 4rs + s2 = 0.
Solving this as a quadratic in r gives √
r = (2 ± 3)s.
r
√
Because r > s, it follows that s =2+ 3.

Alternate Solution: Let ω be the given plane, and let the centers of the four spheres be P1 , P2 , P3 ,
and P4 respectively. Because S1 and S2 have equal radii and S3 and S4 have equal radii, note that P1 P2
and P3 P4 are parallel to ω. Note also that P1 P2 = 2r, P3 P4 = 2s, and P1 P3 = P1 P4 = P2 P3 = P2 P4 = r + s.
Let R be the midpoint of P3 P4 , and let Q denote the point in the plane parallel to ω containing P3 and P4 , such
←−→ √
that QP1 ⊥ ω. Then 4QRS3 is a right triangle with legs of length r and s, and hypotenuse QS3 = r2 + s2 .
But 4QP1 P3 is also a right triangle, with hypotenuse P1 P3 . Because QP1 = r − s, the Pythagorean Theorem
2
yields (r − s)2 + r2 + s2 = (r + s)2 . Hence r2 + s2 = 4rs. Divide by s2 to obtain the equation rs2 + 1 = 4 · rs ;
√
substitute u = rs to obtain the quadratic equation u2 − 4u + 1 = 0. Thus u = 2 ± 3. It is given that r > s,
√
hence rs = 2 + 3.

Solution 7. The average value of the angles is 60◦ , so the nine terms must be 60◦ ± nd, where d is the common
difference and 0 ≤ n ≤ 4. Thus d < 15◦ . Looking at the supplementary angles at E, one angle is 60◦ + kd, and
◦
the other is 60◦ + `d. Let m = k + `. Then d = 60 m . Combining these facts yields only three possible values
for m, namely, m ∈ {5, 6, 7}.

If d = 10◦ , then the supplementary angles must be 80◦ and 100◦ , in some order, and so m∠A ≤ 90◦ . But
then m∠ADE = 10◦ , which is impossible. One can successfully fill the diagram with m∠A = 70◦ , which is the
maximum for d = 10◦ .

If d = 12◦ , then the supplementary angles can be {72◦ , 108◦ } or {84◦ , 96◦ }. Note that m∠A cannot be
108◦ , so try 96◦ . Then m∠AED = 72◦ and m∠ADE = 12◦ . But m∠CDE + m∠BDC = 168◦ , and the two
largest unused angles are 84◦ and 60◦ , so this case is impossible. One can successfully fill the diagram with
m∠A = 84◦ , as shown below. That is the maximum for this case.

6
4BCD has angles with measures (48◦ , 36◦ , 96◦ ).
4ADE has angles with measures (84◦ , 24◦ , 72◦ ).
4CDE has angles with measures (12◦ , 60◦ , 108◦ ).
◦ ◦ ◦ 540 ◦
If d = 60 660
7 , then the largest angles, 7 and 600
7 , must be located at E. The next-largest angle, 7 , is less
than 84 . Hence the largest possible angle is 84◦ .
◦

Problem 8. Let a1 = a2 = a3 = 1. For n > 3, let an be the number of real numbers x such that

x4 − 2an−1 x2 + an−2 an−3 = 0.

Compute the sum a1 + a2 + a3 + · · · + a1000 .

Solution 8. Consider the quartic equation x4 − 2px2 + q = 0, where p and q are nonnegative. This equation can
be rewritten as (x2 − p)2 = p2 − q. Split into cases to determine the number of distinct real roots:
• If p2 − q < 0, there are 0 real roots.
√
• If p2 − q = 0 and p > 0, there are 2 real roots, at x = ± p.
• If p2 − q = 0 and p = 0, then there is 1 real root, at x = 0. (Note that this is just the case x4 = 0.)
q p
• If p2 − q > 0 and q > 0, then there are 4 real roots, at x = ± p ± p2 − q.
√
• If p2 − q > 0 and q = 0, then there are 3 real roots, at x = 0 and x = ± 2p.
To determine an , iteratively apply the above rules, with p = an−1 and q = an−2 an−3 . Because each term an
depends only on the three previous values an−1 , an−2 , and an−3 , it suffices to find a group of three consecutive

7
terms that occurs twice.
n p = an−1 q = an−2 an−3 p2 − q an
4 1 1 0 2
5 2 1 3 4
6 4 2 12 4
7 4 8 8 4
8 4 16 0 2
9 2 16 −12 0
10 0 8 −8 0
11 0 0 0 1
12 1 0 1 3
13 3 0 9 3
14 3 3 6 4
15 4 9 7 4
16 4 12 4 4
The 3-term subsequence 4, 4, 4 occurs starting at a5 and again at a14 . Thus the sequence has the following
9-term period: 4, 4, 4, 2, 0, 0, 1, 3, 3. The sum of these terms is 21, so the sum a1 + a2 + · · · + a1000 is
1000 4
! 994
! 1000
!
X X X X
an = an + an + an
n=1 n=1 n=5 n=995
= (1 + 1 + 1 + 2) + 110 · 21 + (4 + 4 + 4 + 2 + 0 + 0)
= 2329.

Solution 9. Let N = bkc and let r = {k}. The region Rk is the triangle x + y < N , augmented by vertical “slices”
of the region N ≤ x + y < N + 1. There are N slices that are parallelograms
√ of height 1 and width r, as well
r2
as one slice that is a trapezoid of area r − 2 . The region for k = 10 is shown below.

N 2 −r 2
In all, the area is 2 + (N + 1)r. The goal is to compute k so that

N 2 − r2
+ (N + 1)r = 100.
2

8
2
Replace r with zero and solve for N to obtain an estimated value for N . This substitution yields N2 = 100,
√
and so N is approximately 200 which is slightly larger than 14. Trying N = 14, the equation becomes
142 − r2
+ 15r = 100.
2
√
Simplifying this equation yields r2 − 30r + 4 = 0, √
and so r = 15 − 221. Because 142 < 221 < 152 , this value
of r satisfies 0 < r < 1. Hence, k = N + r = 29 − 221.
The region corresponding to this value of k is shown below.

Solution 10. Let N lie on BC such that BEAN is a rectangle with diagonal BA. Then EN = AB = 48. Trape-
zoid EN CD is isosceles, and hence a cyclic quadrilateral, so N lies on the circumcircle of 4CDE. Let G be
the intersection of BA and EN . Then BG = GA = EG = GN = 24.

Recall that the power of a point P with respect to a circle O is the product of the distance P X · P Y , where P ,
X, and Y are collinear, and X and Y are on O; The power of a point P is invariant for any such chord XY .
Because BA is a chord of the circumcircle of 4ABC and EN is a chord of the circumcircle of 4CDE, the
power of point G with respect to both circumcircles is 24 · 24.

Now recall that given two intersecting circles, the locus of points with the same power in each circle is the line
connecting the two points of intersection of the circles, so G lies on CF .

Applying the power of a point theorem again to chord CF gives GC · CF = 24 · 24 = 496, and GC + CF = 50,
so CG and GF are 18 and 32, in some order.

Apply Stewart’s Theorem to triangle ABC to get

AC 2 · BG + BC 2 · GA = CG2 · AB + AG · GB · AB,
or
AC 2 · 24 + 352 · 24 = CG2 · 48 + 24 · 24 · 48
which gives p
2CG2 − 73.
AC =
√ √
The two values for CG then yield AC = 5 79 or AC = 5 23.

9
4 Power Question 2015: The Power of Riffles
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part.
To receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “com-
pute,” you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show
your work or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a
problem says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered
items may be used in solutions to later numbered items, but not vice versa. Pages submitted for credit should be
NUMBERED IN CONSECUTIVE ORDER AT THE TOP OF EACH PAGE in what your team considers to be
proper sequential order. PLEASE WRITE ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM
NUMBER (not the team name) on the cover sheet used as the first page of the papers submitted. Do not identify
the team in any other way.

This power question concerns sequences of positive integers, which will be written as ha1 , a2 , . . .i, or simply han i.
In all cases, the first term of the sequence will have index 1, that is, no term will be denoted a0 .

Throughout this power event, the word “sequence” is equivalent to “positive integer sequence”.
Sequences may not contain non-positive or non-integer values!

A sequence such as h71, 54, 37, 20, 3, 20, 3, 20, 3, 20, 3, 20, 3, . . .i is called periodic; in this case, the period of the se-
quence is 2, and the periodicity begins at the fourth term. Formally, a sequence will be called periodic if there exists
a positive integer p and a positive integer s such that an = an+p for all n ≥ s; the period of the sequence is the least
such value of p, and the beginning of periodicity of the sequence is the least such value of s.

The following are four examples of sequences:

1. A constant sequence such as h2, 2, 2, . . .i is the sequence all of whose terms are 2. The sequence might also be
written as an ≡ 2.
2. The powers of two are given by an = 2n−1 , and begin h1, 2, 4, . . .i.
3. The Fibonacci sequence will be denoted by hFn i; they are defined by F1 = F2 = 1 and the rule Fn = Fn−1 +Fn−2
for n ≥ 3.
4. The modulo sequence for n, h1, 2, 3, . . . , n, 1, 2, 3, . . . , n, 1, 2, . . .i, that is, the periodic sequence satisfying the
conditions (i) that the sequence has period n, (ii) that the periodicity begins with the first term, and (iii) that
the first n terms are 1, 2, 3, . . . , n.
As the examples suggest, a sequence can be defined by an explicit or recursive formula, and need not have any easily
expressible algebraic formula.

For a given sequence han i, the riffle of han i, denoted ha0n i, is given by setting a01 = a1 , and for n ≥ 2:
0
0 an−1 − an if an < a0n−1
an =
a0n−1 + an otherwise.

Note that each term in ha0n i is a positive integer, and so ha0n i is indeed a sequence.

1. Compute a010 for each sequence below. [5 pts]

a. an = n
b. an = n + 1
c. an = n2
d. an = 2n−1
e. an = Fn , the Fibonacci sequence

10
2. Compute a02015 for each sequence below. [3 pts]
a. an = n
b. an = 2n−1
c. the modulo sequence for n = 5

3. Determine the smallest n for which a0n = 2015 for each sequence below, or show that no such n exists. [7 pts]
a. an = n
b. an = Fn
c. the modulo sequence for n = 1000
d. an = 3n − 2n

4. a. Compute examples of two sequences han i and hbn i with equal periods such that ha0n i and hb0n i have different
periods, or show that no such pair of sequences exists. [2 pts]
b. Show that han i is periodic if and only if ha0n i is periodic. [3 pts]

5. Suppose that han i has period p and that ha0n i has period q. [6 pts]
a. Show that p ≤ q ≤ p · M , where M is the maximum value of ha0n i.
b. Determine whether p must be a divisor of q.

6. A sequence han i is invertible if there exists at least one sequence hbn i for which hb0n i = han i; in that case, the
sequence hbn i is an inverse of han i. Determine whether the following sequences han i are invertible. [4 pts]
a. an =n
= n+1

b. an 2
c. an = 2n
d. an = nm , where m > 1 is a fixed integer

7. a. Compute an example of a sequence with period 17 that is invertible, and an example of a sequence with
period 17 that is not invertible. [2 pts]
b. Suppose that han i is periodic with period 2 beginning at n = 1, and that a2 > 2a1 . Show that han i is
invertible. [2 pts]
c. Let han i be a strictly increasing sequence. That is, an < an+1 for all n. Suppose further that han i is
invertible. Prove that an ≥ 2n−1 for all n. [2 pts]
d. Determine the set of real numbers S for which the following statement is true: han i is invertible if and
only if aan+1
n
∈
/ S for all n ≥ 1. [3 pts]

8. Suppose han i is invertible. Let han i−1 denote the inverse of han i. More generally, if k ≥ 1 and han i−k is
invertible, denote its inverse by han i−(k+1) . (It may be helpful to define han i0 = han i.) Define han i to be
k-invertible if the sequences
han i, han i−1 , . . . , han i−(k−1)
are all invertible.
a. Determine whether there exists a set S for which the following statement is true: han i is 2-invertible if
and only if aan+1
n
∈/ S for all n ≥ 1. [4 pts]
b. Determine whether there exists a sequence han i that is 2015-invertible but not 2016-invertible. [3 pts]
c. Determine whether there exists a sequence han i that is k-invertible for all k ≥ 1. (Such a sequence will
be called infinitely invertible, or ∞-invertible.) [4 pts]

11
5 Solutions to Power Question
1. The first 10 terms of a0n for each sequence are given below.
a. an = n: ha0n i = h1, 3, 6, 2, 7, 1, 8, 16, 7, 17, . . .i
b. an = n + 1: ha0n i = h2, 5, 1, 6, 12, 5, 13, 4, 14, 3, . . .i
c. an = n2 : ha0n i = h1, 5, 14, 30, 5, 41, 90, 26, 107, 7, . . .i
d. an = 2n−1 : ha0n i = h1, 3, 7, 15, 31, 63, 127, 255, 511, 1023, . . .i
e. an = Fn : ha0n i = h1, 2, 4, 1, 6, 14, 1, 22, 56, 1, . . .i

2. a. Although no proof is required for this problem, it is useful to determine which values of n satisfy a0n = 1.
If a0k = 1, the next three terms of the sequence will be k + 2, 2k + 4, k + 1—note the two consecutive
increases—after which ha0n i increases and decreases alternately. Thereafter, if a0m and a0m+2 follow con-
secutive decreases, then a0m+2 = a0m − 1, and similarly a0m+2 = a0m + 1 if a0m and a0m+2 follow consecutive
increases. Thus, if a0k+3 = k + 1, the next value of m such that a0m = 1 is m = k + 3 + 2k = 3k + 3.

Using this relation yields a0n = 1 for n = 1, 6, 21, 66, 201, 606, 1821. Subsequently, a01824 = 1822, and
a01825 = 3647. As 2015 is 190 terms later, a02015 = 3647 + 12 · 190 = 3742.

b. Use induction to show that a0n = 2n − 1. The base case is a01 = a1 = 1 = 21 − 1. Assume then that
a0n = 2n −1. Then because an+1 = 2(n+1)−1 = 2n > a0n , it follows that a0n+1 = a0n +an = 2·2n −1 = 2n+1 −1.
Hence a02015 = 22014 .

c. Note that han i is periodic, and therefore bounded. A reasonable conjecture is that ha0n i is periodic.
Although no proof is required for this problem, it is useful to provide one now. First, though, it must
be shown that if all the terms of a sequence han i are at most M , then all the terms of ha0n i are at most 2M :

Proof: Assume towards a contradiction that there exists at least one term in ha0n i that exceeds 2M .
Let k be the smallest index satisfying a0k > 2M . Then a0k−1 ≤ 2M , and so ha0n i has an increase from a0k−1
to a0k . As this is an increase, it must be the case that ak ≥ a0k−1 . Then 2ak ≥ a0k−1 + ak = a0k > 2M ,
which implies ak > M , a contradiction. 2

Note that in general, the 2M bound cannot be improved. For example, the constant sequence h1, 1, 1, . . .i
has a riffle h1, 2, 1, 2, . . .i. Now it can be shown that if han i is periodic, then ha0n i is periodic as well.

Proof: If han i is periodic, then it is bounded. Let its period be p and its maximum value be M .
For any fixed offset r, the subsequence ha0p+r , a02p+r , a03p+r , . . .i must repeat a value in its first 2M + 1
terms. Suppose the first instance of a repeated value in this subsequence is a0kp+r = a0lp+r . Because
akp+r+c = alp+r+c for all positive integers c, it follows by induction that a0kp+r+c = a0lp+r+c . Therefore
ha0n i is periodic, and its period is at most (2M + 1)p. 2

Because ha0n i is periodic with period 5, it is enough to find k, l, and r such that a05k+r = a05l+r . The
first 20 terms of ha0n i are given below.

1 3 6 2 7 6 4 1 5 10
9 7 4 8 3 2 4 1 5 10

The first instance of such a repeated pair of terms is a07 = a017 = 4. Thus ha0n i is periodic with period 10.
From here it is straightforward to get a02015 = a015 = 3.

12
3. a. From the solution to 2a, recall that if k > 1 and if a0k = 1, then the next term to equal 1 will be a03k+3 . Fur-
thermore, it was shown that the terms a0k+1 , a0k+2 , . . . , a03k+3 consist of alternating increases and decreases,
with the terms coming after the increases or decreases themselves increasing or respectively decreasing by
1. That is, a0k+2 , a0k+4 , . . . , a03k+2 = 2k +4, 2k +5, . . . , 3k +4; and a0k+1 , a0k+3 , . . . , a03k+3 = k +2, k +1, . . . , 1.

For each set of terms ak , ak+1 , . . . , a3k+3 , consider the range of values covered by the respective increasing
and decreasing subsequences:
k Up Down
1 6, 7 3, 2, 1
6 16, 17, . . . , 22 8, 7, . . . , 1
21 46, 47, . . . , 67 23, 22, . . . , 1
66 136, 137, . . . , 202 68, 67, . . . , 1
201 406, 407, . . . , 607 203, 202, . . . , 1
606 1216, 1217, . . . , 1822 608, 607, . . . , 1
1821 3646, 3647, . . . , 5467 1823, 1822, . . . , 1
5466 10936, 10937, . . . , 16402 5468, 5467, . . . , 1
It is clear then that the first time 2015 occurs in ha0n i will be in the decreasing subsequence beginning at
a05467 = 5468. Because 5468 − 2015 = 3453, it will take 2 · 3453 = 6906 more terms to reach 2015. That
is, 2015 = a05467+6906 = a012373 .

b. In this case, a0n = 2015 has no solution. Note that ha0n i begins with 1, 2, 4, 1. That is, a04 = 1.

Now suppose a0n = 1, for some index n > 2. Then an+1 = Fn+1 > 1 and an+2 = Fn+2 ≥ Fn+1 + 1.
Therefore a0n+1 = Fn+1 + 1, a0n+2 = Fn+1 + 1 + Fn+2 = Fn+3 + 1, and a0n+3 = (Fn+3 + 1) − Fn+3 = 1.
That is, every third term of ha0n i will equal 1, and the intermediate terms will each be 1 greater than a
Fibonacci number.

This result applies to ha0n i from n = 4 onward. Because 2014 is not a Fibonacci number, 2015 does
not appear in ha0n i.

c. It can be shown by contradiction that 2015 does not appear in ha0n i. If a0n = 2015, then because an ≤ 1000,
a0n−1 must be smaller than a0n . That is, a0n−1 = a0n − an = 2015 − an ≥ 1015. But because a0n−1 ≥ 1015
and an ≤ 1000, it follows that a0n = a0n−1 − an < 1015. Thus 2015 does not appear in ha0n i.

d. The number 2015 does not appear in ha0n i. First, it will be shown that ha0n i has no decreases, or equiv-
alently, an > a0n−1 for all n. Proceed by induction. In the base case, a2 = 5 > 1 = a01 . Then assume
a0k = a1 +· · ·+ak for some k ≥ 1. Note that 31 +32 +· · ·+3k = 21 (3k+1 −3) and 21 +22 +· · ·+2k = 2k+1 −2,
k+1 k+1
so a0k = 3 2 − 2k+1 + 21 . Then ak+1 − a0k = 3 2 −1 > 0, so a0k+1 = a0k + ak+1 , which is another increase.
This completes the inductive step.

Because ha0n i is strictly increasing, it is necessary only to compute terms until reaching one that is at
least 2015. The first several terms of ha0n i are 1, 6, 25, 90, 301, 966, 3025. Hence 2015 does not occur in the
sequence.

4. a. There are many such examples. Consider han i = h1, 2, 1, 2, . . .i and hbn i = h1, 3, 1, 3, . . .i: their respective
riffles, with periods 4 and 6, are shown below.
ha0n i = h1, 3, 2, 4, 3, 1, 2, 4, 3, 1, . . .i
hb0n i = h1, 4, 3, 6, 5, 2, 1, 4, 3, 6, . . .i

b. The forward direction of the proof was shown earlier as a part of the solution to 2c, so now consider the
reverse direction of the proof.

13
Suppose that ha0n i is periodic with period p. Let r be an integer such that a0s = a0s+p for all s ≥ r.
Because an = |a0n − a0n−1 | for all n > 1, it follows that as+p = |a0s+p − a0s+p−1 | = |a0s+2p − a0s+2p−1 | = as+2p
for all s ≥ r. This establishes the periodicity of han i, and completes the proof. 2

5. a. By assumption, the sequence han i has period p, the riffle ha0n i has period q, and maxn a0n = M . Let the
periodicity of ha0n i begin at r, that is, for all k ≥ r, a0k = a0k+q . Then

ak+1 = a0k+1 − a0k = a0k+1+q − a0k+q = ak+1+q

for all such k, and so han i is periodic with period at most q. That is, p ≤ q.

Let s be an integer for which ak = ak+p for all k ≥ s. Consider the M + 1 values a0s , a0s+p , . . . , a0s+M p .
These are M + 1 terms, all of which are values between 1 and M . Hence there must exist integers u and
v such that 0 ≤ u < v ≤ M and a0s+up = a0s+vp . It can be shown via induction that ha0n i is periodic from
a0s+up onwards. Assume a0s+up+z = a0s+vp+z for some z ≥ 0. (The case z = 0 has already been estab-
lished.) Because a0s+up+z = a0s+vp+z and as+up+z+1 = as+vp+z+1 , it follows that a0s+up+z+1 = a0s+vp+z+1 ,
concluding the inductive step. Thus ha0n i is periodic with period at most (v − u)p ≤ M p.

b. Show that q is a multiple of p by contradiction. If p = 1, then q must be a multiple of p. So assume that

p > 1. In the proof that follows, the essential idea is that if q is not a multiple of p, then han i must be
periodic with a period of c = gcd(p, q).

Assume towards a contradiction that q is not a multiple of p. Let c = gcd(p, q), and let d = pc . Note that
c < p, and so d > 1.

Let r be an integer for which ar+k = ar+k+p for all k ≥ 0. Then for at least one integer i with 1 ≤ i ≤ c,
the terms ar+i , ar+i+c , . . . , ar+i+(d−1)c must contain at least two different values. (If this weren’t the case,
then han i would have period at most c, which is a contradiction.)

6 ar+i+yc . Because c = gcd(p, q), pc

So let x, y be integers, with 0 ≤ x < y < d for which ar+i+xc =
q
and c are relatively prime. Hence there exists an integer m > 0 such that m · qc ≡ 1 mod pc . Thus
m(y − x)q ≡ (y − x)c mod p. Then
ar+i+xc = |a0r+i+xc − a0r+i+xc−1 |
= |a0r+i+xc+m(y−x)q − a0r+i+xc+m(y−x)q−1 |
= ar+i+xc+m(y−x)q .
But r + i + xc + m(y − x)q is no less than r and is congruent to r + i + yc mod p, so
ar+i+xc+m(y−x)q = ar+i+yc 6= ar+i+xc .
p
This result contradicts the existence of two distinct values within one of the c subsequences of han i, which
would imply that han i is periodic with period at most c.

6. a. The sequence is not invertible. Note that if hbn i is an inverse of han i, then b3 = |a3 − a2 | = 1. But then
it follows that a3 = b03 = b02 − b3 = 2 − 1 = 1 6= a3 .

b. The sequence is not invertible. The first few terms of han i are 1, 3, 6, 10, 15. So if hbn i is an inverse of
han i, it follows that b4 = |a4 − a3 | = 4. But then a4 = b04 = b03 − b4 = 6 − 4 = 2.

c. The sequence is invertible, with inverse hbn i, where b1 = 1, and bn = 2n−2 for n > 1. It is straightforward
to show by induction that hb0n i = han i.

14
d. This sequence is not invertible for any value of m. As with the earlier examples of sequences√ that were not
invertible, the key is to find a pair of terms an , an+1 for which aan+1 n
< 2. Because m
2 > 1, there exists
an integer k for which ( k ) < 2. So suppose there exists a sequence hbn i for which b1 , b02 , . . . , b0k =
k+1 m 0

a1 , a2 , . . . , ak . Because ( k+1 m
k ) < 2, it follows that ak+1 < 2ak . Therefore bk+1 = |ak+1 − ak | = ak+1 − ak ,
which is less than ak = bk . Thus b0k+1 = b0k − bk+1 = 2ak − ak+1 , which is less than ak+1 .
0

7. a. The periodic sequence h17, 16, 15, . . . , 1, 17, 16, . . .i is invertible, whereas the periodic sequence
h1, 2, 3, . . . , 17, 1, 2, . . .i is not invertible.

b. Let hbn i be defined as follows: b1 = a1 , and for n > 1, bn = a2 − a1 . Then b2 > b1 , and it follows that
b01 = b1 = a1 , b02 = b1 + b2 = a2 , and b03 = b02 − b3 = a1 . Because b01 = b03 and hbn i is constant for n ≥ 2,
hb0n i is periodic from n ≥ 2 onwards. Thus hbn i is an inverse of han i.

c. Proceed by induction on n. For the base case, a1 ≥ 1 = 21−1 . Suppose that hb0n i = han i, and assume
that an ≥ 2n−1 . Then an+1 ≥ an implies that bn+1 ≥ an = 2n−1 , so an+1 = an +bn+1 ≥ 2n−1 +2n−1 = 2n .

d. Given a sequence han i of positive integers, define the ruffle sequence hân i as follows: let â1 = a1 , and for
n > 1 let ân = |an − an−1 |.

Claim: If han i is invertible, it can have only one inverse, namely hân i.

Proof of Claim: Suppose there exists a sequence hbn i for which hb0n i = han i. Then b1 = a1 , and
for n > 1, bn = |an − an−1 |. That is, hbn i = hân i. 2

Note the restricted nature of the claim: it does not guarantee that h(â)0n i = han i. For example, the
ruffle of h1, 2, 3, 4, . . .i is the constant sequence h1, 1, 1, 1, . . .i, but the riffle of that constant sequence
is the period-2 sequence h1, 2, 1, 2, . . .i. The reason the claim appears to generate an extraneous inverse
to h1, 2, 1, 2, . . .i is that this sequence doesn’t have an inverse, which was one of the hypotheses of the claim!

Now consider the question at hand:

an+1
Stronger Claim: A sequence han i is invertible if and only if an ∈
/ [1, 2) for all n ≥ 1.

Proof of Stronger Claim: Suppose han i is a positive integer sequence such that aan+1 n
∈/ [1, 2) for
all n ≥ 1. Then ân > 0 for all n ≥ 1. It is true that â1 = (â)01 = a1 . Proceed by induction to show that
the ruffle sequence hân i is the inverse of han i:

Suppose that (â)0k = ak for some k ≥ 1. Then âk+1 = |ak+1 − ak |. If ak+1 < ak , then âk+1 =
ak − ak+1 < ak = (â)0k , and so (â)0k+1 = (â)0k − âk+1 = ak − (ak − ak+1 ) = ak+1 . If ak+1 > ak ,
then âk+1 = ak+1 − ak ≥ ak = (â)0k , and so (â)0k+1 = (â)0k + âk+1 = ak + (ak+1 − ak ) = ak+1 . This
completes the inductive step. (And the earlier claim entails that there are no other inverses of han i.)

To show that the condition on S is necessary, use proof by contradiction. Suppose k is the smallest
integer for which aak+1
k
∈ [1, 2). By the earlier induction, (â)0k = ak . Then âk+1 = ak+1 − ak < ak = (â)0k ,
and so (â)0k+1 = (â)0k − âk+1 = ak − (ak+1 − ak ) = 2ak − ak+1 , which is smaller than ak and therefore
cannot equal ak+1 . 2

8. a. No such set S exists. For a given sequence han ik , let Rak be the set of values { aa12 , aa32 , aa43 , . . .}. Note that
earlier it was shown that han i0 is invertible if and only if Ra0 ∩ [1, 2) = ∅.

Let hbn i0 be the periodic sequence h1, 3, 1, 3, 1, 3, . . .i. Then Rb0 = { 31 , 3}. This sequence is invertible,

15
and hbn i−1 = hb̂n i = h1, 2, 2, 2, 2, 2, . . .i. Clearly, hbn i−1 is not invertible, because it has repeated values.

Now let hcn i0 be the sequence whose first few terms are 1, 3, 9, 30, 10, and whose subsequent terms
are cn = 10n−4 for n 6= 6. That is, the first eight terms are 1, 3, 9, 30, 10, 100, 1000, 10000. Then
Rc0 = {3, 10 1 0 0
3 , 3 , 10}, and it follows that Rc ∩ [1, 2) = ∅. Thus hcn i is invertible. The first several
terms of hcn i are 2, 6, 21, 20, 90, 900, 9000. Rc = {3, 3 , 21 , 2 , 10}. It follows that Rc−1 ∩ [1, 2) = ∅, so
−1 −1 7 20 9

hcn i0 is indeed 2-invertible.

On the other hand, Rb0 ⊂ Rc0 . In order for the original claim to be true, either 31 or 3 must be an
element of S. But both these values are in Rc0 , and hcn i0 is 2-invertible. Therefore the claim cannot be
true, as no such set S satisfies the condition.

b. Such a sequence exists. For an integer k ≥ 1 and sequence han i, let han ik denote the kth riffle of han i.
That is, han ik+1 is the riffle of han ik . (And as before, let han i0 = han i.)
an+1
Consider the kth riffle of the sequence an = 2n−1 . Note that han i is invertible, because an = 2 for
all n. However, because the first two terms of han i−1 are 1, 1, han i is not 2-invertible.
bn+1
Claim: Let hbn i be the kth riffle of han i, where k ≥ 1. Then hbn i is strictly increasing, with bn > 2 for
all n.

Proof of Claim: The claim is true for k = 1, as han i1 = h1, 3, 7, 15, 31, . . .i, i.e., the sequence 2n − 1.
So assume the claim is true for some positive integer k, and let hcn i = hb0n i = han ik+1 . First show
that hcn i is increasing. Proceed by induction. For the base case, c1 = b1 , and because b2 > b1 , it
follows that c2 = b1 + b2 . Assume (further) now that hcn i is increasing for c1 , c2 , . . . , cj . Note that
b bj+1 bj+1 1 bj+1
bj < j+1 2 , bj−1 < 4 , . . . , b1 < 2j , and so cj = b1 + b2 + · · · + bj < bj+1 (2 − 2j ). Then because bj > 2,
hcn i increases at cj+1 . This completes the (inner) inductive step, demonstrating that hcn i is increasing. 2

It remains to complete the outer induction; that is, that the ratios of consecutive terms are decreas-
ing but are always larger than 2. It can be shown that they are all larger than 2, by noting that
cn+1 b1 + b2 + · · · + bn + bn+1
=
cn b1 + b2 + · · · + bn
bn+1
=1+
b1 + b2 + · · · + bn
bn+1
> 1 + bn+1 bn+1 bn+1
2n + 2n−1 + · · · + 21
1
>1+
1 − 21n
> 2.

This argument completes the proof.

c2 c3 c4
(Note: It can furthermore be shown that c1 > c2 > c3 > · · · > 2.)

It has thus been shown that for any positive integer k, han ik is invertible. (Its ratios are all larger
than 2.) Thus han i2014 can be inverted 2014 times to get han i, and one more time to get 1, 1, 2, 4, 8, . . .,
but cannot be inverted once more (which would be a 2016th inversion).

This solution can easily be amended to find sequences that can be inverted k times but not k + 1 times.

16
c. No such sequence han i exists. Proceed by contradiction: if a sequence han i is invertible, its unique inverse
is the ruffle sequence, hân i. So assume that han i is infinitely invertible. Let a−k
n denote the nth term of
han i−k , which is well-defined. Then a−k
1 = a1 for all k ≥ 1, and a−k
n = |a −k+1
n −a−k+1
n−1 | for all n > 1, k ≥ 1.

Consider the sequence a2 , a−1 −2

2 , a2 , . . . . The differences between consecutive terms are all just a1 . There
must exist an integer s2 such that a−k 2 < a1 for all k ≥ s2 . For example, if a1 , a2 = 4, 17, then
a2 , a−1
2 , a−2
2 , . . . = 17, 13, 9, 5, 1, 3, 1, 3, 1, 3, . . . . That is, after some point, all the values are less than a1 .

Now consider the sequence a3 , a−1 −2

3 , a3 , . . . . The differences between consecutive terms are the terms
of the previous sequence. Note that at some point, all the differences are at most a1 − 1, after which,
there must be some s3 ≥ 1 for which a−s
3
3
< a1 − 1. Then all subsequent terms are at most a1 − 2.

This process can be continued to show that for any r > 1, there exists some integer sr > sr−1 for
which ar , a−1 −2
r , ar , . . . eventually achieves a value less than a1 − (r − 2). Then all subsequent terms are at
most a1 − (r − 2). But this cannot continue indefinitely, as eventually, some term must equal zero, which
is not allowed. Hence there does not exist an infinitely invertible sequence.

Authors’ Note: The authors initially stumbled upon this topic in writing individual questions; in fact, problem 2a
was “almost” an individual question. Many other questions are suggested, such as the following:
1. Does there exist an integer N > 1 such that the first N terms of both han i and ha0n i are permutations of the
set of integers {1, 2, . . . , N }? If so, what are all such integers?
2. For a specific sequence han i, is it possible to determine which positive integers never occur in its riffle ha0n i, or
which occur only finitely many times in its riffle, or which occur infinitely many times in its riffle? For example,
what about the sequence an = n?
Either of these questions, or extensions of the previous questions, would make a promising topic for a mathematics
research project!

17
6 Individual Problems

Problem 1. The perimeter of regular heptagon HEP T GON is 2015 more than the perimeter of square ARM L.
Let x = HE − AR. Compute the greatest possible integer value of x.

27
Problem 2. Let n be the greatest integer such that nn is a divisor of 2727 . Compute the number of positive
divisors of n.

Problem 3. A rectangular box has integer edge lengths. The sum of the numerical values of its volume, its surface
area, and its twelve edge lengths is 2015. Compute the length of the box’s interior diagonal.

Problem 4. Compute the smallest positive integer n such that

n
X 1 2014
log2 1 + 2k ≥ 1 + log2 .
2 2015
k=0

Problem 5. Let f be a function such that for all x, f (x) = f (x + 1) + f (x − 1). Given that f (20) = 15 and
20 = f (15), compute f (20152015).

Problem 6. In trapezoid ABCD with bases AB and CD, AB = 14 and CD = 6. Points E and F lie on AB such
−→
that AD k CE and BC k DF . Segments DF and CE intersect at G, and AG intersects BC at H. Compute
[CGH]
.
[ABCD]

Problem 7. Let f be the function defined by f (x) = x3 − 49x2 + 623x − 2015, and let g(x) = f (x + 5). Compute
the sum of the roots of g.

18
Problem 8. In rectangle W ASH, point E lies on SH such that ∠AW S ∼= ∠HW E. Point D lies on W S such
that ED ⊥ W S. Given that [W ASH] = 100 and [SED] = 32, compute sin ∠SW E.

Problem 9. An (a, r, m, `)-trapezoid is a trapezoid with bases of length a and r, and other sides of length m and
`. Compute the number of positive integer values of ` such that there exists a (20, 5, 15, `)-trapezoid.

Problem 10. Six people of different heights are getting in line to buy donuts. Compute the number of ways they
can arrange themselves in line such that no three consecutive people are in increasing order of height, from
front to back.

19
7 Answers to Individual Problems

Answer 1. 287

Answer 2. 79

√
Answer 3. 5 19

Answer 4. 3

Answer 5. −5

27
Answer 6.
160

Answer 7. 34

√
4+6 6
Answer 8.
25

Answer 9. 29

Answer 10. 349

20
8 Solutions to Individual Problems

Problem 1. The perimeter of regular heptagon HEP T GON is 2015 more than the perimeter of square ARM L.
Let x = HE − AR. Compute the greatest possible integer value of x.

Solution 1. Let a = HE and b = AR so that x = a − b and 7a − 4b = 2015. Substituting a = x + b,

7(x + b) − 4b = 2015
7x + 3b = 2015
2015 − 3b 6 3b
x= = 287 + − < 288.
7 7 7
The greatest possible integer x is thus 287, which occurs when a = 289 and b = 2. Coincidentally, both a and
b are also integers.

27
Problem 2. Let n be the greatest integer such that nn is a divisor of 2727 . Compute the number of positive
divisors of n.

a
Solution 2. Because n must be a power of 3, set n = 3a for some positive integer a. Then 3a·3 must be a divisor
27 3 27 82
of 2727 = 33·(3 ) = 33 . That is, a · 3a ≤ 382 . Trying decreasing integers starting with a = 82, one quickly
arrives at a = 78 as the largest integral value that satisfies the inequality. Thus n = 378 is the largest integer
27
for which nn is a divisor of 2727 , and 378 has 79 positive divisors.

Solution 3. Let a, b, and c denote the length, width, and height of the box. With 4(a+b+c)+2(ab+ac+bc)+abc =
2015, it follows that

2015 − 4b − 4c − 2bc 2015 − 2b(c + 2) − 4c 2023 − 2b(c + 2) − 4(c + 2) 2023

a= = = = − 2,
bc + 2b + 2c + 4 (b + 2)(c + 2) (b + 2)(c + 2) (b + 2)(c + 2)

hence b + 2 and c + 2 must both divide 2023 = 7 · 172 . If b = c = 15, then a = 7 − 2 = 5. If one of b
or √ is 15, then a = 17 − 2 = 15, thus in any case, the length of the interior diagonal is
√ c is 5 and the other
52 + 152 + 152 = 5 19.

Problem 4. Compute the smallest positive integer n such that

n
X 1 2014
log2 1 + 2k ≥ 1 + log2 .
2 2015
k=0

k
Solution 4. It is simpler to replace 1 + 21k with 1 + x2 , and later evaluate at x = 21 . Then the left-hand side of
2
the inequality is
n n
X k Y k
log2 1 + x2 = log2 1 + x2 .
k=0 k=0
Qn k
Consider the product k=0 (1 + x2 ). For n = 0, this product is 1 + x. For n = 1, this product is 1 + x + x2 + x3 .
For n = 2, this product is 1 + x + x2 + x3 + x4 + x5 + x6 + x7 . It is natural to conjecture that the product

21
P2n+1 −1 i
simplifies to i=0 x . This can be proven by induction on n as follows. The base case has already been
Qn k P2n+1 −1 i
checked. If k=0 (1 + x2 ) = i=0 x , then by induction,
n+1
" n #
2k 2k n+1
Y Y
(1 + x ) = (1 + x ) (1 + x2 )
k=0 k=0
 
2n+1
X−1 n+1
= xi  (1 + x2 )
i=0

2n+1
X−1 2n+2
X−1
i
= x + xi
i=0 i=2n+1
2n+2
X−1
= xi .
i=0

n+1
k
1−x2
Qn
Thus the product evaluates to a finite geometric series, and k=0 (1 + x2 ) = 1−x . Set x = 12 , and take
the base-2 logarithm of both sides:
n
X 1 1 1
log2 1 + 2k = log2 1 − − log2 1 −
k=0
2 22n+1 2

1
= log2 1 − + 1.
22n+1

Rephrased in this way, the problem asks for the smallest positive integer n such that

1 1
log2 1 − 2n+1 + 1 ≥ 1 + log2 1 − .
2 2015

Because log2 is an increasing function, it preserves inequalities. It therefore suffices to compute the smallest
positive integer n such that
1 1
1 − 2n+1 ≥ 1 − .
2 2015
n+1
Rearrange this inequality to obtain 22 ≥ 2015. If n = 2, then the left-hand side is 28 = 256 < 2015, but if
16
n = 3, the left-hand side is 2 = 65536 > 2015, so the smallest positive integer n satisfying the criterion is 3.

Problem 5. Let f be a function such that for all x, f (x) = f (x + 1) + f (x − 1). Given that f (20) = 15 and
20 = f (15), compute f (20152015).

Solution 5. Use the given equation to expand f (x + 1) + f (x + 2).

f (x + 1) + f (x + 2) = f (x) + f (x + 2) + f (x + 1) + f (x + 3).

Cancel f (x + 1) + f (x + 2) from both sides:

0 = f (x) + f (x + 3).

Thus if f (1) = a, f (2) = b, and f (3) = c, the sequence f (1), f (2), f (3), . . . is given by a, b, c, −a, −b, −c, a, b, c, . . . .
This sequence is periodic with period six. Thus b = f (2) = f (20) = 15, and c = f (3) = f (15) = 20. Because
20152015 ≡ 1 mod 6, it follows that f (20152015) = f (1) = a. Lastly, recall that f (2) = f (1) + f (3), so
f (20152015) = a = 15 − 20 = −5.

22
Problem 6. In trapezoid ABCD with bases AB and CD, AB = 14 and CD = 6. Points E and F lie on AB such
−→
that AD k CE and BC k DF . Segments DF and CE intersect at G, and AG intersects BC at H. Compute
[CGH]
.
[ABCD]

Solution 6. Let h be the altitude of trapezoid ABCD, so that [ABCD] = 10h. Because AECD and F BCD are
both parallelograms, AE = F B = 6, so EF = 2. Therefore [CEB] = 4h.

CG DG CD CG
Triangles EGF and CGD are similar, so EG = FG = EF = 3. This means that CE = 34 .

Triangles AGF and AHB are also similar, so HB AB 14 DG GF 1

GF = AF = 8 . Because F G = 3, DF = 4 . Because F BCD is
GF
a parallelogram, DF = CB, so CB = 4 . Therefore CB = GF · CB = 8 · 4 = 16 . Thus CH
1 HB HB GF 14 1 7 9
CB = 16 .

Let α = m∠ECB. Then [CEB] = 12 CE ·CB ·sin α and [CGH] = 12 CG·CH ·sin α, so [CGH] = [CEB]· CE CG CH
· CB .
3 9 27 [CGH] 27
Therefore by the results above, [CGH] = 4h · 4 · 16 = 16 h. Because [ABCD] = 10h, [ABCD] = 160 .

Problem 7. Let f be the function defined by f (x) = x3 − 49x2 + 623x − 2015, and let g(x) = f (x + 5). Compute
the sum of the roots of g.

Solution 7. The sum of the roots of f (x) is 49 by Vieta’s formula. If r is a root of f , then r − 5 must be a root
of g and conversely, because g(r − 5) = f ((r − 5) + 5) = f (r). (If all three roots are real, this result can be
seen graphically, because the graph of g(x) is a translation of the graph of f (x) by 5 units to the left.) Thus
the sum of the roots of g(x) is 49 − 3 · 5 = 34.

Problem 8. In rectangle W ASH, point E lies on SH such that ∠AW S ∼= ∠HW E. Point D lies on W S such
that ED ⊥ W S. Given that [W ASH] = 100 and [SED] = 32, compute sin ∠SW E.

23
Solution 8. Let θ = m∠HW E and a = HW . Then HE = a tan θ, W A = a cot θ, and sin ∠SW E = sin(90◦ −2θ) =
cos 2θ. Because [W SA] = 12 [W ASH],
32 [SED] SE 2 (a cot θ − a tan θ)2 (cot θ − tan θ)2 (cos2 θ − sin2 θ)2 cos2 2θ
= = = = = = .
50 [W SA] W S2 a2 + (a cot θ)2 1 + (cot θ)2 cos2 θ cos2 θ

Cross-multiplying,
√
50 cos2 2θ = 32 cos2 θ = 16 + 16 cos 2θ, resulting in a quadratic in√cos 2θ with solutions
cos 2θ = 25 . Because θ < 45◦ , the negative solution is rejected, so the answer is 4+6
4±6 6 6
25 .

Solution 9. Let ARM L be an (a, r, m, `)-trapezoid, with AR = r, RM = `, M L = a, and LA = m, so that

AR k M L, and a ≥ r. Then place N on M L such that N R k LA, as shown in the diagram below.

Because ARN L is a parallelogram, RN = m and LN = r, so N M = a − r. By the triangle inequality,

m + ` > a − r ⇒ ` > a − r − m, m + a − r > `, and a − r + ` > m ⇒ ` > m + r − a. In the case where a = 20,
r = 5, and m = 15, these three inequalities yield ` > 0 (twice) and 30 > `. Because ` is an integer, there are
29 possibilities for `.

Alternate Solution: Using the diagram above, consider the extreme (degenerate) cases. With a 0◦ an-
gle at L, a = ` + m + r, making ` = 20 − 15 − 5 = 0. With a 180◦ angle at L, a + m = r + `, making
` = 20 + 15 − 5 = 30. Thus 0 < ` < 30 for a total of 29 possible integer values.

24
Problem 10. Six people of different heights are getting in line to buy donuts. Compute the number of ways they
can arrange themselves in line such that no three consecutive people are in increasing order of height, from
front to back.

Solution 10. Let A be the event that the first three people are in increasing height order, let B be the event that
the second, third, and fourth people are in increasing height order, let C be the event that the third, fourth, and
fifth people are in increasing height order, and let D be the event that the last three people are in increasing
height order. Letting |x| denote the number of orderings that result in event x occurring, then |A ∪ B ∪ C ∪ D|
is the number of orderings where there are three consecutive people in increasing height order. The desired
answer results by subtracting this number from 6! = 720, the total number of orderings of six individuals.

Proceed by using the principle of inclusion-exclusion:

|A ∪ B ∪ C ∪ D| = (|A| + |B| + |C| + |D|)

− (|A ∩ B| + |B ∩ C| + |C ∩ D| + |A ∩ C| + |B ∩ D| + |A ∩ D|)
+ (|A ∩ B ∩ C| + |B ∩ C ∩ D| + |A ∩ B ∩ D| + |A ∩ C ∩ D|)
− (|A ∩ B ∩ C ∩ D|) .

Observe that |A| = 63 ·3!, because there are 63 ways to choose which 3 people to put in increasing height order

in the first three spots,

and then there are 3! ways to arrange the remaining 3 people in any order. Similarly,
|B| = |C| = |D| = 63 · 3!.

Next, observe that the event A ∩ B really means that persons 1 through 4 are all in increasing height or-
der, because event A says that persons 1 through 3 are in order, and event B says that persons 2 through
4 are in order,
and both of these hold exactly when persons 1 through 4 are all in increasing height order.
There are 64 ways to select which four people to place in these spots, one way to order them, and 2! ways
to order the remaining two people. So |A ∩ B| = 64 · 2!. Identical arguments hold for B ∩ C and C ∩ D, so

|B ∩ C| = |C ∩ D| = 64 · 2!.

Next, observe that |A ∩ C| and |B ∩ D| count the number of ways to arrange the first 5 people in order
and the number of ways to arrange the last 5 people in order, respectively. Therefore following an argument
similar to above, |A ∩ C| = |B ∩ D| = 65 .

The event A ∩ D is the event that the first three people are arranged in order, and the last three people
are arranged in order. There are 63 ways to select which three people are in the first three spots, and then
only one way to arrange them, and only one way to arrange the remaining three people, so |A ∩ D| = 63 .

The events A ∩ B ∩ C is identical to the event A ∩ C, because both events describe the event that the first 5
people in line are in order. Similarly, the event B ∩ C ∩ D is identical to the event B ∩ D. These were calculated
above, so |A ∩ B ∩ C| = |B ∩ C ∩ D| = 65 .

The events A ∩ B ∩ D, A ∩ C ∩ D, and A ∩ B ∩ C ∩ D are all identical to the event that all 6 people are
in increasing height order, so |A ∩ B ∩ D| = |A ∩ C ∩ D| = |A ∩ B ∩ C ∩ D| = 1.

Combine these computations to obtain:

6 6 6 6 6
|A ∪ B ∪ C ∪ D| = 4 · · 3! − 3 · · 2! − 2 · − +2· + 2 · 1 − 1 = 371.
3 4 5 3 5

Therefore the number of orderings where no three consecutive people are in increasing height order is 720−371 =
349.

25
9 Relay Problems

Relay 1-1 The average of six distinct real numbers is 275. The average of the four least numbers is 200. The
average of the four greatest numbers is 340. Compute the average of the middle two numbers.

T 1 T
Relay 1-2 Let T = T N Y W R. A rectangle has area 4 + 8 and a diagonal of length 2 . Compute the perimeter of
the rectangle.

Relay 1-3 Let T = T N Y W R, and let k be a real number such that the region above the x-axis satisfying the
inequalities |2x| + |y| ≤ 2k and |x| + |y| ≥ k has area T . Compute k.

Relay 2-1 Given that 1 − r + r2 − r3 + · · · = s and 1 + r2 + r4 + r6 + · · · = 4s, compute s.

Relay 2-2 Let T = T N Y W R. Two swimming pools Y and P are surrounded by rectangular walkways of width w
feet. The sides of Y are 4 feet longer than the sides of P , and the area of the walkway surrounding Y is 128T
square feet greater than the area of the walkway surrounding P . Compute w.

Relay 2-3 Let T = T N Y W R, and let p and q be real numbers with p < q. Given that p2 − 5p = q 2 − 5q = T ,
1 1
compute − .
p q

26
10 Relay Answers

Answer 1-1 255

Answer 1-2 256

Answer 1-3 16

4
Answer 2-1
7

32
Answer 2-2
7

√
2121
Answer 2-3 −
32

27
11 Relay Solutions

Solution 1-1 Let the numbers be a < b < c < d < e < f . Then a+b+c+d 4 = 200 ⇒ a + b + c + d = 800,
c+d+e+f a+b+c+d+e+f
4 = 340 ⇒ c + d + e + f = 1360, and 6 = 275 ⇒ a + b + c + d + e + f = 1650. From the first
two equations it follows that a + b + 2c + 2d + e + f = 2160, so c + d = 510, and the average is 510
2 = 255.

T 1 T
Relay 1-2 Let T = T N Y W R. A rectangle has area 4 + 8 and a diagonal of length 2 . Compute the perimeter of
the rectangle.

√
Solution 1-2 Let ` and w be the length and width of the rectangle, respectively. Then it is given that `2 + w2 = T2 ,
2 2
hence `2 + w2 = T4 . The area is `w = T4 + 18 = 2T8+1 , so 2`w = 2T4+1 and (` + w)2 = `2 + 2`w + w2 = T +2T
4
+1
.
T +1
Therefore ` + w = 2 , and the perimeter is 2` + 2w = |T + 1|. With T = 255, the perimeter is 256.

Relay 1-3 Let T = T N Y W R, and let k be a real number such that the region above the x-axis satisfying the
inequalities |2x| + |y| ≤ 2k and |x| + |y| ≥ k has area T . Compute k.

Solution 1-3 The region is the concave kite whose vertices are (±k, 0), (0, k), and (0, 2k), as shown below.

(0,2k)

(0,k)
(-k,0) (k,0)

The area of the kite is twice the area of the portion contained in the first quadrant, which is a triangle with
2
base k and height k. So the area of the triangle is k2 and the area of the entire kite is k 2 . With T = 256,
√
k = 256 = 16.

Relay 2-1 Given that 1 − r + r2 − r3 + · · · = s and 1 + r2 + r4 + r6 + · · · = 4s, compute s.

1 1
Solution 2-1 Use the formula for the sum of an infinite geometric series twice to get 1+r = s and 1−r 2 = 4s. Divide
2
1−r s 1
the first equation by the second to yield 1+r = 4s , or 1 − r = 4. Therefore r = 34 , and s = 1
1+ 34
= 74 .

28
Solution 2-2 Let pool P have length x feet and width y feet. Pool P is surrounded by a walkway of width w feet,
so the total rectangular shape has length x + 2w feet and width y + 2w feet. Thus the area of the walkway itself
is (x + 2w)(y + 2w) − xy = xy + 2xw + 2yw + 4w2 − xy = 2(x + y)w + 4w2 . Substitute x + 4 for x and y + 4
for y in this expression to obtain the area of the walkway surrounding pool Y : 2((x + 4) + (y + 4))w + 4w2 .
The difference between these areas is

2((x + 4) + (y + 4))w + 4w2 − 2(x + y)w − 4w2 = 16w.

Equate this quantity with 128T to give 16w = 128T , or w = 8T . The correct value of T is 74 , so w = 32
7 .

Relay 2-3 Let T = T N Y W R, and let p and q be real numbers with p < q. Given that p2 − 5p = q 2 − 5q = T ,
1 1
compute − .
p q

Solution 2-3 Both p and q satisfy the equation x2 −5x−T = 0, so they must be the roots of the polynomial function
f (x) = x2 −5x−T . Vieta’s 2 2
√ formulas thus yield p+q = 5 and pq = −T . Therefore (q−p) = (pq) −4pq = 25+4T .
Because p < q, q − p = 25 + 4T . Therefore
√
1 1 q−p 25 + 4T
− = = .
p q pq −T
32
The correct value of T is 7 , so

√
q q
32 303
1 1 25 + 4 · 7 7 2121
− = = =− .
q p − 32
7
32
−7 32

29
12 Super Relay
1. Triangle T AR has a right angle at R, with RT = 3 and AT = 5. Triangle LM N is similar to triangle T AR,
and M N = 8. Compute AR · M L.
2. Let T = T N Y W R. Compute the unique fixed point of the function f defined by f (x) = T · (x − T + 1). [Note:
A fixed point of a function g(x) is a value z such that g(z) = z.]

3. Let T = T N Y W R. A dessert tray containing T chocolate-covered strawberries has been ordered for an ARML
coaches meeting. Marlys eats some of the strawberries, leaving 24 strawberries for the other attendees to
consume. Then Marlys will have eaten K% of the strawberries. Compute the value of K.
T
4. Let T = T N Y W R, and let a = log 8 . Given that log K = 3 − 2a, compute the value of K.

5. Let T = T N Y W R. In the diagram to the right, circle K has area T π

and is internally tangent to circle J at point E. Chord XL is tangent to
circle K at point O. Given that the product of the radii of circles J and K
is 4T − 20, compute XL.

6. Let T = T N Y W R. Mary places 20 green marbles into a hat, Lydia places 15 purple marbles into a hat, and
Joshua places T blue marbles into a hat. If a marble is randomly selected from the hat, the probability that it
8
will be purple is K . Compute the value of K.
7. Let
√ T = T N Y W R, and let K be the smallest
√ prime factor of T . Suppose that a large cube has edge length
3K. A smaller cube has edge length K − 1 and is placed on a face F of the larger cube so that a face of
the smaller cube completely rests on F. Compute the surface area of the resulting figure.

15. Let A + R = 17, R + M = 1, and M + L = 24. Compute A + L.

14. Let T = T N Y W R. The union of the intervals [4, T ] and [T −40, 17] is a closed interval of length L. Compute L.

13. Let T = T N Y W R. Compute the value of K for which

√ T √4
T = K + 60 .
20

12. Let T = T N Y W R. Circle O has center (T, T ) and radius T . Two lines are tangent to circle O at points (x1 , y1 )
x1 x2 + y1 y2
and (x2 , y2 ) and each line has slope −1. Compute .
T
11. Let T = T N Y W R. Alex, Bailey, and Casey are making candied apples. The recipe calls for 1 bag of apples, 2
cups of butter, and 3 cups of caramel, and yields 8 candied apples. Each person has one of the three ingredients
to make one or more batches of apples, but alas, the amounts of their ingredients are not proportional, so there
will be leftovers of at least one ingredient. Alex has T8 bags of apples, Bailey has 40 cups of butter, and Casey
has T − 17 cups of caramel. Compute the number of candied apples they can make.
10. Let T = T N Y W R. Given that sin(T ◦ ) = cos(10◦ + K ◦ ), compute the smallest positive value of K.
9. Let T = T N Y W R. An ARML coach wants to choose a three-person relay team (without regard to order).
If the coach had T + 1 students tolchoosem from, she could form N more relay teams than if she only had T
√
students to choose from. Compute 2N .

30
8. Let A1 be the number you will receive from position 7 and let A2 be the number you will receive from position
9. Let a1 be the smallest prime factor of A1 and let a2 be the largest prime factor of A2 . Consider the
complex numbers z1 = a1 + b1 i and z2 = a2 + b2 i. Given that |z2 |2 = 2|z1 |2 and that b1 and b2 are positive
integers where b1 + b2 is as small as possible, compute the value of |z2 |2 .

31
13 Super Relay Answers
1. 40
2. 40
3. 40

4. 40
5. 40
6. 40
7. 40

15. 40
14. 40
13. 40

12. 40
11. 40
10. 40

9. 40

8. 26

32
14 Super Relay Solutions

Problem 1. Triangle T AR has a right angle at R, with RT = 3 and AT = 5. Triangle LM N is similar to triangle
T AR, and M N = 8. Compute AR · M L.

MN 8
Solution 1. By the Pythagorean Theorem, AR = 4. By similar triangles, M L = AT · AR = 5· 4 = 10. Thus
AR · M L = 4 · 10 = 40.

Problem 2. Let T = T N Y W R. Compute the unique fixed point of the function f defined by f (x) = T ·(x−T +1).
[Note: A fixed point of a function g(x) is a value z such that g(z) = z.]

Solution 2. Let x = T · (x − T + 1). Then x(T − 1) = T 2 − T = T (T − 1). Thus either T = 1 (in which case, x is
not uniquely determined) or x = T . With T = 40, it follows that x = 40.

Problem 3. Let T = T N Y W R. A dessert tray containing T chocolate-covered strawberries has been ordered
for an ARML coaches meeting. Marlys eats some of the strawberries, leaving 24 strawberries for the other
attendees to consume. Then Marlys will have eaten K% of the strawberries. Compute the value of K.

T −24 K 16
Solution 3. The given information implies that T = 100 . With T = 40, K = 100 · 40 = 40.

T
Problem 4. Let T = T N Y W R, and let a = log 8 . Given that log K = 3 − 2a, compute the value of K.

Solution 4. Note that log K = log 1000 − log( T8 )2 = log 64000

T2 . Thus K = 64000
T2 , and with T = 40, K = 40.

Problem 5. Let T = T N Y W R. In the diagram to the right, circle K has area T π

and is internally tangent to circle J at point E. Chord XL is tangent to
circle K at point O. Given that the product of the radii of circles J and K
is 4T − 20, compute XL.

Solution 5. Let R and r be the radii of the larger and smaller circles, respectively. Then JO = R − 2r, and
XO2 = JX 2 − JO2 = R2 − (R − 2r)2 = 4Rr − 2
√ 4r . The given information implies
√ that r2 = T , so
2
XO = 4(4T − 20) − 4T = 12T − 80. Thus XL = 2 12T − 80. With T = 40, XL = 2 480 − 80 = 40.
−−→
Alternate Solution: Instead of using the Pythagorean Theorem, extend EO to intersect circle J at point D.
Then by the Power-of-the-Point theorem applied to O, XO2 = DO · OE = (2R − 2r)(2r) = 4Rr − 4r2 . As
above, this equation is equivalent to XO2 = 12T − 80, hence XL = 40.

Problem 6. Let T = T N Y W R. Mary places 20 green marbles into a hat, Lydia places 15 purple marbles into a
hat, and Joshua places T blue marbles into a hat. If a marble is randomly selected from the hat, the probability
8
that it will be purple is K . Compute the value of K.

33
8 15 280+8T 600
Solution 6. The given information implies that K = 35+T , so K = 15 . With T = 40, K = 15 = 40.

Problem 7. √ Let T = T N Y W R, and let K be the√ smallest prime factor of T . Suppose that a large cube has edge
length 3K. A smaller cube has edge length K − 1 and is placed on a face F of the larger cube so that a
face of the smaller cube completely rests on F. Compute the surface area of the resulting figure.

Solution 7. The larger cube has a surface area of 6(3K), and the smaller cube has a surface area of 6(K − 1).
When the smaller cube is placed atop the larger cube, the surface area increases by 4K − 4, so that the total
surface area of the resulting solid is 18K + 4K − 4 = 22K − 4. With T = 40, K = 2, hence the answer is
22 · 2 − 4 = 40.

Problem 15. Let A + R = 17, R + M = 1, and M + L = 24. Compute A + L.

Solution 15. Note that A + L = (A + R) + (M + L) − (R + M ) = 17 + 24 − 1 = 40.

Problem 14. Let T = T N Y W R. The union of the intervals [4, T ] and [T − 40, 17] is a closed interval of length L.
Compute L.

Solution 14. Because it is given that the union of the two intervals is itself an interval, it follows that L =
max(T, 17) − min(4, T − 40). With T = 40, conclude that L = 40 − 0 = 40.

Problem 13. Let T = T N Y W R. Compute the value of K for which

√ T √4
T = K + 60 .
20

T4 204
Solution 13. Raise each side to the fourth power to obtain T 2 = 204 (K + 60). Thus K = T2 − 60. With T = 40,
it follows that K = 160000
1600 − 60 = 100 − 60 = 40.

Problem 12. Let T = T N Y W R. Circle O has center (T, T ) and radius T . Two lines are tangent to circle O at
x1 x2 + y1 y2
points (x1 , y1 ) and (x2 , y2 ) and each line has slope −1. Compute .
T

Solution 12. Because the circle is centered at (T, T ), it is symmetric about the line y = x. Because the tangent
lines have a slope of −1, the slopes of the segments joining the points of tangency to the center must equal
1. Thus x1 = y1 and x2 = y2 . The solutions of the equations (x − T )2 + (y − T )2 = T 2 and y = x are
T2 T2

T T
T2 x1 x2 + y1 y2 2 + 2
(x, y) = T ± 2 , T ± 2 . Thus x1 x2 = y1 y2 =
√ √ and = = T . With T = 40, the
2 T T
answer is 40.

Problem 11. Let T = T N Y W R. Alex, Bailey, and Casey are making candied apples. The recipe calls for 1 bag
of apples, 2 cups of butter, and 3 cups of caramel, and yields 8 candied apples. Each person has one of the
three ingredients to make one or more batches of apples, but alas, the amounts of their ingredients are not
proportional, so there will be leftovers of at least one ingredient. Alex has T8 bags of apples, Bailey has 40 cups
of butter, and Casey has T − 17 cups of caramel. Compute the number of candied apples they can make.

34
Solution 11. The ingredients are in the ratio 1 : 2 : 3. So, for example, if 40 ≥ 2 · T8 and T − 17 ≥ 3 · T8 , then
8 · T8 = T candied apples can be made. Similarly, if T8 ≥ 40 3 40
2 and T − 17 ≥ 2 · 40, then 2 · 8 = 160 candied
apples can be made. Finally, if T8 ≥ 31 (T − 17) and 20 ≥ 32 (T − 17), then T −17

3 · 8 candied apples can be
made. With T = 40, only the first pair of inequalities is satisfied, resulting in 40 candied apples.

Problem 10. Let T = T N Y W R. Given that sin(T ◦ ) = cos(10◦ + K ◦ ), compute the smallest positive value of K.

Solution 10. Note that sin(T ◦ ) = cos(90 − T ◦ ). Assuming that 0 ≤ T ≤ 80, the smallest value of K will occur
when 90 − T = 10 + K, in which case, K = 80 − T . With T = 40, K = 40.

Problem 9. Let T = T N Y W R. An ARML coach wants to choose a three-person relay team (without regard to
order). If the coach had T + 1 students tol choose from, she could form N more relay teams than if she only
√ m
had T students to choose from. Compute 2N .

(T +1)T (T −1)
T +1 T
− T (T −1)(T −2) T (T −1) T (T −1)

Solution 9. It follows that N = 3 − 3 = 6 6 = · (T + 1 − T + 2) = .
√ l√ m l6 √ m 2

Thus 2N = T (T − 1), and T − 1 < 2N < T , so 2N = T . With T = 40, 2N = 40.

Problem 8. Let A1 be the number you will receive from position 7 and let A2 be the number you will receive from
position 9. Let a1 be the smallest prime factor of A1 and let a2 be the largest prime factor of A2 . Consider
the complex numbers z1 = a1 + b1 i and z2 = a2 + b2 i. Given that |z2 |2 = 2|z1 |2 and that b1 and b2 are positive
integers where b1 + b2 is as small as possible, compute the value of |z2 |2 .

Solution 8. The given information implies that a22 + b22 = 2(a21 + b21 ). With A1 = A2 = 40, it follows that a1 = 2
and a2 = 5. Thus 25 + b22 = 8 + 2b21 , so 2b21 − b22 = 17 (∗). By inspection, b1 = 3, b2 = 1 is a solution, and
it is easy to check that all ordered pairs of positive integers (b1 , b2 ) that yield smaller values of b1 + b2 do not
satisfy (∗). Thus |z2 |2 = 52 + 12 = 26.
Author’s Note: 2015 marks ARML’s fortieth anniversary, which we felt compelled to commemorate in a special
Super Relay “XL”. Unlike every other answer on this Super Relay, the final answer is 26, not 40. This problem is
dedicated to ARML’s head author, P.J. Karafiol, whose favorite number is 17 — and sure enough, 2617 = 4010 .

35
15 Tiebreaker Problems

Problem 1. Compute the ordered pair of integers (a, b) that minimizes a4 + ab − 2015 .

Problem 2. Compute the number of ways in which the numbers 1, 2, . . . , 8 can be filled into a 3 × 3 array, leaving
the center unoccupied, such that if a is anywhere to the left of b and in the same row, or if a is anywhere
above b and in the same column, then a < b.

Problem 3. Let S = {15, 24, 30, 40, 50, 60, 80}. When one element, k, is removed from S, the product of the
remaining elements is a perfect cube. Compute k.

36
16 Tiebreaker Answers

Answer 1. (−7, 3)

Answer 2. 18

Answer 3. 60

37
17 Tiebreaker Solutions

Problem 1. Compute the ordered pair of integers (a, b) that minimizes a4 + ab − 2015 .

Solution 1. Let f (a, b) = a4 + ab . Note that for fixed values of b, f (a, b) is an increasing function of a when a
is positive (this claim is true even if b is negative, because in that case, (a + 1)4 − a4 > 1 > ab − (a + 1)b ).
Further, f (7, b) = 2401 + 7b > 2015. Note that if a = 7, then decreasing b will always decrease the value of
f (a, b), but will always satisfy f (7, b) > 74 > 2015. In other words, the answer cannot be of the form (7, b),
because (7, b − 1) would make |f (a, b) − 2015| smaller. Thus it suffices to only consider values of a less than 7.
For each positive integer a = 1, . . . , 6, compute the value of b for which f (a, b) is nearest to 2015 in absolute
value. These values are shown in the table below.

a b f (a, b) |f (a, b) − 2015|

1 any value 2 2013
2 11 2064 49
3 7 2268 253
4 5 1280 735
5 4 1250 765
6 3 1512 503

Thus the ordered pair (a, b) minimizing |f (a, b) − 2015| is either (2, 11), or has a < 0. Moreover, if a < 0, but
b is even, then the ordered pair (a, b) produces the same value as the ordered pair (−a, b), and the previous
work also applies. Thus either (2, 11) is the correct ordered pair, or a < 0 and b is odd. Therefore the term ab
is negative. If b > 4, then a4 + ab < 0, and no negative value of a will cause f (a, b) to differ from 2015 by less
than 49, so it must be that b < 4. If b < 0, then ab ≤ 1, and f (a, b) − a4 ≤ 1. By the triangle inequality,

|f (a, b) − 2015| ≥ a4 − 2015 − 1.

As a must be an integer, this quantity is optimized at a = −7, yielding |f (a, b) − 2015| ≥ 385, which is larger
than the value of 49 produced by (2, 11). It follows then, that either b = 1 or b = 3, and that a < 0.

Note that both f (a, 1) and f (a, 3) increase as a is decreased (provided that a ≤ −1). Thus it suffices to
compute successively increasing values of f (a, 1) and f (a, 3) until they become larger than 2015.

f (−1, 1) = 0 f (−1, 3) = 0
f (−2, 1) = 14 f (−2, 3) = 8
f (−3, 1) = 78 f (−3, 3) = 54
f (−4, 1) = 252 f (−4, 3) = 192
f (−5, 1) = 620 f (−5, 3) = 500
f (−6, 1) = 1290 f (−6, 3) = 1080
f (−7, 1) = 2394 f (−7, 3) = 2058

The value nearest to 2015 is 2058. Moreover, 2058 < 2064, so the correct ordered pair is (−7, 3).

Solution 2. The 1 must appear in the top-left cell of the array. Similarly, the 8 must appear in the bottom-right
cell of the array. This leaves six cells remaining which are split into two groups of three: the three cells nearest
the bottom-left corner of the array, and the three cells nearest the top-right corner of the array. Given any
three numbers from {2, 3, . . . , 7}, there is exactly one way to put them into one of these groups to satisfy the

38
conditions, disregarding any relationships between the two groups. There are therefore 63 = 20 such ways

to put numbers into this array. However, it cannot be that the largest number in one group is smaller than
the smallest number in another group, otherwise either the middle column or the middle row would fail the
condition. There are two such possibilities which must be subtracted, yielding a total of 18 possibilities.

Problem 3. Let S = {15, 24, 30, 40, 50, 60, 80}. When one element, k, is removed from S, the product of the
remaining elements is a perfect cube. Compute k.

Solution 3. Factoring the given numbers, the set S is

{3 · 5, 23 · 3, 2 · 3 · 5, 23 · 5, 2 · 52 , 22 · 3 · 5, 24 · 5}.

The product of all of these is 214 · 34 · 57 . If the product is to be a perfect cube, all exponents must be multiples
of 3, thus there is an excess of 22 · 3 · 5 = 60.

39
ARML Competition 2016

Paul J. Karafiol, Head Writer

Andy Soffer, Associate Head Writer
Chris Jeuell, Lead Editor
Paul Dreyer
Edward Early
Zuming Feng
Benji Fisher
Zachary Franco
Winston Luo
Andy Niedermaier
George Reuter
Graham Rosby
Eric Wepsic

June 3–4, 2016

1 Team Problems

x
Problem 1. Compute all values of a for which the intersection of the graphs of y ≥ 2 and y ≤ a|x| + 17 is a
region having area 51.

Problem 2. Let f (x) = logb x and let g(x) = x2 − 4x + 4. Given that f (g(x)) = g(f (x)) = 0 has exactly one
solution and that b > 1, compute b.

Problem 3. Compute all solutions to the following system of equations:

2x
|y| − = −1
|x|
x|x| + y|y| = 24.

Problem 4. Two players are playing a game as follows. Integers N and M are each selected independently at
random from the set {1, 2, . . . , 50}. The players begin with an N × M grid of empty squares. A turn consists
of filling in a single empty square, then filling in all the squares in the same row, and then filling in all the
squares in the same column. The players alternate taking turns until all squares are filled in. Compute the
probability that the player who goes first fills in the last square.

Problem 5. A sequence T1 , T2 , . . . is called (a, b)-nacci if T1 = a, T2 = b, and for n ≥ 3, Tn = Tn−1 + Tn−2 .

Compute the number of (a, b)-nacci sequences of positive integers such that b > a and there exists an integer
k ≥ 3 with Tk = 50.

Problem 6. Complete the number puzzle below. Clues are given for the four rows. (Answers may not begin with
a zero.) Cells inside a region must all contain the same digit, and each region contains a different digit.

1. Product of two primes

2. Multiple of 19

3. A perfect square
4. Multiple of 1-Across

Problem 7. Real numbers x, y, and z are chosen at random from the unit interval [0, 1]. Compute the probability
that max{x, y, z} − min{x, y, z} ≤ 32 .

Problem 8. Square ARM L has sides of length 11. Collinear points X, Y , and Z are in the interior of ARM L
such that 4LXY and 4RXZ are equilateral. Given that Y Z = 2, compute [RLX].

Problem 9. Compute the number of permutations x1 , x2 , . . . , x10 of the integers −3, −2, −1, . . . , 6 that satisfy the
chain of inequalities
x1 x2 ≤ x2 x3 ≤ · · · ≤ x9 x10 .

1000
Problem 10. Compute the greatest integer n for which the decimal representation of n is of the form
1 . T A S T Y 7 . . ., where the digits T , A, S, and Y are not necessarily distinct.

1
2 Answers to Team Problems

Answer 1. − 31
6

√
Answer 2. 3

√
Answer 3. (5, −1) and ( 23, 1)

51
Answer 4. 100

Answer 5. 37

4 6 6 6
4 4 4 6
Answer 6.
4 3 5 6
9 3 3 2

20
Answer 7. 27

√
119 3
Answer 8. 6

Answer 9. 240

Answer 10. 968

2
3 Solutions to Team Problems

x
Problem 1. Compute all values of a for which the intersection of the graphs of y ≥ 2 and y ≤ a|x| + 17 is a
region having area 51.

Solution 1. For a < 0, the graphs of the two functions are shown in the figure below.

In the first quadrant, the two bounding functions are y = x2 and y = ax+17, so they intersect where x2 = ax+17.
Moreover, because the triangle in the first quadrant is half of the total area it has area 512 . Taking the side
along the y-axis as the base, the height is the x-value where the two lines intersect, hence 21 · 17 · x = 51
2 , thus
x = 3. Now 23 = 3a + 17, so a = − 316 .

Problem 2. Let f (x) = logb x and let g(x) = x2 − 4x + 4. Given that f (g(x)) = g(f (x)) = 0 has exactly one
solution and that b > 1, compute b.

Solution 2. Consider the equation f (g(x)) = 0. Because f (x) = logb (x), this equation holds precisely when
g(x) = 1, no matter what b is. Setting g(x) = x2 −4x+4 equal to 1 yields the quadratic equation x2 −4x+3 = 0,
which has two solutions, x = 3 and x = 1. Now consider the equation g(f (x)) = 0: g(x) has one root at 2,
so f (x) = 2, which implies that x √= b2 . Therefore, in order for the equation f (g(x)) = g(f (x)) = 0 to have
exactly one solution, b must equal 3.

Problem 3. Compute all solutions to the following system of equations:

2x
|y| − = −1
|x|
x|x| + y|y| = 24.

Solution 3. Note that x must be real because otherwise, the first equation would not be satisfied.
2x
Case 1: x is positive. Then = 2, and the first equation yields y = ±1. The second equation then becomes
|x|
2
√
x ± 1 = 24, so x = 5 (when y = −1) or x = 23 (when y = 1).
2x
Case 2: x is negative. Then |x| = −2, but in this case, the first equation becomes |y| = −3, which is impossible.
√
Thus the two solutions are (5, −1) and ( 23, 1).

3
Problem 4. Two players are playing a game as follows. Integers N and M are each selected independently at
random from the set {1, 2, . . . , 50}. The players begin with an N × M grid of empty squares. A turn consists
of filling in a single empty square, then filling in all the squares in the same row, and then filling in all the
squares in the same column. The players alternate taking turns until all squares are filled in. Compute the
probability that the player who goes first fills in the last square.

Solution 4. The outcome of the game is determined by the parity of min(N, M ). This is because each time a
player takes a turn, a row and a column that are previously not completely filled in (i.e., the row and col-
umn that contain the square selected on that turn) are now both completely filled in. Therefore, as soon
as min(N, M ) turns are taken, either all the rows will be completely filled, or all the columns, which means
that the entire grid is filled in. Therefore the first player will fill in the last square if and only if min(N, M ) is odd.

To compute the probability than min(N, M ) is odd, count the number of pairs (N, M ) satisfying this con-
dition, and then divide by 50 × 50 = 2500, the total number of possible choices of N and M . First consider the
case when N < M . There are 25 choices for N , from 1 to 49, and each results in 50 − N choices for M (i.e.,
the number of numbers larger than N ). So there are 49 + 47 + . . . + 3 + 1 = 252 = 625 such pairs (N, M ). The
case when M < N is symmetric, so there are another 625 cases there. Then there are another 25 cases where
N = M , because N and M can equal any odd number between 1 and 50. This is a total of 625+625+25 = 1275
1275 51
pairs, so the desired probability is 2500 = 100 .

Problem 5. A sequence T1 , T2 , . . . is called (a, b)-nacci if T1 = a, T2 = b, and for n ≥ 3, Tn = Tn−1 + Tn−2 .

Compute the number of (a, b)-nacci sequences of positive integers such that b > a and there exists an integer
k ≥ 3 with Tk = 50.

Solution 5. The sequence Tk must grow at least as fast as the (1, 2)-nacci sequence 1, 2, 3, 5, 8, 13, 21, 34, 55, . . . ,
thus for k ≥ 9, Tk > 50.

If k = 3, then a + b = 50, which has 24 solutions with a being any integer from 1 to 24 inclusive.
If k = 4, then a + 2b = 50. Now a must be even and a ≤ 16, so there are 8 solutions for k = 4.
If k = 5, then 2a + 3b = 50. Now a ≡ 1 (mod 3) and a < 10, so there are 3 solutions for k = 5.
If k = 6, then 3a + 5b = 50. Now a must be a multiple of 5 and the only solution is (5, 7).
If k = 7, then 5a + 8b = 50. Now b must be a multiple of 5 and the only solution is (2, 5).
If k = 8, then 8a + 13b = 50, which has no solutions with a < b.

Thus the desired number of (a, b)-nacci sequences is 24 + 8 + 3 + 1 + 1 = 37.

1. Product of two primes

2. Multiple of 19

3. A perfect square
4. Multiple of 1-Across

Solution 6. Consider first 1-Across, which is of the form A B B B. 4-Across is a multiple of 1-Across, and of
a different form, so they are not equal. Because both are four-digit numbers, A must be less than 5.

Next look at 2-Across. It is a four-digit multiple of 19 of the form A A A B, where A and B are the same digits

4
as above. Therefore 2-Across equals 1110A + B, which is equivalent to 8A + B mod 19. Because 2-Across
is a multiple of 19 and A < 5, the possibilities for (A, B) are (2, 3) and (4, 6).

Therefore 1-Across is either 2333 or 4666. The four-digit multiples of 2333 are 4666, 6999, and 9332. Only
9332 fits the pattern of 4-Across, so 4-Across is 9332. Because the digits in each region are distinct, 1-
Across must be 4666.

Then 2-Across is 4446. Filling in the digits that are the same in each region, 3-Across must be 43X6,
where X is 0, 1, 5, 7, or 8, because these are the only unused digits. Noting that 3-Across is a perfect square
and that 4356 = 662 , 3-Across is 4356.

4 6 6 6
4 4 4 6
4 3 5 6
9 3 3 2

Problem 7. Real numbers x, y, and z are chosen at random from the unit interval [0, 1]. Compute the probability
that max{x, y, z} − min{x, y, z} ≤ 23 .

Solution 7. Consider points lying in a unit cube in the first octant with one corner at the origin. To find the
volume of the region of points where the inequality holds, take cross-sections for each value of z from 0 to 1.
For any particular value of z, the values of x and y must satisfy 0 ≤ x ≤ 1, 0 ≤ y ≤ 1, − 32 ≤ y − x ≤ 23 ,
z − 23 ≤ x ≤ z + 23 , and z − 23 ≤ y ≤ z + 23 .

Depending on the value of z, the resulting region is a square of area 49 (when z = 0 or 1), a hexagon of
4(1−z)
area 94 + 4z 1 8 1 2 4
3 (when 0 < z < 3 ), a hexagon of area 9 (when 3 ≤ z ≤ 3 ), or a hexagon of area 9 + 3 (when
2 1
3 < z < 1). The region where z = 6 is shown.

In the cases where z is not between 13 and 32 , the area changes linearly with z and hence can be averaged
to 12 49 + 89 = 23 . This area corresponds to 23 of the z-values, with the other 31 having area 89 , for a total
probability of 23 ( 32 ) + 13 ( 89 ) = 20
27 .

5
Problem 8. Square ARM L has sides of length 11. Collinear points X, Y , and Z are in the interior of ARM L
such that 4LXY and 4RXZ are equilateral. Given that Y Z = 2, compute [RLX].

Solution 8. First note that X cannot be between Y and Z, because then the side lengths of triangles LXY and
RXZ would each be less than Y Z = 2, which√would mean that XL and XR would each be less than 2, which
is impossible because XL + XR ≥ LR = 11 2. Without loss of generality, let Y lie between X and Z. Let
XY = Y L = LX = x, so that XZ = ZR = RX = x + 2. In triangle RLX, angle LXR has measure 120◦
because angles LXY and ZXR are each 60◦ . Applying the Law of Cosines then gives
√
x2 + (x + 2)2 − (11 2)2 1
= cos 120◦ = − .
2x(x + 2) 2
q
Solving this quadratic for x gives x = −1 ± 241 , and the positive solution for x is used because it is a side
q q 3
length. So x = −1 + 241 3 and x + 2 = 1 +
241
3 , and the area [RLX] is

√ √
1 1 241 3 119 3
x(x + 2) sin 120◦ = · −1 · = .
2 2 3 2 6

Problem 9. Compute the number of permutations x1 , x2 , . . . , x10 of the integers −3, −2, −1, . . . , 6 that satisfy the
chain of inequalities
x1 x2 ≤ x2 x3 ≤ · · · ≤ x9 x10 .

Solution 9. First note that no two negative numbers can occur consecutively because otherwise, either the re-
sulting positive product would have to be followed by a non-positive product once the negatives are exhausted,
or it would have to occur at the end but there are too many positive terms for that to be possible. Thus the
sequence must start with alternating positive and negative values. Each inequality xi xi+1 ≤ xi+1 xi+2 implies
that xi < xi+2 if xi+1 > 0 and xi > xi+2 if xi+1 < 0 (the inequalities are strict because the xi s are distinct).
The two main cases to consider are x1 = −3 and x2 = −3.

If x1 = −3, then x3 = −2, x5 = −1, and either x6 = 0 or x7 = 0. If x6 = 0, then the positive entries
6!
(x2 , x4 , x7 , x8 , x9 , and x10 ) must satisfy x2 > x4 , x7 < x9 , and x8 < x10 . There are 2!·2!·2! = 90 such sequences.
If x7 = 0, then the positive entries (x2 , x4 , x6 , x7 , x8 , and x10 ) must satisfy x2 > x4 > x6 , and x8 < x10 for
6!
3!·2!·1! = 60 more.

If x2 = −3, then x4 = −2, x6 = −1, and either x7 = 0 or x8 = 0. If x7 = 0, then the positive entries
(x1 , x3 , x4 , x8 , x9 , and x10 ) must satisfy x1 > x3 > x5 , and x8 < x10 for another 60. If x8 = 0, then the positive
6!
entries (x1 , x3 , x5 , x7 , x9 , and x10 ) must satisfy x1 > x3 > x5 > x7 , for 4!·1!·1! = 30 more.

The desired number of permutations is therefore 90 + 60 + 60 + 30 = 240.

1000
Problem 10. Compute the greatest integer n for which the decimal representation of n is of the form
1 . T A S T Y 7 . . ., where the digits T , A, S, and Y are not necessarily distinct.

1000
Solution 10. In order for the integral part of n to be 1, n must be between 501 and 1000. Let m = 1000 − n.
Then
1000 1000 1 m m2 m3
= = m =1+ + 2
+ + ....
n 1000 − m 1 − 1000 1000 1000 10003
For example, if n = 998, then 1000
n begins 1.002004008016032064 . . . and each block of 3 digits following the
decimal point is a power of 2 until they begin to overlap.

6
To maximize n, first consider attempting to find a solution with T = 0, which implies m < 100, by the
previous analysis, because the first block of 3 digits equals either m or m plus the thousands digit of m2 .
Because the ones digit of the second block of 3 digits is 7, and m2 cannot end in 7 (as no squares end in 7),
the m3 block must overlap into the m2 block. Therefore m3 > 1000, so m > 10.

Because T = 0 by (optimistic) assumption, and therefore the part of m2 in the second block must have a
zero in the hundreds place, the goal is to find four-digit squares with a zero in the hundreds place. (Note that
m cannot have a two-digit square because m > 10, and the m3 block will not overlap onto the 0 from the
hundreds place of the second block as long as m3 < 100,000.)

The first square with a zero in the hundreds place is 322 = 1024, which corresponds to m = 32, m2 = 1024,
1000
m3 = 32768. Then the expansion for 1000−m begins

32 1024 32768
1+ + 2
+ + ···
1000 1000 10003
= 1 + 0.032 + 0.001024 + 0.000032768 + · · ·
= 1.033056768 + · · · .

This fits the pattern, except the sixth decimal digit is a 6 instead of a 7. However, m4 = 10242 is just over
1 million, so its first digit overlaps onto the 6, making it a 7. Therefore m = 32, which corresponds to n = 968
giving the maximal n satisfying the given criteria. This can be confirmed by computing 1000 968 = 1.033057 . . . .

Note: The same method as in the solution can be used for negative integers m that are close to 0; the
geometric series would have alternating signs. Also, replacing the 1000s with 10p s yields a similar result, but
instead of blocks of 3 digits apiece, the blocks will each have p digits.

7
4 Power Question 2016: Taxicab Geometry
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part.
To receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “com-
pute,” you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show
your work or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a
problem says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered
items may be used in solutions to later numbered items, but not vice versa. Pages submitted for credit should be
NUMBERED IN CONSECUTIVE ORDER AT THE TOP OF EACH PAGE in what your team considers to be
proper sequential order. PLEASE WRITE ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM
NUMBER (not the team name) on the cover sheet used as the first page of the papers submitted. Do not identify
the team in any other way.

In standard coordinate (Euclidean) geometry, if A = (x1 , y1p

) and B = (x2 , y2 ), then the distance between A and B
is given by the Euclidean distance function: dE (A, B) = (x2 − x1 )2 + (y2 − y1 )2 . In this Power Question, we
investigate what happens when distance is measured using the taxicab distance function,

dT (A, B) = |x2 − x1 | + |y2 − y1 |.

This distance is called the taxi-distance between A and B.

1. This problem will help you become familiar with taxi-distances.

a. Compute the taxi-distance between the points (2, 6) and (7, −2). [1 pt]
b. Compute the coordinates of a point whose Euclidean distance from (3, 2) is 5 and whose taxi-distance
from (3, 2) is also 5. [1 pt]
c. Compute the coordinates of a point whose Euclidean distance from (3, 2) is 5 and whose taxi-distance
from (3, 2) is 7. [1 pt]

In Euclidean geometry, one possible definition of the line segment AB is the set of points P such that AP +P B = AB,
or dE (A, P ) + dE (P, B) = dE (A, B). This definition does not work as well in taxicab geometry!
2. Let A = (0, 4) and B = (4, 0).
a. Show that P = (1, 3) satisfies dT (A, P ) + dT (P, B) = dT (A, B). [1 pt]
b. Find the coordinates of a point P not on the line x + y = 4 satisfying dT (A, P ) + dT (P, B) = dT (A, B).
[1 pt]
c. Find the set of all points P such that dT (A, P ) + dT (P, B) = dT (A, B), and justify your answer. [3 pts]
As in Euclidean geometry, two triangles are congruent if and only if the lengths of corresponding sides and the
measures of corresponding angles are equal. In Euclidean geometry, this measure-congruence is equivalent to
transformation-congruence: two triangles are transformation-congruent if and only if one is the image of the other
under a composition of reflections, rotations, and translations. (Actually, reflections alone suffice, but that result is
beyond the scope of this Power Question.) For the remainder of this Power Question, the symbol ∼ = will be used
exclusively to denote measure-congruence; we write ∼ =T to emphasize that we are using taxicab measure.
3. Let’s investigate whether the equivalence between measure- and transformation-congruence holds in taxicab
geometry. Suppose that A = (xA , yA ) and similarly for other named points.
a. If 4A0 B 0 C 0 is the image of 4ABC under the translation (x, y) 7→ (x + h, y + k), are the taxi-lengths of
the sides of 4A0 B 0 C 0 equal to those of 4ABC? Justify your answer. [1 pt]
b. If 4A0 B 0 C 0 is the image of 4ABC under a reflection in the line y = k, are the taxi-lengths of the sides
of 4A0 B 0 C 0 equal to those of 4ABC? Justify your answer. [1 pt]
c. Let O = (0, 0), R = (8, 0), D = (0, 4), and let 4O0 R0 D0 be the image of 4ORD under a (Euclidean)
rotation of 45◦ counterclockwise about the origin. Is 4ORD ∼=T 4O0 R0 D0 ? Justify your answer. [2 pts]

8
4. Let O have coordinates (0, 0), let P have coordinates (1, 0), and let A be the point (4, 2). The point B has
coordinates (m, n), where m and n are both positive integers, m is relatively prime to n, and ∠BOA ∼= ∠AOP
(considered as Euclidean angles).
a. Compute the ordered pair (m, n). [2 pts]
b. Compute the taxi-lengths of the segments intercepted by ∠BOA and ∠AOP on the line x + y = 1. [2 pts]
The computation in the previous item is problematic, for the following reason. A taxi-circle is the set of all points
that are a fixed taxi-distance from a given point, and the segments whose lengths you computed in 4b are actually
arcs on a unit taxi-circle centered at the origin. So angles congruent in the Euclidean sense do not necessarily inter-
cept taxi-congruent arcs.

One way to fix this problem is to define the measure of an angle analogously to the way the radian measure of
an angle is defined in Euclidean geometry, namely, as a quotient. The numerator is the length of the arc that the
angle intercepts on a circle centered at the angle’s vertex. The denominator is the radius of that circle. If P and R
are equidistant from Q, then the measure of ∠P QR in taxi-radians is given by the formula
taxi-length of the arc from P to R on the taxi-circle centered at Q passing through P
mT ∠P QR = .
dT (Q, P )
5. Let’s flesh out the above definition with an example.
a. Find an equation to describe the taxi-circle centered at the origin of radius 17. [1 pt]
b. Compute the value of taxi-pi, πT , by dividing the semi-taxi-circumference of the taxi-circle from part 5a
by its radius. [1 pt]
c. Show that the value you computed in part 5b is independent of the taxi-circle’s radius. [2 pts]
d. Given that C = (3, 1), O = (0, 0), and B = (−1, 3), compute mT ∠COB. [1 pt]
Using taxi-circles and taxi-angles, define a rotation by θ taxi-radians about the origin as follows. The point P 0 is the
image of P under a counterclockwise rotation of θ taxi-radians around O = (0, 0) if and only if dT (O, P ) = dT (O, P 0 )
and the measure of the taxi-arc from P to P 0 (going counterclockwise around the taxi-circle centered at O) is
dT (O, P ) · θ.
6. Let O = (0, 0), A = (8, 0), and R = (3, 1).
a. Let 4O0 A0 R0 be the image of 4OAR under a rotation of 1 taxi-radian about the origin. Compute the
coordinates of O0 , A0 , and R0 . [3 pts]
0 0 0 ∼
b. Is 4O A R =T 4OAR? Justify your answer. [3 pts]
7. We now consider conditions for two triangles to be measure-congruent in Euclidean geometry and apply them
to taxicab geometry to discover whether or not measure-congruence still holds. Let 4M OD have vertices
M = (0, 2), O = (0, 0), and D = (2, 0).
a. Show that there exists a triangle 4CAB such that dT (C, A) = dT (M, O), dT (A, B) = dT (O, D), and
mT ∠CAB = mT ∠M OD, but 4M OD T 4CAB. (Thus SAS congruence does not hold in taxicab
geometry.) [2 pts]
b. Show that there exists a triangle 4P IG such that dT (P, I) = dT (M, O), dT (I, G) = dT (O, D), and
dT (P, G) = dT (M, D), but 4M OD T 4P IG. (Thus SSS congruence does not hold in taxicab geometry.)
[2 pts]
8. Define the taxi-cosine and taxi-sine of an angle as follows:
dE (C, A) · dE (C, B)
cosT ∠ACB = cos ∠ACB
dT (C, A) · dT (C, B)
dE (C, A) · dE (C, B)
sinT ∠ACB = sin ∠ACB,
dT (C, A) · dT (C, B)
where cos ∠ACB and sin ∠ACB are the ordinary (Euclidean) cosine and sine functions, respectively.

[3 pts]

9. Suppose that triangle ABC is given with AB parallel to the x-axis, and that the perpendicular dropped from
←→
C to AB intersects AB at C 0 and has taxi-length hT . As with conventional trigonometry, let aT = dT (B, C),
bT = dT (A, C), and cT = dT (A, B). Prove the following:
a. aT = bT + cT − 2bT cosT A. [3 pts]
b. bT = aT + cT − 2aT cosT B. [2 pts]
a2T + b2T − 2aT bT cosT C
c. cT = . [3 pts]
aT + bT
aT bT cT
10. Given 4ABC, prove that = = , where aT , bT , and cT are defined as in the previous
sinT A sinT B sinT C
problem. [4 pts]

10
5 Solutions to Power Question
1. Compute from the given definition:
a. |7 − 2| + |−2 − 6| = 13.
b. There are four points that can found by moving from (3, 2) along vectors parallel to the coordinate axes.
These points are (8, 2), (−2, 2), (3, 7), and (3, −3). One of these is sufficient to answer the question.
c. The classic 3−4−5 right triangle gives many more possibilities: (3 ± 3, 2 ± 4) and (3 ± 4, 2 ± 3). (One of
these is sufficient to answer the question.)
2. a. dT (A, P ) + dT (P, B) = |1 − 0| + |3 − 4| + |4 − 1| + |0 − 3| = 8 = |4 − 0| + |0 − 4| = dT (A, B).
b. Many solutions exist (see part c below), but (2, 4) is one such point.
c. Because dT (A, B) = 8, to move from A to B, find all points P (x, y) such that |x−0|+|y−4|+|4−x|+|0−y| =
8. If 0 ≤ x ≤ 4 and 0 ≤ y ≤ 4, the left hand side of the equation simplifies to x + (4 − y) + (4 − x) + y,
which equals 8 for all such x and y. On the other hand, if x < 0 and 0 ≤ y ≤ 4, then the left hand side of
the equation simplifies to 4 − 2x + (4 − y) + y = 8 − 2x, and for x < 0, 8 − 2x > 8. Analogous arguments
show that no points with x > 4, or y < 0, or y > 4 satisfy the condition. Hence the solution set is the
square 0 ≤ x ≤ 4, 0 ≤ y ≤ 4.

Alternate solution: Let P = (x, y), so that dT (A, P ) + dT (P, B) = |x − 0| + |y − 4| + |4 − x| + |0 − y| =

|4−x|+|x−0|+|4−y|+|y−0|. For any real number r, |r| ≥ r, and equality holds if and only if r ≥ 0. Apply-
ing this observation to each term in the above sum, dT (A, P )+dT (P, B) ≥ (4−x)+(x−0)+(4−y)+(y−0) =
8 = dT (A, B), with equality if and only if each of 4 − x, x − 0, 4 − y, y − 0 is nonnegative. Hence the
solution set is the square 0 ≤ x ≤ 4 and 0 ≤ y ≤ 4.
3. a. Let A = (x1 , y1 ) and let B = (x2 , y2 ). Then dT (A, B) = |x2 − x1 | + |y2 − y1 |. On the other hand,
A0 = (x1 +h, y1 +k) and B 0 = (x2 +h, y2 +k), yielding dT (A0 , B 0 ) = |(x2 +h)−(x1 +h)|+|(y2 +k)−(y1 +k)| =
|x2 − x1 | + |y2 − y1 | = dT (A, B). The same argument would apply to the taxi-lengths of BC and AC.
Therefore the taxi-lengths of the sides are preserved under translation.
b. Let A = (x1 , y1 ) and let B = (x2 , y2 ). The reflection takes (x, y) = (x, k + (y − k)) to (x, k − (y − k)) =
(x, 2k − y). Hence A0 = (x1 , 2k − y1 ) and B 0 = (x2 , 2k − y2 ). Thus dT (A0 , B 0 ) = |x2 − x1 | + |(2k − y2 ) −
(2k − y1 )| = |x2 − x1 | + |y1 − y2 | = dT (A, B). Analogous arguments apply to the taxi-lengths of BC and
AC, so the taxi-lengths of the sides are preserved under reflection in a horizontal line. (Note: It would
be straightforward to show the same result for reflection in a vertical line. What might be difficult about
reflecting in an oblique line?)
√ √ √ √
c. O0 = (0, 0), R0 = (4 2, 4 2), D √
0
= (−2 2, 2 2). Notice that dT (O, D) = 4, dT (D, R) = 12, and
dT (O, R) = 8, but dT (O0 , R0 ) = 8 2. Therefore 4ORD T 4O0 R0 D0 .
−→ −−→
4. a. Let θ be the angle between the positive x-axis and the ray OA. Then tan θ = 12 . The ray OB makes
an angle of 2θ with the positive x-axis, so using the double-angle identity from trigonometry yields
2 tan θ 1 4 4
tan 2θ = = = . Hence B lies on the line y = 3 x. Because m and n can have no
1 − tan2 θ 1− 2 1 2 3
prime factors in common, m = 3 and n = 4. The desired point is (3, 4).
−→ −−→ 2 1
3 4
2 1

b. The rays OA and OB intersect the line x+y = 1 at 3 , 3 and 7 , 7 respectively. Note that d T P, ( 3 , 3 ) =
2 1 2 2 1
3 4 3 2 4 1 10
|1 − 3 | + |0 − 3 | = 3 . However, dT 3 , 3 , 7 , 7 = | 7 − 3 | + | 7 − 3 | = 21 . Therefore central angles with
the same Euclidean angle measure can intercept arcs of different taxi-lengths on the unit taxi-circle.
5. a. The taxi-circle is the set of all points (x, y) such that |x| + |y| = 17. This equation describes a square
centered at the origin with vertices (±17, 0) and (0, ±17); the sides of the square lie on the lines x+y = 17,
x + y = −17, x − y = 17, and x − y = −17.
b. The semi-taxi-circumference of this circle is 68, because dT ((17, 0), (0, 17)) = dT ((0, 17), (−17, 0)) = 34.
(Note that the distance along the arc is, as usual, longer than the straight-line taxi-distance
dT ((17, 0), (−17, 0)), which is simply 34.) The radius of the circle is 17. So πT = 68/17 = 4.

11
c. In general, the taxi-circle of radius r centered at the origin is given by the equation |x| + |y| = r,
and one semicircle has endpoints (r, 0) and (−r, 0). Hence the arc length is simply dT ((r, 0), (0, r)) +
dT ((0, r), (−r, 0)) = 2r + 2r = 4r. Therefore πT = 4r/r = 4. (As an interesting exercise, you might try to
show that this result holds for any taxi-semicircle with endpoints (x, y) and (−x, −y).)
d. These points are both found on a taxi-circle of radius 4.
taxi-length of the arc from C to B on taxi-circle O 8
mT ∠COB = = = 2 taxi-radians.
dT (O, C) 4

6. a. First note that O0 = O because O is the center of rotation. To find A0 , note that dT (O, A) = 8, so the
arc from A to A0 on the circle centered at O passing through A must also have length 8, and if A0 is in
the first quadrant, then it lies on the line x + y = 8. These conditions are satisfied by the point (4, 4).
Similarly, R0 must lie on the line x + y = 4 and dT (R, R0 ) = 4. Assuming that 0 < x < 3 and 1 < y < 4
yields dT (R, R0 ) = (3 − x) + (y − 1) = y − x + 2. Hence the desired point satisfies both x + y = 4 and
y − x + 2 = 4, yielding x = 1, y = 3 as the unique solution. Thus R0 = (1, 3).

Alternate solution: First note that O0 = O because O is the center of rotation. Before considering the
other points, let P = (x, y) be any point in the first quadrant and let 0 ≤ r ≤ x. Then Q = (x − r, y + r)
is also in the first quadrant, Q and P are on the same taxi-circle centered at O, and dT (P, Q) = 2r.
Apply this observation to A = (8, 0) with r = 21 dT (O, A) = 4 to find A0 = (8 − 4, 0 + 4) = (4, 4). Similarly,
take P = B = (3, 1) and r = 12 dT (O, B) = 2 to find B 0 = (3 − 2, 1 + 2) = (1, 3).
b. By construction, dT (O, A) = dT (O, A0 ) and dT (O, B) = dT (O, B 0 ). However, dT (A, B) = |3−8|+|1−0| = 6
while dT (A0 , B 0 ) = |1 − 4| + |3 − 4| = 4. Hence the two triangles are not measure-congruent.
7. a. There are many examples. One such example is 4CAB with vertices C(−1, 1), A(0, 0), and B(1, 1). Then
dT (C, A) = dT (M, O) = 2, dT (A, B) = dT (O, D) = 2, and mT ∠A ∼
= mT ∠O = πT /2, but 4M OD T
4CAB because dT (C, B) = 2 6= dT (M, D).
b. There are many examples. One such example is 4P IG with vertices P (−1, 1), I(0, 0), and G(2, 0). Then
dT (P, I) = dT (M, O) = 2, dT (I, G) = dT (O, D) = 2, and dT (P, G) = dT (M, D) = 4, but 4M OD T
4P IG because mT ∠P IG > mT ∠M OD.
8. a. To use the formula, note that dE (O, A) = 3, dE (O, B) = 5, dT (O, A) = 3, dT (O, B) = 7, cos ∠AOB = 53 ,
and sin ∠AOB = 45 . Then cosT ∠AOB = 3·7 3·5 3
· 5 = 73 . (Similarly, sinT ∠AOB = 3·5 4 4
3·7 · 5 = 7 .)
b. Consider the diagram below.

O B' A

−→
Because OA is parallel to the x-axis, dE (O, A) = dT (O, A). Then

dE (O, A) · dE (O, B) dE (O, B)

cosT ∠AOB = cos ∠AOB = cos ∠AOB.
dT (O, A) · dT (O, B) dT (O, B)
←→
Let B 0 be the foot of the perpendicular from B to OA, so that dT (O, B) = dE (O, B 0 ) + dE (B 0 , B). Then

dE (O, B) 1
cosT ∠AOB = cos ∠AOB = dE (O,B 0 ) dE (B,B 0 )
cos ∠AOB.
dE (O, B 0 ) + dE (B 0 , B) +
dE (O,B) dE (O,B)

12
0 0
Substituting | cos ∠AOB| for ddEE(O,B ) dE (B,B )
(O,B) and | sin ∠AOB| for dE (O,B) yields the desired result. (The reader
may wish to show that, under these conditions, cosT ∠AOB + sinT ∠AOB = 1.)
←→
c. Let the feet of the altitudes from A and B to P Q be A0 and B 0 , respectively, as shown in the diagram below.

Q A' P B'

Then dT (Q, A) = dE (Q, A0 ) + dE (A0 , A) and dT (Q, B) = dE (Q, B 0 ) + dE (B 0 , B). Using the definition
yields
dE (Q, A) · dE (Q, B)
cosT ∠AQB = cos ∠AQB
dT (Q, A) · dT (Q, B)
1
= cos ∠AQB
dE (Q, A0 ) + dE (A, A0 ) dE (Q, B 0 ) + dE (B, B 0 )

·
dE (Q, A) dE (Q, B)
1
= cos ∠AQB.
dE (Q, A0 ) dE (A, A0 ) dE (Q, B 0 ) dE (B, B 0 )

+ · +
dE (Q, A) dE (Q, A) dE (Q, B) dE (Q, B)
dE (Q,A0 ) dE (A0 ,A) dE (Q,B 0 )
On the other hand, dE (Q,A) = | cos ∠AQP |, dE (Q,A) = | sin ∠AQP |, and similarly dE (Q,B) = | cos ∠BQP |,
dE (B 0 ,B)
dE (Q,B) = | sin ∠BQP |. Hence the right side can be rewritten as

dT (Q, A) = dE (Q, A0 ) + dE (A0 , A)

Lemma: Suppose that 4P QR is a right triangle with right angle at Q and P Q parallel to the x-axis. Then
cosT ∠P = ddTT (P,Q)
(P,R) .

13
cos ∠P
Proof of Lemma: Because P Q is parallel to the x-axis, the result from 8b applies, and cosT ∠P = | cos ∠P |+| sin ∠P | .
dE (P,Q) dE (Q,R)
In this situation, cos ∠P = dE (P,R) and sin ∠P = dE (P,R) , so substitute to obtain the following:
dE (P,Q)
dE (P,R)
cosT ∠P =
dE (P,Q) dE (Q,R)
dE (P,R) + dE (P,R)

dE (P, Q)
= .
|dE (P, Q)| + |dE (Q, R)|
However, P and Q lie on the same horizontal line and Q and R lie on the same vertical line, so dE (P, Q) =
dT (P, Q) and dT (P, R) = dE (P, Q) + dE (Q, R), yielding the desired result. 2

Consider the diagram below.

A C' B

a. Because dT (A, C) = dT (A, C 0 ) + dT (C 0 , C), it follows that dT (C, C 0 ) = dT (A, C) − dT (A, C 0 ) and analo-
gously that dT (C, C 0 ) = dT (B, C) − dT (B, C 0 ). Then dT (B, C) = dT (A, C) + dT (A, B) − 2 · dT (A, C 0 ) =
dT (A, C 0 )
bT + cT − 2dT (A, C 0 ). The lemma applies to angle A in 4AC 0 C, and cosT ∠CAC 0 = implies
dT (A, C)
dT (A, C 0 ) = dT (A, C) · cosT ∠A. Substituting this expression for dT (A, C 0 ) into the equation for dT (B, C)
yields the first equation.
b. The second equation is derived similarly. As before, dT (A, C) = dT (B, C) + dT (A, B) − 2 · dT (A, C 0 ) =
dT (B, C 0 )
aT + cT − 2dT (A, C 0 ). Again, apply the lemma to angle B in 4BC 0 C to obtain cosT ∠CBC 0 = ,
dT (B, C)
0
yielding dT (B, C ) = dT (B, C) · cosT ∠B. Substituting this result into the previous equation yields
dT (A, C) = aT + cT − 2aT cosT ∠B.
dE (A, C) · dE (B, C)
c. Apply the definition of taxi-cosine to obtain cosT ∠ACB = cos ∠ACB. By the Law
dT (A, C) · dT (B, C)
(dE (B, C))2 + (dE (A, C))2 − (dE (A, B))2
of Cosines applied to 4ABC, cos ∠C = . Hence
2 · dE (A, C) · dE (B, C)
dE (A, C) · dE (B, C) (dE (B, C))2 + (dE (A, C))2 − (dE (A, B))2
cosT ∠ACB = ·
dT (A, C) · dT (B, C) 2 · dE (A, C) · dE (B, C)
(dE (B, C)) + (dE (A, C)) − (dE (A, B))2
2 2
= .
2 · dT (A, C) · dT (B, C)
Use the Pythagorean Theorem in 4AC 0 C and 4BC 0 C to obtain
(dE (B, C 0 ))2 + (dE (C, C 0 ))2 + (dE (A, C 0 ))2 + (dE (C, C 0 ))2 − (dE (A, C 0 ) + dE (B, C 0 ))2
cosT ∠ACB = .
2 · dT (A, C) · dT (B, C)
h2T − pT (cT − pT )
By algebra, this equation is equivalent to cosT ∠ACB = , where pT = dT (A, C 0 ).
aT · bT
Now, obtain expressions for hT , pT and cT − pT by algebraic manipulation of previously-found equations

14
relating lengths of sides.
First, find an expression
for pT . Note that dT (A, C) = bT = aT + cT − 2aT cosT ∠B, which implies
dE (B, C) · (cT − pT ) cT − pT
bT = aT + cT − 2aT · . After cancellation, bT = aT + cT − 2(cT − pT ) →
aT · (cT − pT ) dE (B, C)
bT + cT − aT
bT = aT − cT + 2pT → pT = .
2
aT + cT − bT
A similar manipulation to the equation bT = aT + cT − 2(cT − pT ) yields cT − pT = .
2
aT + cT − bT 2aT aT + cT − bT
Now, notice that hT = aT − (cT − pT ) = aT − , which simplifies to − =
2 2 2
aT − cT + bT
, as needed. Substitution yields
2
2
bT + cT − aT bT + cT − aT aT + cT − bT
−
2 2 2
cosT ∠ACB = ,
aT · bT

a2T + b2T − cT (aT + bT )

which simplifies to cosT ∠ACB = .
2 · aT · bT
This equation can be algebraically manipulated to obtain the desired result.

Alternate solution: To simplify the computations, introduce notation for various Euclidean lengths. Let

a = dE (B, C), b = dE (A, C), c = dE (A, B) = cT ;

0 0
p = dE (A, C ) = dT (A, C ) = b cos A;
q = dE (B, C 0 ) = dT (B, C 0 ) = a cos B.

Then
aT = q + hT , bT = p + hT , cT = p + q.
b·c
a. From the definition of the taxi-cosine, bT cosT A = bT · bT ·cT cos A = b cos A = p. Therefore bT + cT −
2bT cosT A = (p + hT ) + (p + q) − 2p = q + hT = aT .
b. Similarly, aT cosT B = q, so aT + cT − 2aT cosT B = (q + hT ) + (p + q) − 2q = bT .
c. The Euclidean radian measures of the angles add up to π, so cos C = cos(π − A − B) = − cos(A + B) =
sin A sin B − cos A cos B. Substitute into the definition of the taxi-cosine and multiply by aT bT :
ab
aT bT cosT C = aT bT (sin A sin B − cos A cos B)
aT bT
= (b sin A)(a sin B) − (b cos A)(a cos B)
= h2T − pq
= h2T − (bT − hT )(aT − hT )
= (aT + bT )hT − aT bT .

Therefore a2T + b2T − 2aT bT cosT C = a2T + b2T − 2(aT + bT )hT + 2aT bT = (aT + bT )(aT + bT − 2hT ) =
(aT + bT )cT . Solving for cT gives the desired result.
10. In general, the area of a triangle ABC is equal to half of the area of the parallelogram determined by any two
dE (O, A) · dE (O, B)
sides of triangle ABC. The definition of taxi-sine, namely that sinT ∠AOB = sin ∠AOB,
dT (O, A) · dT (O, B)
implies that dT (O, A) · dT (O, B) · sinT ∠AOB = dE (O, A) · dE (O, B) · sin ∠AOB. The latter expression is equal
−* −−*
to |OA × OB|T . Thus the cross-product of two vectors can be interpreted in a taxicab space as it is in a

Alternate solution: As in the solution to problem 9, let a, b, and c denote the Euclidean lengths of the
sides. With this notation, the definition of the taxi-sine gives
ab ab sin C cT aT bT cT c
sinT C = sin C = , thus = .
aT bT aT bT sinT C abc sin C
aT bT cT
Hence the Taxicab Law of Sines follows from the Euclidean version, multiplying by .
abc

Note: Taxicab geometry is an active area of mathematical research, so many important questions remain unanswered.
Students who are interested in taxicab geometry might find the following sources helpful; both were inspirational in
the creation of this Power Question.
Eugene Krause, Taxicab Geometry: An Adventure in Non-Euclidean Geometry, Dover, 1975.
Ayşe Bayar, Süheyla Ekmekçi and Münevver Özcan, On Trigonometric Functions and Cosine and Sine
Rules in Taxicab Plane, International Electronic Journal of Geometry, 2 (2009), 17-24. Retrieved from
http://www.iejgeo.com/matder/dosyalar/makale-11/paper2(bayar-ekmekci-ozcan).pdf.

16
6 Individual Problems

Problem 1. Given that a, b, and c are positive integers such that ab · bc is a multiple of 2016, compute the least
possible value of a + b + c.

Problem 2. Triangle ABC is isosceles. An ant begins at A, walks exactly halfway along the perimeter of 4ABC,
and then returns directly to A, cutting through the interior of the triangle. The ant’s path surrounds exactly
90% of the area of 4ABC. Compute the maximum possible value of tan A.

√ √ √ √
b 4 1c · b 4 3c · b 4 5c · · · b 4 2015c
Problem 3. √
Compute 4 √ √ √ .
b 2c · b 4 4c · b 4 6c · · · b 4 2016c

Problem 4. Compute the number of permutations x1 , . . . , x6 of the integers 1, . . . , 6 such that xi+1 ≤ 2xi for all i,
1 ≤ i < 6.

Problem 5. Compute the least possible non-zero value of A2 + B 2 + C 2 such that A, B, and C are integers
satisfying A log 16 + B log 18 + C log 24 = 0.

Problem 6. In 4LEO, point J lies on LO such that JE ⊥ EO, and point S lies on LE such that JS ⊥ LE.
Given that JS = 9, EO = 20, and JO + SE = 37, compute the perimeter of 4LEO.

Problem 7. Compute the least possible area of a non-degenerate right triangle with sides of length sin x, cos x,
and tan x, where x is a real number.

Problem 8. Let P (x) be the polynomial x3 +Ax2 +Bx+C for some constants A, B, and C. There exist constants
D and E such that for all x, P (x + 1) = x3 + Dx2 + 54x + 37 and P (x + 2) = x3 + 26x2 + Ex + 115. Compute
the ordered triple (A, B, C).

Problem 9. An n-sided die has the integers between 1 and n (inclusive) on its faces. All values on the faces of
the die are equally likely to be rolled. An 8-sided die, a 12-sided die, and a 20-sided die are rolled. Compute
the probability that one of the values rolled is equal to the sum of the other two values rolled.

Problem 10. Compute the largest of the three prime divisors of 134 + 165 − 1722 .

17
7 Answers to Individual Problems

Answer 1. 14

√
Answer 2. 3 11

5
Answer 3. 16

Answer 4. 144

Answer 5. 105

Answer 6. 120

√
4 √
4
2 8
Answer 7. 2 − 4

Answer 8. (20, 11, 5)

23
Answer 9. 240

Answer 10. 1321

18
8 Solutions to Individual Problems

Problem 1. Given that a, b, and c are positive integers such that ab · bc is a multiple of 2016, compute the least
possible value of a + b + c.

Solution 1. Factoring yields 2016 = 25 · 32 · 7, which means that 2 · 3 · 7 = 42 | ab. To minimize a + b, first look
for cases where ab = 42. If a = 2, then b = 21, so setting c = 2 is sufficient, yielding 25 for the sum. If a = 3,
then b = 14, but c must be 5 (or greater) to satisfy the condition that 25 | ab bc , yielding a + b + c = 22. If
a = 6 and b = 7, then c = 1 is sufficient, yielding a + b + c = 14. This value is minimal because a = 7 forces
b = 6 and c ≥ 5, while the next larger value of a is 14. Thus the minimal value of a + b + c is 14.

Solution 2. Point A cannot be the vertex of the isosceles triangle, because if it were, the ant would reach the
midpoint of the base, and then the area surrounded by the ant’s path would be half the triangle’s area rather
than 90%. So A is one of the two base angles. Let P be the point reached by the ant and, without loss of
generality, let m∠B be the triangle’s vertex angle. Note that by the triangle inequality, P lies on BC. There
are two cases to consider: either the ant passes through vertex B first, making [AP B] greater than [AP C], or
the ant passes through vertex C first, making [AP C] greater than [AP B]. In the first case, because 4AP B
and 4AP C have collinear bases (along BC) and the same altitude (from point A), [AP B] PB
[AP C] = P C . Because
[AP B] 90% PB
[AP C] = 10% = 9,
= 9. Without loss of generality, set P B = 9 and P C = 1. Then AB = 10, and from
PC
1
√
AB + P B = 2 (AB + BC + AC), obtain AC = 18. By the Pythagorean Theorem, the
altitude to AC is 19,
√
19
and tan A =
√ 9 ≈ 0.5. In the second case,
√
analogous reasoning yields P B = 1, P C =
9, and AC = 2, so that
tan A = 199 > 0.5. Thus the answer is 3 11.

√ √ √ √
b 4 1c · b 4 3c · b 4 5c · · · b 4 2015c
Problem 3. Compute √ √ √ √ .
b 4 2c · b 4 4c · b 4 6c · · · b 4 2016c

√ √
Solution 3. Notice that b 4 n − 1c = b√4 nc except
√
when
√
n is a perfect fourth power. So all of the factors in the
4
15c b 4 255c b 4 1295c
product can be ignored except for bb √
4 , √ , √
16c b 4 256c b 4 1296c
. These fractions simplify to 21 , 34 , and 56 , respectively,
1 3 5
and their product is 2 · 4 · 56 , or 16
.

Problem 4. Compute the number of permutations x1 , . . . , x6 of the integers 1, . . . , 6 such that xi+1 ≤ 2xi for all i,
1 ≤ i < 6.

Solution 4. First notice that some pairs are impossible: 1 can only be followed by 2, and 2 can only be followed
by 1, 3, or 4. These observations suggest that the problem can be divided into several cases:
Case 1: x6 = 1. If x5 = 2, then there are 4! arrangements of the other numbers. If x5 6= 2, then holding either 23
or 24 as a consecutive pair yields 4! permutations. So there are 3 · 24 = 72 possible arrangements.
Case 2: x5 = 1 and x6 = 2. Then there are 4! = 24 arrangements of the other numbers.
Case 3: Either 123 or 124 occur in that order. Then there are 4! permutations of each cluster and the other three
numbers, yielding 2 · 4! = 48 possible arrangements.
Hence the desired number of permutations of the numbers is 144.

19
Problem 5. Compute the least possible non-zero value of A2 + B 2 + C 2 such that A, B, and C are integers
satisfying A log 16 + B log 18 + C log 24 = 0.

Solution 5. Use the laws of logarithms to obtain (4 log 2)A + (log 2 + 2 log 3)B + (3 log 2 + log 3)C = 0. Collecting
terms with common logarithm coefficients yields (4A + B + 3C) log 2 + (2B + C) log 3 = 0. Because 3 is not
a rational power of 2, in order to satisfy this equation with integral A, B, C, both 4A + B + 3C = 0 and
2B + C = 0. The second equation is equivalent to C = −2B, yielding 4A + B − 6B = 0 or 4A = 5B. Thus A
is divisible by 5 and B is divisible by 4. Set A = 5, B = 4, and C = −8 to obtain 52 + 42 + (−8)2 = 105.

Problem 6. In 4LEO, point J lies on LO such that JE ⊥ EO, and point S lies on LE such that JS ⊥ LE.
Given that JS = 9, EO = 20, and JO + SE = 37, compute the perimeter of 4LEO.

Solution 6. Optimistically, one might hope that SE = 12 and JO = 25, and in fact this turns out to be the case!
Let JO = x and SE = y. Then JE 2 = 92 + y 2 = 81 + y 2 and JE 2 + 202 = x2 , hence y 2 + 481 = x2 . Thus
x2 − y 2 = 481. Because x + y = 37, x − y = 481/37 = 13. Hence 2x = 50, x = 25, and y = 12. That information
by itself does not solve the problem, but the fact that the sides of right triangles JES and JOE are known
makes a combination of angle-chasing and trigonometry feasible.

Let m∠JOE = θ. Then m∠JES = θ because triangles JSE and JEO are both 3−4−5 right triangles. Then
m∠L = 180◦ − (m∠LEO + m∠LOE) = 180◦ − (90◦ + θ + θ) = 90◦ − 2θ. Hence tan L = tan1 2θ and sin L = cos 2θ.
2
2
Using double-angle identities for tangent and sine yields tan L = 1−tan θ 2
2 tan θ and sin L = cos θ − sin θ. In this
3 4 3
case, sin θ = 5 , cos θ = 5 , and tan θ = 4 . Substitute these values into the preceding equations to obtain
tan L = 7/16 7 7 9 216 9 225 441
6/4 = 24 , while sin L = 25 . Then LS = tan L = 7 and LJ = sin L = 7 . Thus LS + LJ = 7 = 63.
With SE = 12, EO = 20, JO = 25, the perimeter is 120.

Alternate Solution: Let t = SE. Then JE 2 = t2 + 92 = (37 − t)2 − 202 , which implies that t = 12. Note that
24
4JEO ∼ 4JSE. Let α = m∠EOJ; then m∠JES = α and m∠LJS = 2α. Thus sin 2α = 2(sin α)(cos α) = 25 ,
JS
and 4LJS is similar to a 7–24–25 triangle. Letting x = 9 , it follows that LJ + LS = 25x + 24x = 49x = 63.
Thus the perimeter of 4LEO is 63 + 37 + 20 = 120.

Problem 7. Compute the least possible area of a non-degenerate right triangle with sides of length sin x, cos x,
and tan x, where x is a real number.

Solution 7. Let 4ABC be a right triangle with hypotenuse BC and side lengths sin x, cos x, and tan x. There are
three cases to consider, depending on the length chosen for BC. If BC = tan x, then tan2 x = sin2 x + cos2 x =
1 ⇒ tan x√ = 1 ⇒ x = 45◦ . In this case, ABC would be an isosceles right triangle with legs of length
sin 45◦ = 22 , so [ABC] would be AB·AC
2 = 14 .

Because sin x, cos x, and tan x must all be positive,

sin x sin x
tan x = ≥ = sin x.
cos x 1
Thus it cannot be the case that BC = sin x, because the hypotenuse is the longest side of the triangle.
2
2
Finally, if BC = cos x, then [ABC] = AB·AC
2 = 2sin x 2 2
cos x and cos x = sin x + tan x. The latter equation can be
rewritten as

cos4 x = (sin2 x)(cos2 )x + sin2 x

cos4 x = (1 − cos2 x) cos2 x + (1 − cos2 x)
2 cos4 x = 1,

20
√
4 √
and so cos x = 1
√
4
2
= 8
2 , sin2 x = 1 − 2
2 . Thus
√
√
2 4
4
8 1 √
4
√
4

[ABC] = − = 2 2− 8 .
2 4 4

The last step is to compare the value above with 14 . To see that the value above is less than 14 , observe that
√ √ √ √
4 4 2 < 5 because (4 4 2)4 = 512 < 625 = 54 . Next, observe that 2 4 8 > 3 because (2 4 8)4 = 27 = 128 > 81 =
34 . Put these facts together to see that
1 √
4
√
4
1 √
4
√
4
1 1
2 2− 8 = 4 2 − 2 8 < (5 − 3) = .
4 8 8 4
√
4 √
4
2 8
Thus the answer is 2
− 4
.

Solution 8. Plug x = 0 into the given equation involving P (x + 2) to obtain P (2) = 115. Next, plug in x = 1 into
the given equation involving P (x + 1) to obtain P (2) = 1 + D + 54 + 37. Thus D = 115 − (1 + 54 + 37) = 23.
Then P (x) = P ((x − 1) + 1) = (x − 1)3 + 23(x − 1)2 + 54(x − 1) + 37 = x3 + 20x2 + 11x + 5, hence (A, B, C) =
(20, 11, 5). It can be verified that E = 103.

Solution 9. Let a, b, and c be the values rolled on the 8-, 12-, and 20-sided dice, respectively. Call a triple (a, b, c)
of rolls great if one roll equals the sum of the other two.

Imagine that the first two dice are rolled (so that a and b are determined), and then the 20-sided die is
rolled. Then the triple (a, b, c) is great if either c = a + b, c = a − b, or c = b − a. The first case is possible
1
for any pair of rolls (a, b), and the probability is 20 for each such pair. The second and third cases are only
possible if a > b or if b > a, respectively, but again, in each of these cases, the probability of obtaining the
1
proper value of c is 20 . So the total probability of rolling a great triple satisfying either c = a − b or c = b − a
1 8 1 11
is simply 20 · P (a > b or b > a). But P (a > b or b > a) = 1 − P (a = b) = 1 − 96 = 1 − 12 = 12 . (This value is
simply the probability that whatever number was rolled on the 8-sided die also comes up on the 12-sided die.)
1 11 23
So the total probability of rolling a great triple is 20 + 240 = 240 .

Problem 10. Compute the largest of the three prime divisors of 134 + 165 − 1722 .

Solution 10. Let N = 134 + 165 − 1722 . Notice that 165 = (24 )5 = 220 and 134 − 1722 = 1692 − 1722 =
(169 − 172)(169 + 172) = −3 · 341 = −1023 = 1 − 210 . Thus N = 220 − 210 + 1. Multiply by 1025 = 210 + 1 to
obtain

1025 · N = (220 − 210 + 1)(210 + 1)

= 230 + 1.

21
The right-hand side can be refactored via the identity 4u4 + 1 = (2u2 + 2u + 1)(2u2 − 2u + 1), using u = 27 :

1025 · N = (2 · 214 + 2 · 27 + 1)(2 · 214 − 2 · 27 + 1)

52 · 41 · N = (32768 + 256 + 1)(32768 − 256 + 1)
= 33025 · 32513
= (25 · 1321)(41 · 793)
N = 1321 · 793.

Searching for small prime divisors leads to 793 = 13 · 61, so 1321 is the largest prime factor of N .

22
9 Relay Problems
2
Relay 1-1 Compute the number of lattice points on the graph of y = (x − 153)(153−x ) .

√ √
Relay 1-2 Let T = T N Y W R. Let r and s be the zeros of x2 − 4T x + T 2 . Compute r+ s.

Relay 1-3 Let T = T N Y W R. Among all triples of integers (a, b, c) satisfying 2a + 4b = 8c , compute the least value
of a + b + c greater than T 2 .

Relay 2-1 Compute the number of ways to assign the integers {1, 2, . . . , 7} to each of the
hexagons in the figure to the right such that every pair of numbers in edge-
adjacent hexagons is relatively prime.

2T
Relay 2-2 Let T = T N Y W R. In right triangle ABC with hypotenuse BC, AB = T and BC = √
3
. A circle passes
through A and is tangent to BC at its midpoint. Compute the radius of the circle.

Relay 2-3 Let T = T N Y W R. Compute the number of ways to make $T from an unlimited supply of $10 bills, $5
bills, and $1 bills.

23
10 Relay Answers

Answer 1-1 27

√
Answer 1-2 9 2

Answer 1-3 169

Answer 2-1 72

Answer 2-2 24

Answer 2-3 9

24
11 Relay Solutions
2
Relay 1-1 Compute the number of lattice points on the graph of y = (x − 153)(153−x ) .

2
Solution 1-1 For an integer x, (x − 153)(153−x )
will be an integer if and only if one of the following three conditions
is satisfied:

(1) 153 − x2 ≥ 0,
(2) x − 153 = −1 or 1,
(3) x − 153 = 0 and 153 − x2 > 0.
For the first case, there are 25 solutions for −12 ≤ x ≤ 12. For the second case, there are two solutions
(x = 152 or 154). There are no solutions for the third case, as 153 − 1532 < 0. Hence there are 27 lattice
points on the graph.

√ √
Relay 1-2 Let T = T N Y W R. Let r and s be the zeros of x2 − 4T x + T 2 . Compute r+ s.

√ √ √
Solution 1-2 Let u = r + s. Then u2 = r + s + 2 rs. The sum of the zeros 2 2
√ of x − 4T x + T is 4T and their
2 2 2
product is T . Thus u = 4T + 2|T |. With T = 27, u = 6 × 27, so u = 9 2.

Relay 1-3 Let T = T N Y W R. Among all triples of integers (a, b, c) satisfying 2a + 4b = 8c , compute the least value
of a + b + c greater than T 2 .

Solution 1-3 Rewrite the equation with common bases to obtain 2a + 22b = 23c . The only case in which the sum
of two integer powers of 2 equals a third integer power of 2 occurs when the summands have equal exponents,
so a = 2b, 3c = a + 1, and a + b + c = a + a2 + a+1
3 =
11a+2
6 . For all three values to be integers, a must be an
√
even number of the form 3k − 1. As T = 9 2, T = 162 and 11a+2
2
6 ≥ 162 ⇒ 11a + 2 ≥ 972 ⇒ a ≥ 970 11 > 88.
The least even integer value a of the form 3k − 1 greater than 88 is 92, and the sum is 92 + 46 + 31 = 169.

Solution 2-1 No two even numbers can be adjacent, so the three even numbers must be in alternating hexagons on
the boundary. There are two ways to select alternating hexagons on the boundary, and six ways to assign the
even numbers to the three hexagons. The number 3 cannot be adjacent to 6, hence the 3 must be opposite the
6. Then the remaining odd numbers can be assigned in any way in the remaining three hexagons. Hence the
total number of arrangements is 2 × 6 × 6 = 72.

2T
Relay 2-2 Let T = T N Y W R. In right triangle ABC with hypotenuse BC, AB = T and BC = √
3
. A circle passes
through A and is tangent to BC at its midpoint. Compute the radius of the circle.

25
Solution 2-2 Note that 4ABC is a 30-60-90 right triangle, so AC = T
√
3
. Let point C 0 be obtained by reflecting
point C about AB so that 4CBC 0 is equilateral. Note that the circle which passes through A and which is
tangent to BC at its midpoint happens to be the incircle of 4CBC 0 ! The radius of the incircle is one-third
the altitude of 4CBC 0 , so the radius is T3 . As T = 72, the radius is 24.

Relay 2-3 Let T = T N Y W R. Compute the number of ways to make $T from an unlimited supply of $10 bills, $5
bills, and $1 bills.

Solution 2-3 Let T = 10t + 5f + n, where t, f , and n are non-negative integers representing the number of tens,
fives, and ones, respectively. Then 2t + f ≤ b T5 c, with the remaining total consisting of $1 bills. The number of
ways to make $T is equal to the number of lattice points (t, f ) contained in the triangle bounded by the lines
t = 0, f = 0, and 2t + f = b T5 c. When t = 0, there are b T5 c + 1 possible values of f . Incrementing t by one
reduces by two the number of values of f for which there is a solution. As T = 24, b T5 c + 1 = 5, and there are
5 + 3 + 1 = 9 solutions: (t, f, n) = (0, 0, 24), (0, 1, 19), (0, 2, 14), (0, 3, 9), (0, 4, 4), (1, 0, 14), (1, 1, 9), (1, 2, 4), and
(2, 0, 4).

26
12 Super Relay
1. Given that a, b, and c are positive integers with ab = 20 and bc = 16, compute the positive difference between
the maximum and minimum possible values of a + b + c.

2. Let T = T N Y W R. Mary circles a number in one of the first six rows of Pascal’s
Triangle, shown at right. Jim then circles another number — possibly having the
same value as Mary’s number — but Jim’s circle cannot coincide with Mary’s circle.
For example, if Mary circles the top “1”, then Jim can circle any number other than
the top “1”. The product of the two circled numbers is then computed. Compute the
number of possible distinct products that are greater than T .

3. Let T = T N Y W R. The diagram at right shows part of a 2×T checkerboard

pattern. A 1 × 3 tile is randomly placed on the board so that it completely
covers three unit squares. Compute the probability that two of the covered
squares are black.

4. Let T = T N Y W R. Leo rides his bike at 16T miles per hour and Carla and Rita both ride their bikes at 12
miles per hour. Leo and Carla start at the points (0, 0) and (60, 0), respectively, and they begin biking towards
each other (along the x-axis), where the units of both coordinate axes are in miles. At the same time, Rita
starts biking from the point (a, b), where a > 0, and she is riding along the line x = a, towards the x-axis.
Compute the value of a such that Leo, Carla, and Rita will meet at the same moment.

5. Let T = T N Y W R. In rectangle ARM L, diagrammed at right (not drawn to scale),

N is the midpoint of AR, N L = 40, and AN = T . Arc LP \ M is an arc of a circle
centered at N , P is the midpoint of arc LP
\ M , and arc LP
\ 0 M is the reflection of

LP M across LM . Given that P is the midpoint of arc LP M , compute N P 0 .

\ 0 \ 0

6. Let T = T N Y W R. Let p and q be distinct primes and for each positive integer n, let d(n) be the number
d(pT +2 q k )
of positive divisors of n. Compute the least possible positive integer k such that the quotient is an
d(pT +1 q 2 )
integer.

An ARML team of 15 students contains k boys and 15 − k girls. The value of k satisfies
7. Let T = T N Y W R.
the equation T +1
3 = 325k. Compute the probability that a randomly chosen student from the team is a boy.

15. A dartboard is made up of two concentric circles that have radii 20 and 16. A dart is thrown at random and
hits the board. Compute the probability that the dart lands in the circle of radius 16.

14. Let T = T N Y W R. Let A = 2K , R = A20 , M = R16 , and L = M T . Given that K and L are positive integers,
compute the least possible value of A.
13. Let T = T N Y W R. Compute 2logT 8 − 8logT 2 .
12. Let T = T N Y W R. Compute the value of k such that the following system of equations has a solution.

x+y = T
20x + 16y = 8
kx + 20y = 16.

11. Let T = T N Y W R. John has T distinct baseball cards and Benson has 2T distinct baseball cards, which include
the same T cards that John has. John randomly chooses two cards from Benson’s deck. The probability that
T
exactly one of the chosen cards is not in John’s deck can be expressed in the form . Compute K.
K

27
10. Let T = T N Y W R. In circle O, perpendicular chords AR and M L intersect at N . Given that AN = T ,
RN = 5, and M N = 25, compute the area of circle O.
9. Let T = T N Y W R, and let K = Tπ . Consider the sequence defined by a1 = 20, a2 = 16, and for n ≥ 3, an is

a
the units digit of an−1 + an−2 . Compute K .
10

8. Let t be the number you will receive from position 7 and let s be the number you will receive from position 9.
In 4ABC, point H lies on BC such that AH ⊥ BC. Given that the sides of 4ABC are integers, tan ∠B = t,
and sin ∠CAH = s, compute the least possible perimeter of 4ABC.

28
13 Super Relay Answers
1. 24
2. 6
1
3. 2

4. 24
5. 24
6. 25
8
7. 15
16
15. 25

14. 32
13. 0
12. 28

11. 55
10. 949π
3
9. 5

8. 48

29
14 Super Relay Solutions

Problem 1. Given that a, b, and c are positive integers with ab = 20 and bc = 16, compute the positive difference
between the maximum and minimum possible values of a + b + c.

Solution 1. Note that 5 must divide a, hence a must be either 5, 10, or 20. Thus the possible solutions (a, b, c)
are (5, 4, 4), (10, 2, 8), and (20, 1, 16). The first of these gives the minimum possible sum a + b + c of 13 while
the last of these gives the maximum possible sum a + b + c of 37. Thus the answer is 37 − 13 = 24.

Problem 2. Let T = T N Y W R. Mary circles a number in one of the first six rows of Pascal’s
Triangle, shown at right. Jim then circles another number — possibly having the
same value as Mary’s number — but Jim’s circle cannot coincide with Mary’s circle.
For example, if Mary circles the top “1”, then Jim can circle any number other than
the top “1”. The product of the two circled numbers is then computed. Compute the
number of possible distinct products that are greater than T .

Solution 2. Before receiving a value of T , it makes sense to list out all the possible products. There are 6 products
that are greater than T (= 24): 25 (5 × 5), 30 (5 × 6 = 3 × 10), 40 (4 × 10), 50 (5 × 10), 60 (6 × 10), and 100
(10 × 10).

Problem 3. Let T = T N Y W R. The diagram at right shows part of a 2×T checkerboard

pattern. A 1 × 3 tile is randomly placed on the board so that it completely
covers three unit squares. Compute the probability that two of the covered
squares are black.

Solution 3. Note that for any possible position of the 1 × 3 tile’s left boundary, there are exactly two placements
of the tile (in the first row or in the second row). Exactly one of these placements will cover two white squares
and one black square, and the other will cover two black squares and one white square. Because these options
are equally likely, the probability that the 1 × 3 tile covers two black squares is therefore 12 (independent of the
value of T ).

Problem 4. Let T = T N Y W R. Leo rides his bike at 16T miles per hour and Carla and Rita both ride their bikes
at 12 miles per hour. Leo and Carla start at the points (0, 0) and (60, 0), respectively, and they begin biking
towards each other (along the x-axis), where the units of both coordinate axes are in miles. At the same time,
Rita starts biking from the point (a, b), where a > 0, and she is riding along the line x = a, towards the x-axis.
Compute the value of a such that Leo, Carla, and Rita will meet at the same moment.

Solution 4. First note that because Carla and Rita ride at the same speed, Carla should ride a distance of b (i.e.,
Rita’s initial y-coordinate) and Leo should ride a distance of a = 60 − b in order for them to meet at the same
time. If k is the ratio of Leo’s speed to Carla’s and Rita’s speed, then 60−b
b
a
= 60−a 60r
= r, hence a = 1+r 240T
= 3+4T .
1
With T = 2 , it follows that a = 24.

Problem 5. Let T = T N Y W R. In rectangle ARM L, diagrammed at right (not drawn to scale),

N is the midpoint of AR, N L = 40, and AN = T . Arc LP \ M is an arc of a circle
centered at N , P is the midpoint of arc LP M , and arc LP 0 M is the reflection of
\ \
LP
\ M across LM . Given that P 0 is the midpoint of arc LP
\ 0 M , compute N P 0 .

30
Solution 5. Let N P intersect LM at Q. Because N P = N L = 40, it follows that P Q = 40 − AL. Because
P 0 Q = P Q, it follows that N P 0 = AL − (40 − AL) = 2 · AL − 40. With T = 24, it follows that 4N AL is
similar to a 3−4−5 triangle, hence AL = 32, and N P 0 = 24.

Problem 6. Let T = T N Y W R. Let p and q be distinct primes and for each positive integer n, let d(n) be the
d(pT +2 q k )
number of positive divisors of n. Compute the least possible positive integer k such that the quotient
d(pT +1 q 2 )
is an integer.

Solution 6. Note that d(pT +2 q k ) = (T + 3)(k + 1) because a positive divisor of pT +2 q k is of the form px q y , where
x and y are integers satisfying 0 ≤ x ≤ T + 2 and 0 ≤ y ≤ k. Similarly, d(pT +1 q 2 ) = 3(T + 2). Thus the
(T + 3)(k + 1)
quotient is . Note that T + 3 and T + 2 are relatively prime, so either k = T + 1 if 3 divides T + 3
3(T + 2)
or k = 3T + 5 if 3 does not divide T + 3. With T = 24, the former case applies, thus k = 25.

Problem 7. Let T = T N Y W R. An ARML team of 15 students contains k boys and 15 − k girls. The value of k
satisfies the equation T +1
3 = 325k. Compute the probability that a randomly chosen student from the team
is a boy.

26 26·25·24 8

Solution 7. With T = 25, 3 = 3! = 2600. Thus k = 2600/325 = 8, and the desired probability is 15 .

Problem 15. A dartboard is made up of two concentric circles that have radii 20 and 16. A dart is thrown at
random and hits the board. Compute the probability that the dart lands in the circle of radius 16.

Solution 15. The inner circle has radius 16 and therefore has area π · 162 . The outer circle has radius 20 and
therefore has area π · 202 . The probability that a dart lands in the inner circle is the ratio of these areas, so
16 2
2 2
the desired probability is π·16 = 45 = 16

π·202 = 20 25 .

Problem 14. Let T = T N Y W R. Let A = 2K , R = A20 , M = R16 , and L = M T . Given that K and L are
positive integers, compute the least possible value of A.

Solution 14. Note that R = (AK )20 = A20K , M = R16 = A320K , and L = M T = A320KT . If T is a positive
integer, then K = 1 suffices. On the other hand, if T is a rational number (say dc in lowest terms), but not a
d 16 25
positive integer, then K = gcd(320,d) . With T = 25 , K = gcd(320,25) = 5. Thus A = 25 = 32.

Problem 13. Let T = T N Y W R. Compute 2logT 8 − 8logT 2 .

3
Solution 13. Note that 2logT 8 = 2logT 2 = 23 logT 2 = 8logT 2 . Thus, so long as T is positive and T 6= 1 (to ensure
that the function logT is defined), the expression 2logT 8 − 8logT 2 equals 0. Because T = 32 satisfies these
constraints, the answer is 0.

Problem 12. Let T = T N Y W R. Compute the value of k such that the following system of equations has a
solution.
x+y = T
20x + 16y = 8
kx + 20y = 16.

31
Solution 12. Subtract sixteen times the first equation from the second equation to obtain 4x = 8 − 16T . Thus
x = 2 − 4T . Plug this into the first equation to get y = T − x = 5T − 2. Plug in these values for x and y
28 − 50T
into the third equation to obtain k(2 − 4T ) + 20(5T − 2) = 16 → k = . With T = 0, it follows that
1 − 2T
28
k= 1 = 28.

Problem 11. Let T = T N Y W R. John has T distinct baseball cards and Benson has 2T distinct baseball cards,
which include the same T cards that John has. John randomly chooses two cards from Benson’s deck. The
T
probability that exactly one of the chosen cards is not in John’s deck can be expressed in the form . Compute
K
K.

Solution 11. There are 2T

2 = T (2T − 1) possible selections of two cards that John can choose from Benson’s
deck. Of those, there will be T 2 pairs of cards, exactly one of which is not in John’s deck. Hence the desired
T2 T
probability is T (2T −1) = 2T −1 , hence K = 2T − 1. With T = 28, K = 55.

Problem 10. Let T = T N Y W R. In circle O, perpendicular chords AR and M L intersect at N . Given that
AN = T , RN = 5, and M N = 25, compute the area of circle O.

·RN
Solution 10. Note that by Power of a Point, LN = AN MN = T5 . With T = 55, LN = 11. Let P be the foot
of the perpendicular from O to AR and let Q be the foot of the perpendicular from O to M L. Note that
AP = 12 (55 + 5) = 30 and M Q = 12 (25 + 11) = 18. Because chords AR and M L are perpendicular, it follows
that OP = QN = M N − M Q = 7 and that OQ = P N = P R − N R = 30 − 5 = 25. Letting r be the radius of
circle O, it follows that r2 = 302 + 72 = 182 + 252 = 949, hence the answer is 949π.

Problem 9. Let T = T N Y W R, and let K = Tπ . Consider the sequence defined by a1 = 20, a2 = 16, and for

a
n ≥ 3, an is the units digit of an−1 + an−2 . Compute K .
10

Solution 9. Ignoring the tens digits of the first two terms, the sequence is

0, 6, 6, 2, 8, 0, 8, 8, 6, 4, 0, 4, 4, 8, 2, 0, 2, 2, 4, 6, 0, 6, 6, 2, . . . .

Note that this sequence is periodic with period 20. Thus when m is an integer, a20m = 6. With T = 949π,
K = 949, and a940 = 6. Counting 9 terms into the next block of 20 terms, conclude that a949 = 6, hence the
desired ratio is 53 .

Problem 8. Let t be the number you will receive from position 7 and let s be the number you will receive from
position 9. In 4ABC, point H lies on BC such that AH ⊥ BC. Given that the sides of 4ABC are integers,
tan ∠B = t, and sin ∠CAH = s, compute the least possible perimeter of 4ABC.

Solution
√ 8. Let BH = x and AC = y. Then AH = txtx and CH = sy.8 By the Pythagorean Theorem, AB =
3
x 1 + t2 and t2 x2 + s2 y 2 = y 2 , hence AC = y = √1−s 2
. With t = 15 and s = 5 , note that triangles AHB
and CHA are similar to 8−15−17 and 3−4−5 triangles, respectively. Setting x = 15 gives AH = 8, AB = 17,
CH = 6, and AC = 10. Thus the minimum perimeter of 4ABC is 17 + 10 + (15 + 6) = 48.

32
15 Tiebreaker Problems

Problem 1. Compute the least value of N such that there are exactly 43 ordered quadruples of positive integers
(a, b, c, d) satisfying N = a2 + b2 + c2 + d2 .

Problem 2. Regular octagon HEP T AGON has legs of length 8. Segments ET and P A intersect at X. Compute
the area of heptagon HEXAGON .

Problem 3. Compute the base-10 value of 1112 + 2223 + 3334 + · · · + 8889 .

33
16 Tiebreaker Answers

Answer 1. 100

√
Answer 2. 112 + 112 2

Answer 3. 2016

34
17 Tiebreaker Solutions

Problem 1. Compute the least value of N such that there are exactly 43 ordered quadruples of positive integers
(a, b, c, d) satisfying N = a2 + b2 + c2 + d2 .

Solution 1. To understand how to count the number of ordered quadruples that satisfy the given equation, it is
useful to consider a particular value of N and a particular solution to the resulting equation. So, for example,
with N = 18, one solution is (1, 2, 2, 3). Because of the symmetry of the expression a2 + b2 + c2 + d2 , any
4!
permutation of (1, 2, 2, 3) is also a solution, so there are 2! = 12 different quadruples corresponding to the same
values; call the set of all such quadruples a solution class. This observation suggests starting by identifying
different possible sizes of solution classes. Assuming that the initial values satisfy a ≤ b ≤ c ≤ d, the different
solution classes and their sizes are given in the following table.

Solution class Size

a<b<c<d 4! = 24
a=b<c<d
4!
a<b=c<d 2! = 12
a<b<c=d
4!
a=b<c=d 2!2! =6
a=b=c<d 4!
3! =4
a<b=c=d
a=b=c=d 1

The goal is then to identify the least value of N such that the sizes of the different solution classes sum
to 43. Note that 43 is odd, while all but one of the class sizes are even. So for some positive integer m,
N = m2 + m2 + m2 + m2 = 4m2 = (2m)2 . Thus N is not only even, it is the square of an even number.
Furthermore, note that all other solution classes except the case a = b < c = d have sizes divisible by four.
Because 42 is not divisible by four, there must also be a solution of the form N = a2 + a2 + c2 + c2 = 2a2 + 2c2 ,
which implies that half of N is a sum of squares of distinct integers.

These two insights rule out several small cases, because for m = 1, 2, 3, 4, N/2 = 2, 8, 18, 32 respectively,
and none of these is the sum of two distinct squares. On the other hand, m = 5 yields N = 100, and
N/2 = 50 = 12 + 72 , which is promising! In fact,

100 = 92 + 32 + 32 + 12
= 82 + 42 + 42 + 22
= 72 + 52 + 52 + 12 ,

yielding a total of 3 · 12 + 6 + 1 = 43 solutions. For N = 100, note that if the resulting equation had a solution
of the form a = b = c < d, then because 100 is not a multiple of 3, d cannot be a multiple of 3. Similarly, if a
solution of the form a < b = c = d existed, then a cannot be a multiple of 3. But because none of (100 − 12 )/3,
(100 − 22 )/3, (100 − 42 )/3, (100 − 72 )/3, or (100 − 82 )/3 is a perfect square, and because (100 − 52 )/3 = 25 = 52 ,
corresponds to the case 100 = 52 + 52 + 52 + 52 , which has already been accounted for, it follows that there are
no solutions of the form a = b = c < d or a < b = c = d.

Lastly, for N = 100, the equation has no solutions of the form a < b < c < d. To see this, note that a perfect
square is either congruent to 1 or 0 modulo 4, depending on its parity. Given that a2 + b2 + c2 + d2 = 100 ≡ 0
(mod 4), it follows that a, b, c and d must all have the same parity. If all are even, then the least possible value
of a2 + b2 + c2 + d2 with a < b < c < d would be 22 + 42 + 62 + 82 = 120 which is larger than 100. If all are odd,
then the least value is 12 + 32 + 52 + 72 = 84 6= 100. The second smallest value is 12 + 32 + 52 + 92 = 116. Thus
there are no solutions in the solution class a < b < c < d, and so the least value of N for which the equation
a2 + b2 + c2 + d2 = N has 43 solutions is 100.

35
Problem 2. Regular octagon HEP T AGON has legs of length 8. Segments ET and P A intersect at X. Compute
the area of heptagon HEXAGON .

Solution 2. Dissect heptagon HEXAGON into hexagon HEAGON and 4AXE. First compute the area of
hexagon HEAGON . Extend HE and ON to meet at Q, and extend ON and AG to meet at R. Note that
4HN Q and √ 4GOR are isosceles right triangles each√with hypotenuse of length 8. Thus each triangle has legs
of length 4 2 and therefore [HN Q] = [GOR] = 21 (4 2)2 = 16. Therefore

[HEAGON ] = [AEQR] − [HN Q] − [GOR]

= QR · QE − 16 − 16
= (QN + N O + OR)(QH + HE) − 32
√ √
= (8 + 8 2)(8 + 4 2) − 32
√
= 96 + 96 2.

Next, compute [AXE]. Draw in altitude XY from X to EA, so that

√
[AXE] = 12 EA · XY = (4 + 4 2) · XY.

To complete this computation, determine XY as follows. Draw HA. Because HA k ET and HE k XY ,

triangles HEA and XY E are similar. Specifically, EY = 21 EA, so XY = 12 HE = 4. Thus
√ √
[AXE] = (4 + 4 2) · 4 = (16 + 16 2).

Finally, combine these computations to obtain

[HEXAGON ] = [HEAGON ] + [AXE]

√ √
= (96 + 96 2) + (16 + 16 2)
√
= 112 + 112 2.

Problem 3. Compute the base-10 value of 1112 + 2223 + 3334 + · · · + 8889 .

Solution 3. Note that m m mm+1 = (m + 1)3 − 1. So the desired sum S is

S = (23 − 1) + · · · + (93 − 1)
= (23 + · · · + 93 ) − 8
= (13 + 23 + · · · + 93 ) − 8 − 13 .
2
n(n+1)
Use the identity 13 + 23 + · · · + n3 = (1 + 2 + · · · + n)2 = 2 to obtain
2
9 · 10
S= − 9 = 452 − 9 = 2016.
2

36
ARML Competition 2017

Paul J. Karafiol, Head Writer

George Reuter, Associate Head Writer
Andy Soffer, Associate Head Writer
Chris Jeuell, Lead Editor
Paul Dreyer
Edward Early
Zuming Feng
Benji Fisher
Zachary Franco
Silas Johnson
Winston Luo
Andy Niedermaier
Graham Rosby
Eric Wepsic

June 2–3, 2017

Head Author’s Note
A little over twenty years ago, my friend and mentor Don Barry suggested that I read through a draft of the coming
year’s ARML contest over the winter holiday; he wanted my suggestions and feedback. I agreed, and one snowy
afternoon I pulled a chair up in front of the fireplace and opened the packet he had mailed me, eager to hunker down
for a few hours of challenging mathematics. You can imagine my surprise when I discovered that each “problem”
had three or four candidates, so that instead of doing one ARML contest, I was doing four—and instead of a few
hours of math, I was looking at somewhere between ten and twelve!

As I began working the problems, my initial dismay gave way to delight. While I couldn’t work them all that
first time through, I was immediately hooked on the challenge of creating and solving questions whose solution
demanded creativity, planning, and strategy, not just brute-force calculation. I couldn’t then conceive of being able
to write such questions. Three years later, however—exactly seventeen years ago—inspiration struck, and my first
Power Question, “Power of Association,” was born. I wrote several more, and a few questions to fill in gaps in the
rest of the contest, and in 2008, Don asked me to succeed him as head author. I was honored, terrified, and grateful.

Ten years and ten contests later, I’m still honored and grateful, although marginally less terrified. I’ve learned
that writing a contest of exceptional quality requires a committee of exceptional quality, and I’ve been fortunate to
enjoy working with the best. Our annual meeting includes math late into the night, wonderful meals and fellowship,
and jokes that only the nerdiest would get—and some for which you really “had to be there.” A not-insubstantial
added benefit of expanding that committee has been embracing several former students as colleagues and friends.

This year marks my last as head author, although I hope and plan to continue contributing problems for many years
to come. We couldn’t have come this far without many people’s work: among the authors, I want to particularly
thank Lead Editor and perpetual Super Relay author Chris Jeuell for his unwavering eye for errors (mathematical,
typographical, and LATEXnical), unceasing patience, and unlimited generosity of spirit and good humor. (Evidence
of this last is evident in this year’s Super Relay, which was a wonderful parting gift.) Micah Fogel, Paul Zeitz, and
the other PQ graders have provided valuable insights not just on individual Power Questions but on what makes a
Power Question challenging, exciting, and feasible to grade. Don Barry and the late Bryan Sullivan trusted in me
and my authorship and provided needed but always-deferential suggestions and feedback; Don, you’re right that it’s
impossible to write a relay that’s too easy. And a huge thanks goes out to all the coaches, teachers, parents, and
mentors who make the contest possible for thousands of young people every year. Seeing the throngs of students
taking the contest every year makes the whole thing worthwhile.

On a personal note, I want to thank my wife, Allison, and my children, Ari, Jonah, and Helen, for putting up
with the many hours—often on family “vacation”—that ARML has taken. I, and they, will probably forever asso-
ciate our spring break trips with late nights spent on last-minute finishing touches.

Next year’s contest will take place under the leadership of George Reuter, who has long authored the NYSML
contest. His accession returns us to our roots, in that ARML itself was an outgrowth of NYSML in the early 1970’s.
George joined our committee three years ago, and with his background in high school teaching, shares my perspective
on what actual high school students can do, as well as a passion for challenging, unusual mathematics. I’m excited
for his leadership, and grateful, again; but not at all terrified. Have a wonderful ride!

Paul J. Karafiol
April, 2017

1
1 Team Problems
√
Problem 1. Compute the number of ordered triples of positive integers (a, b, c) such that ab + c! = 28.

Problem 2. Chris the frog begins on a number line at 0. Chris takes jumps of lengths 1, 2, 3, . . . , 2017, in that
order. If Chris’s current location is an even integer, he jumps in the positive direction; P
otherwise, he jumps in
the negative direction. Let P (n) denote Chris’s location after the nth jump. Compute 2017 j=1 P (j).

Problem 3. The diagram below shows arc ABC, ˘ which has a measure of 210◦ . Points X and Y lie on the arc so
◦ ∼
that m∠AXB√ = 90 and 4ABX = 4Y CB. Given that AX = 8, the value of [ABC] can be expressed in the
form a + b c, where a, b, c are integers and c is not divisible by any perfect square greater than 1. Compute
the ordered triple (a, b, c).
A
Y

C
B X

Problem 4. Ari repeatedly rolls a standard, fair, six-sided die. Let R(n) be the nth number rolled, and let
Q(n) = R(1) · R(2) · . . . · R(n). Compute the probability that there exists an n such that Q(n) = 100 and for
all m < n, Q(m) is not a perfect square.

Problem 5. Given that C, A, T , F , I, S, and H are digits, not necessarily distinct, and that
3 · C A T · F I S H = C A T F I S H,
compute the greatest possible value of C A T F I S H.

1 Pn−1
Problem 6. Let {an } be a sequence with a0 = 1, and for all n > 0, an = 2 i=0 ai . Compute the greatest value
of n for which an < 2017.

Problem 7. On July 17, 2017, the nation of Armlandia will turn n2 years old and the nation of Nysmlistan will
turn n years old. The next four anniversaries for which Nysmlistan’s age divides Armlandia’s age will occur,
in order, on July 17 in the years 2027, 2032, 2038, and M . Compute M .

Problem 8. Ellipse E has center O, major axis of length 10, and minor axis of length 4. Ellipse E 0 is obtained by
rotating E counterclockwise about O by 60◦ . Proceeding clockwise around the perimeter of E, the intersection
points of E and E 0 are labeled A, R, M , L. Compute [AOL].

Problem 9. Let E(n) denote the least integer strictly greater than n whose base-10 representation contains only
even digits, and let O(n) denote the least integer strictly greater than n whose base-10 representation contains
only odd digits. Compute the least positive integer N for which E(N ) + O(N ) is a multiple of 2017.

Problem 10. Compute the number of tilings of a 4 × 7 rectangle using only 1 × 1, 2 × 2, 3 × 3, and 4 × 4 tiles.

2
2 Answers to Team Problems

Answer 1. 9

Answer 2. 1

Answer 3. (64, −32, 3)

3
Answer 4. (or 0.012)
250

Answer 5. 6673335

Answer 6. 21

Answer 7. 2062

√
200 2923
Answer 8.
2923

Answer 9. 64026

Answer 10. 1029

3
3 Solutions to Team Problems
√
Problem 1. Compute the number of ordered triples of positive integers (a, b, c) such that ab + c! = 28.

Solution 1. Squaring both sides of the given equation gives ab + c! = 784. Accordingly, only values of c up to 6
need be considered, as 7! > 784.
c 784 − c! prime factorization of 784 − c!
1 783 33 · 29
2 782 2 · 17 · 23
3 778 2 · 389
4 760 23 · 5 · 19
5 664 23 · 83
6 64 26
For 1 ≤ c ≤ 5, the only ordered triple satisfying the equation is (784 − c!, 1, c). For c = 6, there are four triples:
(2, 6, 6), (4, 3, 6), (8, 2, 6), and (64, 1, 6), for a total of 9 ordered triples.

Solution 2. List where Chris has landed after his first few jumps.
Jump Before Direction After
1 0 + 1
2 1 − −1
3 −1 − −4
4 −4 + 0

In fact, this pattern continues: Chris lands back at the origin after every fourth jump. Suppose that P (4n) = 0.
Then the next four jumps are as follows.
Jump Before Direction After
4n + 1 0 + 4n + 1
4n + 2 4n + 1 − −1
4n + 3 −1 − −4n − 4
4n + 4 −4n − 4 + 0
This inductive argument shows that Chris lands back at the origin after every four jumps. Moreover, the sum
of the four values of P (j) is constant:
P (4n + 1) + P (4n + 2) + P (4n + 3) + P (4n + 4)
= (4n + 1) + (−1) + (−4n − 4) + 0
= −4.
Compute P (1) + · · · + P (2017) by grouping terms in groups of four:
2017 503
!
X X
P (j) = P (4k + 1) + P (4k + 2) + P (4k + 3) + P (4k + 4) + P (2017)
j=1 k=0
503
!
X
= −4 + P (2017)
k=0
= −2016 + P (2017).

4
Because 2016 is divisible by 4, after jump 2016, Chris will be at the origin. After jump 2017, Chris will be at
2017, so P (2017) = 2017, and the sum is
2017
X
P (j) = −2016 + 2017 = 1.
j=1

A
Y

C
B X

Solution 3. Because ∠AXB is a right angle, it follows that AB is a diameter of the circle containing points A, B,
and C. Thus ∠ACB is also right angle. Also ∠BAC ∼ = ∠BY C because both angles subtend BC. ˜ Because
∠BY C ∼ = ∠XAB, conclude that 4ABC ∼ = 4ABX. Next, note that m∠XAB = m∠BY C = 12 · mBC ˜ =
1 ◦ ◦ ◦ 1 1 ◦ ◦
2 (210 − 180 ) = 15 . Thus [ABC] = [ABX] = 2 · AX · BX = 2 · 8 · (8 tan 15 ) = 32 tan 15 . The value
◦ x sin x ◦
of tan 15 can be computed by using the tangent half-angle identity: tan( 2 ) = 1+cos x with x = 30 , yielding
sin 30◦ 1/2 √ √ √
tan 15◦ = 1+cos 30◦ = 1+ 3/2 = 2 − 3. Thus [ABC] = 32(2 − 3) = 64 − 32 3, and the desired ordered triple
√

(a, b, c) is (64, −32, 3).

Note: The value of tan 15◦ can also be computed using the identity tan(A − B) = tan A−tan B
1+tan A tan B , with A = 45◦
and B = 30◦ .

Alternate Solution: Proceed as in the previous solution to conclude that 4ABC ∼ = 4ABX. Then AB =
◦
AX
cos 15 ◦ and [ABC] = 1
2 ·AX · AX
◦ ·sin
cos 15 √ √ 15 . Use the subtraction identities to compute cos 15◦ = cos(45◦ −30◦ ) =
√ √
cos 45◦ cos 30◦ +sin 45◦ sin 30◦ = 6+ 2
and sin 15 ◦
= sin(45◦
−30 ◦
) = sin 45 ◦
cos 30◦
−cos 45 ◦
sin 30◦
= 6− 2
.
√ √ √ √4 √ √ √ √ 4
◦ ( 6− 2)( 6− 2)
sin 15 6−√2 6−2 12+2
Dividing, cos 15◦ = 6+ 2 = ( 6+ 2)( 6− 2) =
√ √ √ √ √
4 = 2− 3. Substituting into the expression for [ABC]
1 2
Ä √ ä √
yields 2 · 8 · 2 − 3 = 64 − 32 3, as in the previous solution.

Solution 4. In order for the conditions in the problem to be met, the first roll must be either 2 or 5. After the
first roll, rolls of 1 can be ignored, because they are essentially equivalent to simply re-rolling; there are five
other possibilities. Because of the requirement that no partial product before 100 be a perfect square, there
are only five sequences of non-1 rolls that result in success: 2-5-2-5, 2-5-5-2, 5-4-5, 5-2-5-2, and 5-2-2-5. The
3
probability of obtaining each sequence of length four is 61 · 15 = 750 1
, while the probability of obtaining the

5
2
sequence of length three is 61 · 51 = 750
5
. Because there are four sequences of length four and one sequence of
1 5 9 3
length three, the desired probability is 4 · 750 + 1 · 750 = 750 = 250 .

Problem 5. Given that C, A, T , F , I, S, and H are digits, not necessarily distinct, and that

3 · C A T · F I S H = C A T F I S H,

compute the greatest possible value of C A T F I S H.

Solution 5. First rewrite the given equation as

3 · C A T · F I S H = 10000 · C A T + F I S H

which gives
10000 1
3= + .
F I SH C AT

Therefore, in order to compute the greatest possible value of C A T , and hence of C A T F I S H, F 10000
I S H
must be as close as possible to 3 without going over, which means F I S H must be as close to 3333 as possible.
Trying F I S H = 3334 gives C A T = 1667, which is impossible. Trying F I S H = 3335 gives C A T = 667,
which is the greatest possible value of C A T .

Thus the greatest possible value for C A T F I S H is 6673335.

1 Pn−1
Problem 6. Let {an } be a sequence with a0 = 1, and for all n > 0, an = 2 i=0 ai . Compute the greatest value
of n for which an < 2017.

Solution 6. Examine the first few terms in the sequence until a pattern emerges.

1 30
a1 = = 1
2 2
31
Å ã
1 1 3
a2 = 1+ = = 2
2 2 4 2
32
Å ã
1 1 3 9
a3 = 1+ + = = 3
2 2 4 8 2
n−1
The following argument shows that for n ≥ 1, an = 3 2n . The base case is shown above. For the induction
k−1
hypothesis, assume that ak = 3 2k for some k ≥ 1; then use the geometric series sum formula with r = 23 to
compute ak+1 :
1
ak+1 = (a0 + a1 + · · · + ak )
2Ç å
1 30 31 32 3k−1
= 1 + 1 + 2 + 3 + ··· + k
2 2 2 2 2
k k
!! !!
1 1 − 32k 1 1 − 32k
Ç å
1 1 1 3k 3k
= 1+ = 1 − = = .
2 2 1 − 32 2 2 1
2
2 2k 2k+1

3n−1 1 3 n−1

Thus for n > 0, an = 2n = 2 · 2 .

Consequently, the number which must be computed is the greatest value of n for which 1.5n−1 < 2·2017 = 4034,
i.e., 1.5n < 6051. Note that 1.54 = 2.252 = 5.0625 > 5. Therefore 1.520 = (1.54 )5 > 55 = 3125, and so

6
1.522 > 3125 · 2.25 which is clearly larger than 6051. Thus the desired value of n is no more than 21. Moreover,
5
because 1.54 is only slightly larger than 5 and 1.520 = 1.54 , one should expect that 1.520 will be only slightly
larger than 55 = 3125. This argument suggests that 3125 · 1.5 = 4627.5 is a reasonable approximation for 1.521 ,
making 21 a reasonable guess for the greatest n with 1.5n < 6051.

To verify this guess, it suffices to verify that 1.520 < 4034. Expanding 1.520 using the binomial theorem
gives
1.520 = (1.54 )5
= (5 + 0.0625)5
Ç å Ç å Ç å
5 5 5 4 5
= ·5 + · 5 · 0.0625 + · 53 · 0.06252 +
0 1 2
Ç å Ç å Ç å
5 2 3 5 4 5
· 5 · 0.0625 + · 5 · 0.0625 + · 0.06255 .
3 4 5

The largest binomial coefficient is 52 = 53 = 10, but 0.0625 < 0.1, so when positive powers of 0.0625 are

multiplied by binomial coefficients, the result is never more than 1. Replacing the binomial coefficients and
powers of 0.0625 with 1 yields an upper bound of
56 − 1
1.520 < 55 + 54 + 53 + 52 + 51 + 50 = = 3906 < 4034.
5−1
Thus 1.521 < 4034 · 1.5 = 6051, and so 21 is indeed the correct answer.

Solution 7. First note that the age of Nysmlistan in each of 2027, 2032, and 2038 is n + 10, n + 15, and n + 21,
respectively. Similarly, the age of Armlandia in those years is n2 + 10, n2 + 15, and n2 + 21. Next suppose that
n + k | n2 + k, for some integers n and k. Then using polynomial long division,
k(k + 1)
n2 + k = (n − k)(n + k) +
n+k
which implies that n + k must divide k(k + 1).

Therefore n + 10 must divide 110, n + 15 must divide 240, and n + 21 must divide 462. Factoring each
value shows that only two values of n satisfy these three conditions: n = 1 and n = 45. The condition that
2027 is the first year after 2017 in which Nysmlistan’s age divides Armlandia’s age precludes n = 1, so n = 45.

Let m = M − 2017, so that Nysmlistan’s age in year M is m + n. Because M > 2038, m + n > 21. Then,
following the above, n + m | (n2 + m), so n + m divides m(m + 1). Substituting n = 45 gives m + 45 | (m2 + m),
and using polynomial long division once again yields
1980
m2 + m = m − 44 + .
m + 45
The first factors of 1980 greater than 45 are 55, 60, 66, and 90, corresponding to m = 10, m = 15, m = 21,
and m = 45, respectively. Hence m = 45 and M = 2062.

7
Solution 8. Place the ellipse and its images on the coordinate plane. While ordinarily it is more convenient to place
ellipses so that their major and minor axes are parallel to (or coincide with) the x- and y-axes, in this situation,
it’s more strategic to place the ellipses so that their intersection points lie on the axes. Accordingly, place E on
the xy-plane so that its center is at the origin and its major axis is rotated 30◦ clockwise relative to the x-axis.
Then E 0 will also have its center at the origin, but will have its major axis rotated 30◦ counterclockwise relative
to the x-axis. Because this diagram is symmetric about both x− and y− axes, the points of intersection A, R,
M , and L all lie on the coordinate axes, as shown in the diagram below.

4
E´
L
2

M A

-6 -4 -2 0 2 4 6

-2
R E
-4

-6

Without loss of generality, let A = (a, 0) be the point of intersection on the positive x-axis. Then L = (0, `) is
the point of intersection on the positive y-axis, and [AOL] = a` 2 .

To compute the lengths a and `, rotate E by 30◦ counterclockwise to a new ellipse E 00 so that its major
axis coincides with the x-axis. The points A and L will also be rotated by 30◦ to new points A00 and L00 ,
respectively. Because 4AOL ∼ = 4A00 OL00 , it follows that [AOL] = [A00 OL00 ] = 21 A00 O · L00 O. Now A00 is the
intersection of E and a line through the origin making a 30◦ angle with the x-axis, as shown in the diagram
00

below.

L´´
2 A´´
E´´

-6 -4 -2 0 2 4 6

-2

-4

-6

x2 y2
The equation for E 00 is + = 1, or 4x2 + 25y 2 = 100. Because tan 30◦ = √1 ,
3
the equation for the line is
25 4

8
y= √1 x. Substitute to obtain
3
Å ã2
1
4x2 + 25 √ x = 100
3
12x + 25x2 = 300
2

300
x2 =
37…
3
x = 10 .
37
»
3
√ 10 37
Because a = A00 O, it suffices to compute A00 O using trigonometry: x
A00 O = cos 30◦ = 3
2 , so A00 O = √ =
3
2
√20 .
37
√
Similarly, L00 is the intersection of E 00 with the line y = − 3 x. Solve for the intersection point to obtain
Ä √ ä2
4x2 + 25 − 3x = 100
79x2 = 100
10
x = −√ .
79

Because ` = L00 O and = cos 120◦ = − 21 , conclude that ` =

x
L00 O
√20 .
79
Ä
√20
äÄ
√20
ä √
00 00 a` 37 79 200 2923
Then [AOL] = [A OL ] = = = .
2 2 2923

Solution 9. Let S(n) = E(n) + O(n). In general, either S(n) consists of only odd digits, or S(n) has initial digit 2,
and then only odd digits. To prove this observation, notice that one of E(n) and O(n) will consist of an initial
digit, and then either all 0’s or all 1’s, respectively: if d is the first (leftmost) digit of n and d is odd, then the
next larger number with all even digits must have first digit d + 1 (or 2 if d = 9) followed by all 0’s, while if
d is even, then the next larger number with all odd digits must have first digit d + 1 followed by all 1’s. To
compute the sum of E(n) and O(n), consider the sum of each pair of corresponding digits separately, beginning
with the digits immediately to the right of their leftmost digits. If d is odd, this process is equivalent to adding
0’s to O(n)’s digits. If d is even, this process is equivalent to adding 1’s to E(n)’s digits, and none of E(n)’s
digits is a 9. In neither case is there any need to carry. Thus when the E(n) and O(n) are added, there will
be no carrying, except possibly in the initial digit. If E(n) and O(n) have the same number of digits, then the
initial digit of S(n) will be odd. If E(n) and O(n) do not have the same number of digits, then O(n) must have
initial digit 9 and E(n) will have one more digit than O(n), with an initial 2 and 0’s everywhere else. Either
way, S(n) will fit the above form. This argument proves the lemma.

Therefore, to determine possible values of S(N ), look for multiples of 2017 that either consist of only odd
digits or have an initial 2 and then all odd digits. Consider the thousands place of the first few multiples of
2017, which are 2017, 4034, 6051, . . . . The thousands place will continue to be even, until enough multiples of
17 are accumulated to roll over into the thousands place. As long as the thousands place is even, the multiple
of 2017 cannot be a possible value of S(N ), because all the digits will not be even, and the thousands place is
not the initial digit of the multiple (except for 2017, which does not work because 0 is not odd). Therefore,
noting that 58 < 1000/17 < 59, skip directly to 59 · 2017 = 119003. In order to force the hundreds place to be

9
odd, note that 5 < 97/17 < 6, so skip ahead by 6 · 2017 to 65 · 2017 = 131105. This value still doesn’t work, so
noting that S(N ) must be odd, consider S(N ) = 67 · 2017 = 135139.

Note that S(N ) does not begin with a 2, so E(N ) and O(N ) have the same number of digits, and the first
digits of E(N ) and O(N ) must differ by 1, so E(N ) must begin with a 6 and O(N ) must equal 71111. Thus
E(N ) = 135139 − 71111 = 64028, and hence N = min(64028, 71111) − 2 = 64026.

Problem 10. Compute the number of tilings of a 4 × 7 rectangle using only 1 × 1, 2 × 2, 3 × 3, and 4 × 4 tiles.

Solution 10. Let T (n) be the number of ways to tile a 4 × n rectangle with the given types of tiles, with T (0) = 1.
Because there is only one way to tile a 4 × 1 rectangle, i.e., with 1 × 1 squares, T (1) = 1.

Use the term fault line to describe a vertical line from top to bottom that does not intersect any tiles and that
is not either the left or the right border of the figure. For example, if a 4 × 4 region is tiled with two 2 × 2
tiles on the left, and then eight 1 × 1 tiles, as shown in the left figure below, then the tiling has two fault lines,
while if the two 2 × 2 tiles are next to each other at the top of the region, then the tiling has only one fault
line, as shown in the right figure below.

Two fault lines One fault line

Under this definition, it is possible for a tiling to have no fault lines; for example, if the region is completely
filled by a single square, or if 2 × 2 squares are staggered so that the right half of one is directly above or below
the left half of the other. Thus, in general, a 4 × n rectangle can have between 0 and n − 1 fault lines, inclusive.
The number of tilings can be counted by conditioning on the location of the leftmost fault line.

For ease of notation, label the vertical lines (whether or not they are fault lines) one unit to the right of
the leftmost edge, two units to the right of the leftmost edge, etc. as `1 , `2 , etc. respectively, as shown below.
In that example, the leftmost fault line is at `2 .

1 2 3 4

Let F (n) represent the number of ways to tile a 4 × n rectangle with no fault lines. Note that F (0) = T (0) = 1
and F (1) = T (1) = 1.

A 4 × 2 rectangle can be tiled with no fault line in four ways (two 2 × 2 squares or one 2 × 2 square and
four 1 × 1 squares), so F (2) = 4. If there is one fault line, it is at `1 , and divides the rectangle into two 4 × 1
strips, each of which can be tiled in F (1) = T (1) ways, for F (1)·T (1) = 1 tiling. So T (2) = F (2)+F (1)·T (1) = 5.

10
A 4 × 3 rectangle can be tiled with no fault line in four ways: either there is a 3 × 3 square (and the rest
of the area is filled in with 1 × 1 squares), which can happen in two ways, or there are two 2 × 2 squares, one
above and to either the left or right of the other, with the other spaces filled in with 1 × 1 squares.

Hence F (3) = 4. If the leftmost fault line is at `1 , then there are F (1) ways to tile the left side and T (2) ways
to tile the right side. If the leftmost fault line is at `2 , then there are F (2) ways to tile the left side and T (1)
ways to tile the right side. Hence T (3) = F (3) + F (1) · T (2) + F (2) · T (1) = 4 + 1 · 5 + 4 · 1 = 13.

A 4 × 4 rectangle can be tiled with no fault line in three ways: either using a single 4 × 4 square, or by
alternating 2 × 2 squares in the top and bottom row (and filling in with four 1 × 1 squares on the ends) as in
the 4 × 3 case, which can happen in two ways. So F (4) = 3. Otherwise, with the first fault line occurring at
`1 , `2 , or `3 , there are F (1) · T (3), F (2) · T (2), and F (3) · T (1) possible tilings, respectively, so that

T (4) = F (4) + F (1) · T (3) + F (2) · T (2) + F (3) · T (1)

= 3 + 1 · 13 + 4 · 5 + 4 · 1
=⇒ T (4) = 40.

The foregoing analysis shows that, in general,

T (n) = F (n) + F (1) · T (n − 1) + F (2) · T (n − 2) + · · · + F (n − 1) · T (1).

Indeed, the proof is trivial, because if there is no fault line, there are F (n) tilings, and if the leftmost fault line
is at `i , then there are F (i) ways of tiling the region to the left of the fault line without any fault lines, and
there are T (n − i) ways of tiling the region to the right of the fault line (with or without fault lines). Also, for
n > 4, F (n) = 2, because the only way to tile such a region without fault lines is to alternate 2 × 2 squares in
the top and bottom rows. So a simpler version of the formula above for n > 4 is simply

T (n) = 1 · T (n − 1) + 4 · T (n − 2) + 4 · T (n − 3) + 3 · T (n − 4) + 2 · T (n − 5) + · · · + 2 · T (0).

Using this formula, construct a table of values, as shown below.

n 0 1 2 3 4 5 6 7
T (n) 1 1 5 13 40 117 348 1029

Alternate Solution: Let T (n) be the number of ways to tile an 4 × n rectangle with square integer-length
tiles, and let S(n, k) be the number of ways to tile the same rectangle conditional on the largest tile touching
the left-hand border being of size k × k. Then T (n) = S(n, 1) + S(n, 2) + S(n, 3) + S(n, 4).

Because there is only one way to arrange a strip of 1 × 1 squares along the left side, S(n, 1) = T (n − 1).
Because there is only one way to arrange a single 4 × 4 square on the left side, S(n, 4) = T (n − 4). As the
diagram below shows, there are two ways to arrange a 3 × 3 square and three 1 × 1 squares so that the 3 × 3
square touches the left side, so S(n, 3) = 2 · T (n − 3).

11
The remaining term to address is the S(n, 2) term. There are three ways to fit one 2 × 2 tile and four 1 × 1
tiles in the leftmost 4 × 2 subregion. There is also one way to fit two 2 × 2 tiles in that subregion, as illustrated
below.

However, it is also possible for a series of two or more 2 × 2 squares to alternate across an 4 × m range, where
m can be any integer larger than 2. Note that for any such m, this alternating set of 2 × 2 squares can occur
in two ways, depending on whether the leftmost 2 × 2 square is at the top or the bottom. The diagram below
shows both examples with m = 4.

Thus S(n, 2) = 4 · T (n − 2) + 2 · T (n − m).

P
m≥3

Combine the above results to obtain the following recursion:

X
T (n) = T (n − 1) + 4 · T (n − 2) + 2 · T (n − 3) + T (n − 4) + 2 · T (n − m).
m≥3

Use the recursion to build the table below.

n 0 1 2 3 4 5 6 7
T (n) 1 1 5 13 40 117 348 1029

Hence there are 1029 tilings. They are illustrated on the following page.

12
13
4 Power Question 2017: The Paral-Elle Universe
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part.
To receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “com-
pute,” you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show
your work or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a
problem says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered
items may be used in solutions to later numbered items, but not vice versa. Pages submitted for credit should be
NUMBERED IN CONSECUTIVE ORDER AT THE TOP OF EACH PAGE in what your team considers to be
proper sequential order. PLEASE WRITE ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM
NUMBER (not the team name) on the cover sheet used as the first page of the papers submitted. Do not identify
the team in any other way.

Elle is a student at Springfield High School. One day, she’s escorted from her math class by a friendly space alien,
who takes her to a parallel universe to learn math the way they do math, which is different than the way we do
math. In the parallel universe, students are taught two operations: ⊕ and . Elle notices that

a ⊕ b = min(a, b) and a b=a+b

for all real numbers a and b. Elle decides to investigate the properties of these two new operations.

Throughout this Power Question, a ⊕ b and a b will always denote parallel-universe addition and multiplication,
respectively, while a + b and a · b will denote addition and multiplication in the usual sense. The standard order of
operations ( first, then ⊕) will apply, although parentheses are sometimes included for clarification. All variables
refer to real numbers unless otherwise specified.

1. a. Determine whether there is a real number O such that O ⊕ x = x for all x. [2 pts]
b. Show that there is a real number I such that I x = x for all x. [2 pts]
2. Show that distributes over ⊕; that is, a (b ⊕ c) = (a b) ⊕ (a c) for all a, b, c. [3 pts]
3. Find all solutions to each of the following equations:
a. (x x) ⊕ (2 x) ⊕ 1 = x ⊕ 2; [3 pts]
b. (x x) ⊕ (−1) = (x ⊕ 1) (x ⊕ (−1)). [3 pts]

Next, Elle investigates the properties of exponents and polynomials. She uses the notation
b
a =a a a ··· a,
| {z }
b copies of a

0
where a is a real number and b is a positive integer. Furthermore, by definition, a = 0. Elle wonders if exponenti-
ation works in the parallel universe the way it works in ours.
4. Show that for any real numbers x and y, and any positive integer n:
n n n
a. (x y) =x y ; [2 pts]
n n n
b. (x ⊕ y) =x ⊕y ; [2 pts]
2 2 2 2 2
c. (x ⊕ y) =x ⊕y = (x ) ⊕ (x y) ⊕ (y ). [3 pts]
2
5. a. Show that for any a, there is exactly √
one solution to the equation x = a. This is called the parallel
square root of a and is denoted by ◦ a. [2 pts]
b. Show that for any a, there is exactly one solution to the equation a x = I, where I is defined in
Problem 1b. This is called the parallel reciprocal of a and is denoted by b
a. [2 pts]
√
◦ √
◦
c. Show that for any a, b a=d a. [3 pts]

14
In the parallel universe, a parallel-universe polynomial (PUP) of degree n is a finite “sum” (i.e., parallel-universe
addition) of n + 1 parallel products whose exponents decrease from n to 0, inclusive:
n (n−1) 1 0
f (x) = (an x ) ⊕ (an−1 x ) ⊕ · · · ⊕ (a1 x ) ⊕ (a0 x ),
1 0
where each ai is a real number. As a convenient shorthand, x will simply be written as x and a0 x will simply
be written as a0 . For example:
• f (x) = (5 x) ⊕ 3 is a degree-1 PUP, but not a degree-2 PUP.
2
• f (x) = (1 x ) ⊕ 1 is not a PUP at all, because the x term is missing.
2
• f (x) = (0 x ) ⊕ (0 x) ⊕ 0 is a degree-2 PUP.
2 2
• f (x) = (x ) ⊕ x ⊕ 1 is not a PUP; it should instead be written as (0 x ) ⊕ (0 x) ⊕ 1.
Elle now examines the properties of degree-2 (quadratic) PUPs. In the parallel universe, a root of a PUP f is a real
number x such that f (x) = 0.

6. In our universe, a quadratic equation has either zero, one, or two distinct real roots.
a. Find a quadratic PUP that has exactly one root. [2 pts]
b. Could a quadratic PUP have zero roots? Justify your answer. [2 pts]
c. Could a quadratic PUP have infinitely many distinct roots? Justify your answer. [2 pts]
d. Suppose that f (x) is a PUP of positive degree. Prove that if f has more than one root, then it has
infinitely many distinct roots. [3 pts]
7. Suppose f (x) = (a x 2 )⊕(b x)⊕c is a quadratic PUP that has exactly one root. Find a formula for this root
in terms of a, b, and c. Express the formula using only the operations ⊕, , parallel-universe exponentiation,
parallel square roots, and parallel reciprocals. [5 pts]

Suppose f (x) = (a x 2 ) ⊕ (b x) ⊕ c is a quadratic PUP. A factorization of f is an ordered triple of real numbers

(k, r, s) such that r ≤ s, and
f (x) = k (x ⊕ r) (x ⊕ s),
for all real numbers x.
8. Prove that if (k, r, s) is a factorization of f , then:
a. k = a, and [2 pts]
b. r s=c a.
b [2 pts]
9. Prove that every quadratic PUP has exactly one factorization. [5 pts]

15
5 Solutions to Power Question
1. a. No. This equation would mean that min(O, x) = x for all x. However, if O exists, then min(O, O + 1) =
O=6 O + 1. This is a contradiction, so no such O exists.
b. I = 0. This makes I x = I + x = 0 + x = x for all real numbers x.
2. Translating this into “normal” arithmetic, the left-hand side is a + min(b, c), and the right-hand side is
min(a + b, a + c). If b ≤ c, then both sides are equal to a + b, and if b > c, then both sides are equal to
a + c. Thus this equation holds for all real numbers a, b, c.
1
3. a. The equation translates to min(2x, x + 2, 1) = min(x, 2). If x < , the equation is equivalent to 2x = x,
2
1
which implies x = 0. If x ≥ , the equation is equivalent to 1 = min(x, 2), which implies x = 1. These
2
both satisfy the original equation, and there cannot be any other solutions. Thus the two solutions are
x = 0 and x = 1.
b. This equation translates to min(2x, −1) = min(x, 1) + min(x, −1). If x ≤ −1, then both sides are equal to
2x, so every x ≤ −1 is a solution. If −1 < x < − 12 , then the equation becomes 2x = x − 1, which has no
solution on the interval. If − 12 ≤ x ≤ 1, then the equation is −1 = x − 1, so x = 0 is a solution. Finally,
if x > 1, then the equation becomes −1 = 0, which is not true for any value of x. Thus the solutions to
this equation are x = 0 and all x ≤ −1.
b
4. Note that, translating into normal arithmetic, a = b · a. Each of the statements will be proven by translation.
a. This equation translates to n(x + y) = nx + ny, which is just the normal distributive property of multi-
plication over addition.
b. This translates to n · min(x, y) = min(nx, ny). If x ≤ y, then both sides are equal to nx, and if x > y,
then both sides are equal to ny. Either way, the equation holds for all real x and y and all integers n.
c. The first equality holds by part (b). For the second, the left-hand side is min(2x, 2y), and the right-hand
side is min(2x, x + y, 2y). But note that x + y is the arithmetic mean of 2x and 2y, so it must be greater
than or equal to one of them and less than or equal to the other. Thus min(2x, x + y, 2y) = min(2x, 2y),
as desired.
√
5. a. This equation translates to 2x = a, which has exactly one solution: x = ◦ a = a2 .
b. This equation translates to a + x = I = 0, which has exactly one solution: x = b
a = −a.
c. As shown in parts (a) and (b), the left-hand side is equal to 21 (−a), and the right-hand side is equal to
1

− 2 a , and these are equal by associativity of (normal) multiplication.

6. a. Let f (x) = (0 x 2 ) ⊕ (0 x) ⊕ 1. Then f (x) = min(2x, x, 1), which is 2x for x ≤ 0, x for 0 < x < 1,
and 1 for x ≥ 1. The only root of this PUP is x = 0.
b. The answer is yes. As an example, let f (x) = (4 x 2 )⊕(3 x)⊕(−1). Then f (x) = min(2x+4, x+3, −1),
so f (x) ≤ −1 for all x, which means f has no roots.
c. The answer is yes. As an example, let f (x) = (0 x 2 ) ⊕ (0 x) ⊕ 0. Then f (x) = min(2x, x, 0) = 0 for
all x ≥ 0, so all nonnegative real numbers are roots of f .
Note: A general quadratic PUP (a x 2 ) ⊕ (b x) ⊕ c has no roots if c < 0, infinitely many roots if
c = 0, and one root if c > 0. A proof of this fact is in the solution to Problem 7.
d. Consider a general PUP, f (x), as defined in the background to Problem 6. Then

f (x) = min(nx + an , (n − 1)x + an−1 , . . . , 2x + a2 , x + a1 , a0 ).

This is the minimum of n + 1 functions, all of whose graphs are lines of nonnegative slope. Suppose x ≤ y.
Then kx + ak ≤ ky + ak for each k ≥ 0, and therefore f (x) ≤ f (y). Now suppose f has two different roots,
so there are real numbers r1 , r2 with r1 < r2 and f (r1 ) = f (r2 ) = 0. Now, for any z with r1 < z < r2 ,
f (r1 ) ≤ f (z) ≤ f (r2 ), so it follows that f (z) = 0. Thus every z in the interval [r1 , r2 ] is a root of f ,

16
hence f has infinitely many roots. As seen in parts (a) and (b), if f does not have infinitely many roots,
then it can only have either one root or zero roots. This shows that f has either zero, one, or infinitely
many roots.
7. There are multiple ways to solve this problem; in particular, the root can be found by translating to normal
arithmetic. However, a solution using almost exclusively parallel-universe arithmetic follows.
Setting f (x) = 0 and translating to normal arithmetic, note that the given equation is equivalent to
f (x) = min(2x + a, x + b, c) = 0. First, note that if x > max c−a 2 , c − b , then f (x) = c, and f (x) ≤ c
for all x. Thus if c < 0, then f has no roots, and if c = 0, then f has infinitely many roots. Hence if f has
n ) = (b n
exactly one root, it follows that c > 0. Notice that in general, (a’ a) . This is because this equation
√
translates to −(na) = n(−a), which holds by associativity of (normal) multiplication. Now let x = ◊ ◦
a ⊕ b;
this is the root of f . Substituting and using the result from the above paragraph yields
√
◦
2 √
◦

f (x) = a a⊕b ⊕b a ⊕ b ⊕ c.
¤ Ÿ

Form the parallel-universe equivalent of a common denominator:

√
◦
2 √
◦
√
◦
2 √
◦
2 √
◦
2
f (x) = a a⊕b ⊕ b a⊕b a⊕b ⊕ c a⊕b a⊕b .
¤ ¤ ¤

Now use the distributive property, effectively combining like terms:

î √
◦
√
◦
2
ó √
◦
2
f (x) = a ⊕ b a⊕b ⊕c a⊕b a⊕b .
¤

Using the results of Problems 2 and 4 yields:

√
◦
2 2
√
◦
2
f (x) = a ⊕ b a ⊕b ⊕ (c a) ⊕ c b a⊕b .
¤

2 2 2

Now recall the assumption that c > 0, and therefore a ⊕ (c a) = a, and b ⊕ c b =b . Thus

√
◦
2
√
◦
2
f (x) = a ⊕ b a ⊕b a⊕b .
¤

By Problem 4c, this becomes:

√
◦
2 √
◦
2
f (x) = a⊕b a⊕b = 0,
¤

as desired.
√
q
◦ ◦
Other forms of the answer, such as (a
Ÿ ⊕ b 2 ), are also possible. However, answers like b a ⊕ bb are not
valid, because Problem 4a does not hold if n is a negative integer.
8. Before proving any part of this problem, let (k, r, s) be a factorization of f , and let

g(x) = k (x ⊕ r) (x ⊕ s),

so that f (x) = g(x) for all x. By the result of Problem 2,

2
g(x) = (k x ) ⊕ (k (r ⊕ s) x) ⊕ (k r s).

Also, note as in Problem 7 that f (x) = min(2x + a, x + b, c).

a. If x < min c−a x 2 . Similarly, if x < r, then g(x) = k x 2 . Choose a

2 , b − a , then f (x) = 2x + a = a
value of x small enough to satisfy both inequalities. Because f (x) = g(x), it follows that a x 2 = k x 2 .
Multiplying (in the parallel universe) by (x
’ 2 ) yields k = a, as desired.

b. As argued in Problem 7, if x > max c−a

2 , c − b , then f (x) = c. Similarly, if x > s, then g(x) = k r s.
Choose x large enough to satisfy both inequalities. Because f (x) = g(x), it follows that c = k r s. By
part (a), k = a, and then by the definition of the parallel reciprocal, r s = c b a.

17
9. As in Problem 8, let (k, r, s) be a factorization of a quadratic PUP f (x), and let g(x) = k (x ⊕ r) (x ⊕ s).
Note by the distributive property (Problem 2), that
2
g(x) = (k x ) ⊕ (k r x) ⊕ (k r s)
because r ≤ s and thus r ⊕ s = r. Furthermore, from Problems 8a and 8b, this becomes
2
g(x) = (a x ) ⊕ (a r x) ⊕ c.

Claim: If 2b < a + c, then the only factorization of f is a x ⊕ (b − a) x ⊕ (c − b) (where “−” denotes
normal subtraction), and if 2b ≥ a + c, then the only factorization of f is a (x ⊕ d) (x ⊕ d), where d = c−a
2 .

To prove the claim, consider each case of the claim separately:

c−a
Case 1: Assume 2b < a + c, so that b − a < c − b. Let x = Then 2 .
c−a
f (x) = min(2x + a, x + b, c) = min c, + b, c .
2
Because 2b < a + c, it follows that c − a < 2c − 2b, and thus x < c − b. Then x + b < c, so f (x) = x + b.

Note also that c+a

g(x) = min(2x + a, x + a + r, c) = min c, + r, c .
2
Because f (x) = x + b < c, conclude that g(x) 6= c, hence g(x) = c+a 2 + r. Setting this equal to f (x) yields
r = b − a, and the result of Problem 8b gives s = c − b.

This is the only possible factorization of f , and it remains to show that it is a factorization. Recall that
2
g(x) = (a x ) ⊕ (a r x) ⊕ (a r s).
Using the definition of ,a r = b and a r s = c, so g(x) = f (x) for all x, as desired.

Case 2: Now assume 2b ≥ a + c, so that c − b ≤ b − a. First, note that for every x, either x ≤ b − a
(and then 2x + a ≤ x + b), or c − b ≤ x (and then c ≤ x + b). Thus
f (x) = min(2x + a, x + b, c) = min(2x + a, c).
Now consider the claim that r ≥ c − a − r. If not, then there exists an x with r < x < c − a − r. Because
r < x, it follows that x + a + r < 2x + a. Note that x < c − a − r is equivalent to x + a + r < c. This implies
that g(x) = x + a + r. And because f (x) is equal to either 2x + a or c, it follows that x = r or x = c − a − r, a
contradiction. Therefore r ≥ c − a − r, and rearranging, r ≥ c−a
2 .

c−a
Now consider s. By Problem 8b, r + s = c − a, and r ≤ s by assumption, so 2r ≤ c − a, and r ≤ 2 .
Thus the only possible factorization of f has r = s = c−a
2 (and k = a).

c−a
It remains to verify that this is a factorization of f . For convenience, let d = 2 . An earlier result showed
that
f (x) = min(2x + a, c) = (a x 2 ) ⊕ c.
Now apply the result of Problem 4b to g:
2 2 2 2
g(x) = a (x ⊕ d) = (a x ) ⊕ (a d ) = (a x ) ⊕ c = f (x),
as desired.
Authors’ Note: Elle has in fact discovered the relatively new field of tropical algebra, which is currently an area of
active mathematical research. For a further introduction to the field, including applications to algebraic geometry and
computational biology, see the article “Tropical Mathematics” by David Speyer and Bernd Sturmfels, appearing in the
June 2009 issue of Mathematics Magazine. This article was also invaluable to the authors as a source of inspiration and
a few problems in this Power Question. It is available online at https://math.berkeley.edu/~bernd/mathmag.pdf.

18
6 Individual Problems

Problem 1. Compute the number of perfect squares in the set {11 , 22 , 33 , . . . , 20172017 }.

√ √
Problem 2. Trapezoid ARM L has AR k M L. Given that AR = 4, RM = 26, M L = 12, and LA = 42,
compute AM .

Problem 3. Compute the number of ordered pairs of integers (a, b) such that the polynomials x2 − ax + 24 and
x2 − bx + 36 have one root in common.

−→
Problem 4. In 4ABC, m∠A = 90◦ , AC = 1, and AB = 5. Point D lies on ray AC such that m∠DBC =
2m∠CBA. Compute AD.

− 23 +2 cos θ 1
4 +cos θ
Problem 5. Given that 2 +1 = 2 , compute cos 2θ.

Problem 6. A diagonal of a regular 2017-gon is chosen at random. Compute the probability that the chosen
diagonal is longer than the median length of all of the diagonals.

√
Problem 7. Given that i = −1, compute (i+1)3 (i−2)3 +3(i+1)2 (i+3)(i−2)2 +3(i+1)(i+3)2 (i−2)+(i+1)3 .

Problem 8. In triangle ABC, m∠C = 90◦ and BC = 17. Point E lies on side BC such that m∠CAE = m∠EAB.
The circumcircle of triangle ABE passes through a point F on side AC. Given that CF = 3, compute AB.

Problem 9. Let S be the set of divisors of 67 · 9! + 27 · 8!. Compute the median of S.

Problem 10. Rhombus ARM L has its vertices on the graph of y = bxc − {x}. Given that [ARM L] = 8, compute
the least upper bound for tan A.

19
7 Answers to Individual Problems

Answer 1. 1030

√
Answer 2. 66

Answer 3. 12

Å ã
37 4
Answer 4. or 3 or 3.36
11 11

1
Answer 5. (or 0.125)
8

503
Answer 6.
1007

Answer 7. −24i − 20 or −20 − 24i

Å ã
149 2
Answer 8. or 49 or 49.6
3 3

Answer 9. 5040

100
Answer 10.
621

20
8 Solutions to Individual Problems

Problem 1. Compute the number of perfect squares in the set {11 , 22 , 33 , . . . , 20172017 }.

Solution 1. For a positive integer k, consider whether k k is even or odd. If k k is even, then k is also even, and
k k = (k k/2 )2 is a perfect square. There are b2017/2c = 1008 even k k in the set. If k k is odd, then k is also
odd, and k k will be a perfect square if and only if k is a perfect square. Because 442 < 2017 < 452 , and 22 of
the first 44 natural numbers are odd, there are 22 odd squares in the set. The total number of perfect squares
in the given set is therefore 1008 + 22 = 1030.

√ √
Problem 2. Trapezoid ARM L has AR k M L. Given that AR = 4, RM = 26, M L = 12, and LA = 42,
compute AM .

Solution 2. Draw lines perpendicular to LM through A and R, meeting LM at B and C, respectively, as shown
in the diagram below. (As discussed in the note below, it can be shown that B and C must lie on the segment
LM .)

A R

L B C M

By the Pythagorean Theorem, AB 2 + LB 2 = 42 and RC 2 + CM 2 = 26. Because AB = RC, LB + BC + CM =

12, and RC 2 + CM 2 = 26, it follows that

AB 2 + (8 − LB)2 = 26
=⇒ AB 2 + 64 − 16LB + LB 2 = 26.
2 2
Substitute √ + LB to obtain 64 − 16LB + 42 = 26, from which it follows that LB = 5, and
√ 42 for AB
2
AB = 42 − 5 = 17. Apply the Pythagorean Theorem once more to obtain
p
AM = AB 2 + BM 2
»
= AB 2 + (LM − LB)2
»
= AB 2 + (12 − 5)2
√
= 66.

Note: In general, trapezoids with a given set of side lengths are unique, i.e., there is a SSSS congruence
theorem for trapezoids. In this case, assuming that C does not lie on LM leads to a contradiction as follows.
Let CM = x and RC = h, implying BL = (x + 12) − 4 = x + 8. Then x2 + h2 = 26 and (x + 8)2 + h2 = 42,
yielding x = −3, but side lengths must be positive, so this is a contradiction.

Problem 3. Compute the number of ordered pairs of integers (a, b) such that the polynomials x2 − ax + 24 and
x2 − bx + 36 have one root in common.

21
Solution 3. In general, if r is a root of f (x) and of g(x), then r is a root of f (x) − g(x). The common root of
x2 − ax + 24 and x2 − bx + 36 is a root of (b − a)x − 12. It is evident that there are no solutions where b = a,
because the constant polynomial 0x − 12 has no roots. It follows that the common root of the two polynomials
12
is . Substitute this root into x2 − ax + 24, and notice that
b−a
144 12
0= 2
−a· + 24
(b − a) b−a
0 = 144 − 12a(b − a) + 24(b − a)2
−12 = 3a2 + 2b2 − 5ab
−12 = (a − b)(3a − 2b).

Because a − b and 3a − 2b are integers, a − b must be an integer that divides 12. If k divides 12, then a − b = k
implies that a = k + b, so 3(k + b) − 2b = − 12 12
k . Solve for b to obtain b = − k − 3k, which is also an integer.
Thus choosing any integer that divides 12 for a − b will result in a distinct ordered pair of integers (a, b). There
are 12 integers that divide 12 (namely ±1, ±2, ±3, ±4, ±6, and ±12), so there are 12 such ordered pairs of
integers.

−→
Problem 4. In 4ABC, m∠A = 90◦ , AC = 1, and AB = 5. Point D lies on ray AC such that m∠DBC =
2m∠CBA. Compute AD.

Solution 4. Consider the diagram below.

A B
5

AD AD
Because m∠DBC = 2m∠CBA, it follows that m∠DBA = 3m∠CBA. Thus = = tan(3m∠CBA).
AB 5
3 tan θ − tan3 θ AD 3 · 15 − 125
1
Using the triple-angle formula, tan 3θ = , substitute to obtain = 1 , so AD =
1 − 3 tan2 θ 5 1 − 3 · 25
1
3 − 25 75 − 1 37
3 = 25 − 3 = 11 .
1 − 25

− 32 +2 cos θ 1
4 +cos θ
Problem 5. Given that 2 +1 = 2 , compute cos 2θ.

√ √ » √
Solution 5. Let x = 2cos θ and let k = 2−3/2 . Then k1 = 8 ⇒ 2k 1
= 2 ⇒ 1
2k = 4 2. Then the given
»
1
» √ »1
equation becomes kx2 + 1 = x 2k , which is equivalent to x2 − xk 2k
1
+ k1 = x2 − x · 8 · 2k + k1 =
» » 2 »
x2 − 2x k1 + k1 = x − k1 = 0. Thus 2cos θ = x = 1
k = 2
3/4
and therefore cos θ = 34 . Hence
2
cos 2θ = 2 cos2 θ − 1 = 2 · 43 − 1 = 18 .

22
Alternate Solution: The expressions in the exponents cos θ and 2 cos θ are suggestive of a quadratic equation.
1 1
Let x = 2 4 +cos θ . Then x2 = 2 2 +2 cos θ ; note that this exponent is 2 more than the exponent on the left-hand
side. Because 22 = 4, rewrite the left side as 14 x2 + 1 to obtain the equation 14 x2 + 1 = x. Hence x2 − 4x + 4 = 0,
yielding x = 2. Then 41 + cos θ = 1 ⇒ cos θ = 34 . Proceed as in the first solution to obtain cos 2θ = 18 .

Problem 6. A diagonal of a regular 2017-gon is chosen at random. Compute the probability that the chosen
diagonal is longer than the median length of all of the diagonals.

Solution 6. Consider the 2014 diagonals emanating from a particular vertex. By symmetry, the 1007 distinct
lengths occur in pairs. The complete list of diagonal lengths consists of 2017 copies of the 1007 distinct lengths.
As 1007 is odd, 1007−1
2 = 503 of the 1007 lengths exceed the median length and the probability is 1007 503
.
Considering that making 2017 copies of the list simply multiplies the numerator and denominator by 2017, it
does not change the answer.

√
Problem 7. Given that i = −1, compute (i+1)3 (i−2)3 +3(i+1)2 (i+3)(i−2)2 +3(i+1)(i+3)2 (i−2)+(i+1)3 .

Solution 7. Let S denote the sum. Rather than expand each term, note that [(i + 1)(i − 2) + (i + 3)]3 =

(i + 1)3 (i − 2)3 + 3(i + 1)2 (i − 2)2 (i + 3) + 3(i + 1)(i − 2)(i + 3)2 + (i + 3)3

and (i + 1)(i − 2) + (i + 3) = i2 + 1 = 0. By subtraction, S − 0 = (i + 1)3 − (i + 3)3 = (i3 + 3i2 + 3i + 1) −

(i3 + 9i2 + 27i + 27) = −6i2 − 24i − 26 = −24i − 20.

Solution 8. First note that because m∠CAE = m∠EAB, minor arcs F ˜E and EB˜ have the same length, and so
EF = EB. Then if CE = x, EB = 17 − x = EF , so by the Pythagorean Theorem, x2 + 32 = (17 − x)2 , or
34x = 280, yielding x = 140 149
17 and EB = 17 − x = 17 . Proceed by constructing G on AB such that EG ⊥ AB,
yielding the diagram below.

B
G

17 - x

F
3
A
C

23
Then 4ACE ∼ = 4AGE because both are right triangles sharing hypotenuse AE and with congruent angles at A.
So EG = CE. Hence 4CEF ∼ = 4GEB and GB = CF = 3. However, AG = AC implies that GB = AB −AC.
Then AB 2 − AC 2 = BC 2 = 289 and AB 2 − AC 2 = (AB + AC)(AB − AC) yields AB + AC = 289 3 . Then
(AB−AC)+(AB+AC) 3+289/3 149
AB = 2 = 2 = 3 .

Alternate Solution: As in the first solution, compute CE = 140 149

17 and EB = 17 . Then by the Power of
140 140
the Point theorem, CE · CB = CF · CA, so 17 · 17 = 3 · CA, yielding CA = 3 . From this point there are at
least two ways to compute AB. Using the Pythagorean Theorem is messy because it results in AB 2 = 22201 9 ,
√
although a reasonable guess at 22201 is 149 (because 22201 is just slightly less than 22500 = 1502 and the last
digit of 22201 is 1), which turns out to be correct. Using the Angle Bisector Theorem yields AB BE 149
AC = EC = 140 ,
140 149
and given that AC = 3 , AB = 3 .

Problem 9. Let S be the set of divisors of 67 · 9! + 27 · 8!. Compute the median of S.

Solution 9. Rewrite 67 · 9! + 27 · 8! as (67 · 9 + 27)8! = 630 · 8! = 5040 · 7! = (7!)2 . Divisors occur in pairs except
for the central divisor, 7! = 5040, which is the median of S.

Problem 10. Rhombus ARM L has its vertices on the graph of y = bxc − {x}. Given that [ARM L] = 8, compute
the least upper bound for tan A.

Solution 10. As x = bxc + {x},√ y = 2bxc − x or y = 2n − x on the interval n ≤ x < n + 1 for integers n. These
line segments have length 2 and are separated by that same length as shown in the diagram below.

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
-1

-2

-3

-4

-5

The following argument shows that for any rhombus to lie on this graph, one pair of opposite vertices must
lie on the same diagonal segment. In fact, if U and V are points on different segments of the graph with
coordinates (xU , yU ) and (xV , yV ) respectively, then the slope of segment U V must be positive. Suppose,
without loss of generality, that xU < xV . Then yU < yV because each segment is above the segment to its
immediate left. Thus the slope of U V is xyVV −y
−xU > 0. Because the diagonals of a rhombus are perpendicular and
U

perpendicular segments have opposite reciprocal slopes, it is impossible that both diagonals of the rhombus
have positive slopes. Hence one pair of opposite vertices of the rhombus must lie on the same diagonal segment.
This argument yields a further conclusion: because the slope of the diagonal with negative slope is −1, the
other two vertices must lie on a line parallel to the line y = x. A rhombus satisfying these conditions is shown
in the diagram below.

24
5

3
M
2

1
R
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
-1
L
-2

-3

A -4

-5

In order for tan A to be positive, ∠A must be acute. Hence A must be the lower-left (or upper-right) vertex
of the rhombus, with R and √ √ diagonal segment; the slope of AM is 1. Because the distance
L on the same
between adjacent√segments is 2, AM = k 2 for some integer k, and because R and L lie on the same diagonal
segment, RL < 2. (The inequality is strict because the right endpoint
√
of each segment is not included in
1 16 8 2
the graph.) Because [ARM L] = 2 AM · RL = 8, RL = k 2 = k . Hence k > 8, and to maximize ∠A, it
√

suffices to maximize RL. If k = 9, however, the midpoint of AM does not lie on any segment. Thus k = 10
√ √ √ RL/2
8 2
and RL = 10 = 5 while AM = 10 2. Hence tan A2 = AM/2
4 2 RL
= AM 2
= 25 . Using the double-angle identity,

2 tan A2 4
25 100 100
tan A = 2 A
= 4 = 625 − 4 = 621 .
1 − tan 2 1 − 625

Note that this least upper bound for tan A is in fact attainable, although the problem was stated to avoid
having to assume that conclusion.

25
9 Relay Problems

Relay 1-1. Compute the greatest integer b for which logb (136!) − logb (135!) − logb (17) is an integer.

Relay 1-2. Let T = T N Y W R. Given that x and y are integers satisfying x2 − y 2 = T , compute the least possible
value of x2 + y 2 .

Relay 1-3. Let T = T N Y W R. Rectangle SHAW has side lengths SH = 24 and SW = T . Point D lies on AW
such that HD ⊥ SA. Point E is the intersection of HD and SA. Compute DE.

Relay 2-1. Compute the number of three-digit positive integers that are divisible by 11 and have middle digit 6.

Relay 2-2. Let T = T N Y W R. Compute the least positive integer N such that when a fair N -sided die whose faces
are numbered consecutively from 1 through N is rolled once, the probability of rolling a factor of N is less
1
than .
T

T2
Relay 2-3. Let T = T N Y W R. Given that 1 − r + r3 − r4 + r6 − r7 + · · · = , compute r.
1 + T + T2

26
10 Relay Answers

Answer 1-1. 8

Answer 1-2. 10

Å ã
125 47
Answer 1-3. or 1
78 78

Answer 2-1. 8

Answer 2-2. 17

1
Answer 2-3.
17

27
11 Relay Solutions

Relay 1-1. Compute the greatest integer b for which logb (136!) − logb (135!) − logb (17) is an integer.

Solution 1-1. Using the property logb x − logb y = logb xy , logb (136!) − logb (135!) = logb 136!

135! = logb (136) and
logb (136) − logb (17) = logb (136/17) = logb (8). Therefore the greatest integer b for which logb (8) is an integer
is b = 8.

Relay 1-2. Let T = T N Y W R. Given that x and y are integers satisfying x2 − y 2 = T , compute the least possible
value of x2 + y 2 .

Solution 1-2. Note that T = (x + y)(x − y). Let m = x + y and n = x − y. Solving the system for x and y yields
(x, y) = m+n m−n
2 , 2 . Therefore x and y are both integers if and only if m and n have the same parity. Thus
it suffices to consider all factorizations of T = mn such that m and n have the same parity and m ≥ n ≥ 1.
(The reader can verify that the value of x2 + y 2 does not change in the cases where n = x + y and m = x − y
or where −m = x + y and −n = x − y.) With T = 8, the only ordered pair of positive integers (m, n) of the
same parity with m ≥ n and mn = 8 is (4, 2), resulting in (x, y) = (3, 1). Hence x2 + y 2 = 32 + 12 = 10.

Alternate Solution: Noting that x2 + y 2 = x2 − y 2 + 2y 2 = T + 2y 2 , if T is fixed, then x2 + y 2 is minimized

when y 2 is minimized. With T = 8, the integers x = 3 and y = 1 satisfy the given equation, so the least
possible value of x2 + y 2 is 10.

Relay 1-3. Let T = T N Y W R. Rectangle SHAW has side lengths SH = 24 and SW = T . Point D lies on AW
such that HD ⊥ SA. Point E is the intersection of HD and SA. Compute DE.

Solution 1-3. Note that 4SHE ∼ 4ADE because ∠HSE ∼ = ∠DAE and ∠HES ∼ = ∠DEA. Let HE = a and
SE = b. Then for some positive real number k < 1, it follows that DE = ka, AE = kb, and AD = k · SH =
24k. Also note that 4SHE ∼ 4HAE ∼ 4SAH, hence HE/SE = AE/HE = AH/SH, which implies
a/b = kb/a = T /24. Using a/b = kb/a, conclude that k = a2 /b2 = (T /24)2 . Using a/b = T /24, conclude that
b = 24a/T , and because a2 + b2 = 242 (by the Pythagorean Theorem), it follows that a = √T24T 2 +242
. Hence
3
DE = ka = 24√TT2 +242 (?). With T = 10, the denominator of the fraction in (?) simplifies to 24 · 26 and the
numerator is 103 = 1000. Because 1000 and 24 have a common factor of 8, the expression for DE simplifies to
125 125
= .
3 · 26 78

Relay 2-1. Compute the number of three-digit positive integers that are divisible by 11 and have middle digit 6.

Solution 2-1. Represent the number as A 6 B. Then A − 6 + B must be a multiple of 11. This can only happen
if A + B = 6 or A + B = 17. The first case has six possibilities (A = 1, 2, 3, 4, 5, 6) and the latter case has two
possibilities (A = 8, 9) for a total of 8.

28
Relay 2-2. Let T = T N Y W R. Compute the least positive integer N such that when a fair N -sided die whose faces
are numbered consecutively from 1 through N is rolled once, the probability of rolling a factor of N is less
1
than .
T

Solution 2-2. Let P (N ) denote the probability of rolling a factor of N on a fair N -sided die. Then P (1) = 1,
P (2) = 1, P (3) = 32 , P (4) = 43 , P (5) = 25 , and P (6) = 23 . More generally, P (N ) = N2 when N is prime and
P (N ) is greater than N2 when N is composite. Try the least prime N such that N2 < T1 , or N > 2T . As
2
T = 8, try N = 17. It is straightforward to see that P (17) = 17 < 81 , and it can also easily be checked that for
1
1 ≤ N ≤ 16, P (N ) ≥ T . Thus N = 17.

T2
Relay 2-3. Let T = T N Y W R. Given that 1 − r + r3 − r4 + r6 − r7 + · · · = , compute r.
1 + T + T2

1−r 1
Solution 2-3. The left-hand side of the given equation is (1 − r) + r3 (1 − r) + r6 (1 − r) + · · · = 1−r 3 = 1+r+r 2 for

|r| < 1. The right-hand side of the given equation is 1 +11 +1 , so r = T1 is a solution as long as |T | > 1. More
T2 T
1 T2 1 1 1 1 1 1
generally, 1+r+r 2 = 1+T +T 2 ⇒ r2 + r + 1 = T2 + T + 1 ⇒ r2 + r − T2 − T = 0 ⇒ (r − T )(r + 1 + T ) = 0 so r
1 1
is either T or −1 − T1 . As T = 17, the series only converges for r = .
17

29
12 Super Relay
1. Compute the units digit of 171 + 172 + 173 + . . . + 1716 + 1717 .
2. Let T = T N Y W R. Given that the polynomial x2 − 17x + T K can be factored over the integers and that K is
an integer, compute the greatest possible value of K.
3. Let T = T N Y W R and let K = 17 − T . Let
» p √ K
P = K+ K+ K + ... and J =1+ .
K
1+
K
1+
1 + ...

Compute bP − Jc.
4. Let T = T N Y W R. In 4P JK, each of m∠P and m∠J is an integral multiple of 17◦ and m∠P ≥ m∠J. Let
S be the number of triangles that satisfy these conditions, where no two of these triangles are similar to one
another. Compute S + T .

5. Let√T = T N Y W R. Square IJKL has area T . Diagonal KI is extended past I to point P such that P K =
17 2. Points E and O lie in the plane such that EIOP is a square and E and L lie on the same side of P K.
Compute the perimeter of trapezoid P OLE.
6. Let T = T N Y W R. Donald and John each play the trumpet. They take turns (starting with Donald), where
each person plays a note subject to the constraint that neither person can play the same note that was last
played by the other person. There are a total of 17 different notes they can play and between them, they play
a total of bT + 1c notes. Given that the number of melodies they can play can be expressed in the form w · xy ,
where w, x, and y are positive integers and w and x are relatively prime, compute the least possible value of
w + x + y.
7. Let T = T N Y W R. A triangle is similar to an 8-15-17 triangle and one of its sides is T . Given that the
perimeter of this triangle is an integer, compute the least possible perimeter this triangle can have.

15. Compute the number of ordered triples of integers (a, b, c) that are solutions to the equation abc = 17.
√ 20 17
14. Let T = T N Y W R and let P = T . The complex number + can be expressed in the form
i+P i−P −1
J + Ki, where J and K are real. Compute J + K.
13. Let T = T N Y W R. The circle defined by x2 + y 2 = 17 intersects line ` : y = −T x + 3 in two points (x1 , y1 ) and
1 1
(x2 , y2 ). Let P be the product · , let J be the sum x1 + x2 , and let K be the slope of a line perpendicular
x1 x2
to `. Compute P JK.
12. Let T = T N Y W R. The ages of Catherine, Charlie, and Elizabeth are integers. Charlie’s age is b4T + 68c, and
the sum of Catherine’s and Elizabeth’s ages is 100. Nine years ago, Elizabeth’s age was a positive multiple of
17 and that number had a common factor (greater than 1) with Charlie’s current age. Compute Elizabeth’s
current age.
11. Let T = T N Y W R, let K = T − 10, and let P = 217 . Compute logP 20 + 21 + 22 + . . . + 2K−1 + 2K .

10. Let T = T N Y W R. Iris uses voice recognition software to queue up her favorite songs. When she says a song
1
title aloud, the probability that the software identifies the wrong song is 100 max T, 17
T . Given that Iris says
40 song titles from her favorite band, Iron Maiden, into her voice recognition software, compute the expected
number of songs that the software correctly identifies.

30
9. Let T = T N Y W R. Bryan has 1700 sheets of graph paper and is stuffing ARML team envelopes with 30 sheets
in each envelope. Being exacting, it takes him 11 minutes and 20 seconds to count and stuff bT − 3c sheets of
paper into one or more envelopes. He completely stuffs as many envelopes as he can, until he has fewer than
30 sheets remaining, at which point, he stops. Rounded to the nearest minute, compute the number of minutes
it will take Bryan to complete this arduous task.

8. Let J be the number you will receive from position 7 and let K be the number you will receive from position
9. Let A be the largest prime factor of J and let P be the largest prime factor of K. Consider the following
system of equations:

x3 + y = A
x2 + 4y = P.

This system has three solutions, two of which are (x1 , y1 ) and (x2 , y2 ), where x1 and x2 are non-real. Compute
the value of x1 + x2 .

31
13 Super Relay Answers
1. 7
2. 10
3. 0

4. 25
5. 54
6. 87
7. 232

15. 12
14. −1
3
13. −
4
12. 94
11. 5
10. 38

9. 544

11
8. −
4

32
14 Super Relay Solutions

Problem 1. Compute the units digit of 171 + 172 + 173 + . . . + 1716 + 1717 .

Solution 1. The units digit of 17n is the same as the units digit of 7n . The sequence of units digits of 7n is readily
found to be 7, 9, 3, 1, 7, 9, . . ., which is periodic with period 4. Moreover, the sum of any four consecutive terms
of this sequence has the same units digit as 7 + 9 + 3 + 1 = 20. Thus 171 + 172 + 173 + . . . + 1716 has a units
digit of 0, and the units digit of the given sum is the same as the units digit of 1717 , which is 7.

Problem 2. Let T = T N Y W R. Given that the polynomial x2 − 17x + T K can be factored over the integers and
that K is an integer, compute the greatest possible value of K.

Solution 2. Suppose that x2 − 17x + T K can be factored as (x − r)(x − s). By Vieta’s Formulas, r + s = 17 and
rs = T K. Now consider different possibilities for the sign of T . If T = 0, then K is not uniquely determined.
r(17−r)
If T < 0, then K = rs T = T can be written as a quadratic function of r with a positive leading coefficient,
so K has no maximum value. If T > 0, then to determine the possible values of K, examine pairs of positive
integers r and s that sum to 17 such that T divides rs. With T = 7, consider r = 7, s = 10, so that K = rs
T = 10,
or r = 14, s = 3, so that K = rsT = 6. Thus the greatest possible value of K is 10.

Problem 3. Let T = T N Y W R and let K = 17 − T . Let

» p √ K
P = K+ K+ K + ... and J =1+ .
K
1+
K
1+
1 + ...

Compute bP − Jc.

√
Solution 3. To compute P , write P = K + P . This is equivalent to P 2 − P = K. To compute J, write
J =1+ K 2 2
J . This is equivalent to J −√J = K. Thus P and J are both roots of the equation x − x − K = 0.
The roots of this equation are x = 1± 1+4K
2 . If K = 0, then by inspection, P = 0 and J = 1. If K > 1, then
√
1+ 1+4K
P =J = 2 because the other root is negative, yet P > 0, J > 0. With T = 10, K = 17 − 10 = 7 > 0.
Thus P − J = 0, and the answer is 0.

Problem 4. Let T = T N Y W R. In 4P JK, each of m∠P and m∠J is an integral multiple of 17◦ and m∠P ≥
m∠J. Let S be the number of triangles that satisfy these conditions, where no two of these triangles are similar
to one another. Compute S + T .

Solution 4. Let p = m∠P = (17a)◦ and j = m∠J = (17b)◦ , where a and b are positive integers. Then a ≥ b and
2 ≤ n ≤ 10, where n = a + b. By examining small values of n, conclude that the equation a + b = n has b n2 c
solutions with a ≥ b. Thus there are 2(1 + 2 + 3 + 4) + 5 = 25 possible solutions, so S = 25. With T = 0,
S + T = 25.

Problem 5. √ Let T = T N Y W R. Square IJKL has area T . Diagonal KI is extended past I to point P such that
P K = 17 2. Points E and O lie in the plane such that EIOP is a square and E and L lie on the same side
of P K. Compute the perimeter of trapezoid P OLE.

33
√ √
Solution 5. The side of√square √ IJKL is T and the diagonal’s length is √2T . Thus the diagonal of square
EIOP√has length
√ 17 2 − 2T , hence the side√of square EIOP √ is 17 − T . Note that OL = OI + IL =
(17 − T ) + T = 17, PpO = EP = EI = 17 − T , and LE = EI 2 + IL2 . Thus the perimeter of trapezoid
√ √ √
P OLE is 51 − 2 T + 2T − 34 T + 289. With T = 25, T = 5. Thus square EIOP has side length
17 − 5 = 12, 4LIE is a 5-12-13 triangle, and the perimeter of P OLE is 54.

Problem 6. Let T = T N Y W R. Donald and John each play the trumpet. They take turns (starting with Donald),
where each person plays a note subject to the constraint that neither person can play the same note that was
last played by the other person. There are a total of 17 different notes they can play and between them, they
play a total of bT + 1c notes. Given that the number of melodies they can play can be expressed in the form
w · xy , where w, x, and y are positive integers and w and x are relatively prime, compute the least possible
value of w + x + y.

Solution 6. The first note can be played in 17 ways. Each note after the first can be played in 16 ways. Thus
the total number of melodies that Donald and John can play is 17 · (24 )bT c . Thus w = 17. With T = 54, the
possible values of x are numbers of the form 2d where d is a divisor of 216 = 4 · 54. Some quick computations
reveal that x + y = 2d + 216
d is minimized when d = 4 so that x = 16 and y = 54. Thus the least possible value
of w + x + y is 17 + 16 + 54 = 87.

Problem 7. Let T = T N Y W R. A triangle is similar to an 8-15-17 triangle and one of its sides is T . Given that
the perimeter of this triangle is an integer, compute the least possible perimeter this triangle can have.

Solution 7. Let the triangle have sides 8x, 15x, 17x, where x > 0. Then the triangle’s perimeter is P = 40x and
T ∈ {8x, 15x, 17x}. If T = 8x, then P = 5(8x) = 5T . If T = 15x, then P = 8T 3 . Finally, if T = 17x, then
40T 40 8
P = 17 . Note that 17 < 3 < 5. With T = 87, note that 17 does not divide 87, hence P is not an integer. But
because 3 divides 87, P = 8T
3 is integer and is equal to 232.

Problem 15. Compute the number of ordered triples of integers (a, b, c) that are solutions to the equation abc = 17.

Solution 15. If a, b, and c are all positive, then two of the variables must be 1 and the third variable must be 17.
This gives 3 solutions. If not all of a, b, and c are positive, then exactly two variables must be negative. For
each of the 3 positive solutions, there are 3 choices for which variables could be negative. Thus the answer is
3 + 3 · 3 = 12.

√ 20 17
Problem 14. Let T = T N Y W R and let P = T . The complex number + can be expressed
i+P i−P −1
in the form J + Ki, where J and K are real. Compute J + K.

20 20 −i + P −20i + 20P 17 17 −i − P − 1
Solution 14. Note that = · = . Similarly, = · =
i+P i + P −i + P P2 + 1 i−P −1 i − P − 1 −i − P − 1
−17i − 17(P + 1) √ −20i + 60
. With T = 12, P = 12 = 3. Thus the given expression is equal to +
(P + 1)2 + 1 10
−17i − 68
= (−2i + 6) + (−i − 4) = 2 − 3i. Hence J = 2, K = −3, and J + K = −1.
17

Problem 13. Let T = T N Y W R. The circle defined by x2 + y 2 = 17 intersects line ` : y = −T x + 3 in two points
1 1
(x1 , y1 ) and (x2 , y2 ). Let P be the product · , let J be the sum x1 + x2 , and let K be the slope of a line
x1 x2
perpendicular to `. Compute P JK.

34
−1
Solution 13. Because the slope of ` is −T , it follows that K = −T = T1 . Substitute y = −T x+3 into the equation
2 2 2 2 2 2
x + y = 17 to obtain x + (−T x + 3) = 17 or (1 + T )x − 6T x − 8 = 0. The two roots of this equation are
6T −8 1+T 2 6T
x1 and x2 . By Vieta’s Formulas, x1 + x2 = 1+T 2 and x1 x2 = 1+T 2 . Hence P = −8 and P J = −8 . Finally,
Ä ä
6T
P JK = −8 · T1 = − 34 (independent of T ).

Problem 12. Let T = T N Y W R. The ages of Catherine, Charlie, and Elizabeth are integers. Charlie’s age
is b4T + 68c, and the sum of Catherine’s and Elizabeth’s ages is 100. Nine years ago, Elizabeth’s age was
a positive multiple of 17 and that number had a common factor (greater than 1) with Charlie’s current age.
Compute Elizabeth’s current age.

Solution 12. List the multiples of 17 that are less than 100: 17, 34, 51, 68, 85. Thus Elizabeth’s current age is one
of 26, 43, 60, 77, or 94. With T = − 43 , Charlie’s current age is 65 = 5 · 13. The only possible age of Elizabeth
nine years ago that shares a common factor greater than 1 with 65 is 85. Hence Elizabeth’s current age is 94.

Let T = T N Y W R, let K = T −10, and let P = 217 . Compute logP 20 + 21 + 22 + . . . + 2K−1 + 2K .

Problem 11.

Solution 11. By the formula for the sum of the terms of a geometric series, it follows that 20 + 21 + 22 + . . . +
K−1 K K+1 0 1 2 K−1 K
/ log217 2K+1 = K+1

2 +2 = 2 − 1. Thus logP 2 + 2 + 2 + . . . + 2 +2 17 . With T = 94,
K+1 0 1 2 K−1 K

K = 84, and 17 = 5. Thus logP 2 + 2 + 2 + . . . + 2 +2 = 5.

Problem 10. Let T = T N Y W R. Iris uses voice recognition software to queue up her favorite songs. When she
1
max T, 17

says a song title aloud, the probability that the software identifies the wrong song is 100 T . Given
that Iris says 40 song titles from her favorite band, Iron Maiden, into her voice recognition software, compute
the expected number of songs that the software correctly identifies.

Solution 10. When a song title is read aloud, the probability that the software identifies the correct song is
1
1 − 100 max T, 17 . Thus the expected number of correctly identified songs is 40 − max 2T 34
5 , 5T . With T = 5,
2T 34
T 34

max 5 , 5T = max 2, 25 = 2. Thus the answer is 40 − 2 = 38.

Problem 9. Let T = T N Y W R. Bryan has 1700 sheets of graph paper and is stuffing ARML team envelopes
with 30 sheets in each envelope. Being exacting, it takes him 11 minutes and 20 seconds to count and stuff
bT − 3c sheets of paper into one or more envelopes. He completely stuffs as many envelopes as he can, until
he has fewer than 30 sheets remaining, at which point, he stops. Rounded to the nearest minute, compute the
number of minutes it will take Bryan to complete this arduous task.

Solution 9. Bryan can stuff 1700

30 = 56 envelopes, so he stuffs a total of 56 · 30 = 1680 sheets of paper. Note also
that 11 minutes and 20 seconds is equivalent to 34
3 minutes. Let S be the number of minutes Bryan needs to stuff
the 1680 sheets. Then bT − 3c/ 34 S = 56·30·34 56·30·34

3 = 1680/S. Thus 3bT −3c . With T = 38, S = 3·35 = 8 · 2 · 34 = 544.

Problem 8. Let J be the number you will receive from position 7 and let K be the number you will receive from
position 9. Let A be the largest prime factor of J and let P be the largest prime factor of K. Consider the
following system of equations:

x3 + y = A
x2 + 4y = P.

35
This system has three solutions, two of which are (x1 , y1 ) and (x2 , y2 ), where x1 and x2 are non-real. Compute
the value of x1 + x2 .

Solution 8. Eliminate y by multiplying the first equation by 4 and subtracting 1 the second equation to obtain
4x3 −x2 −4A+P = 0. The sum of the three roots of this equation is − −1 4 = 4 . Let f (x) = 4x 3
−x2
−4A+P .
As long as A and P are real, the equation f (x) = 0 will have at least one real root (to see this, consider the
graph of y = f (x)). The problem implies that two of the roots of the equation f (x) = 0 are non-real and are
equal to x1 and x2 . To determine x1 + x2 , it suffices to determine the real root r of f (x) = 0, and then x1 + x2
will be equal to 41 − r. With J = 232 = 23 · 29 and K = 544 = 25 · 17, it follows that A = 29, P = 17, and
f (x) = 4x3 − x2 − 99. Note that f (3) = 108 − 9 − 99 = 0, hence x = 3 is the real root of f (x) = 0, and it can
be verified that the quadratic factor of f (x) (i.e., 4x2 + 11x + 33) has a negative discriminant, hence no real
roots. Thus the answer is 41 − 3 = − 11
4 .

36
15 Tiebreaker Problems

Problem 1. Compute the least positive N such that there exists a quadratic polynomial f (x) with integer coeffi-
cients satisfying
f (f (1)) = f (f (5)) = f (f (7)) = f (f (11)) = N.

Problem 2. Cube ARM LKHJC, with opposite faces ARM L and HJCK, is inscribed in a cone, such that A is
the vertex of the cone, edges AR, AL, AH lie on the surface of the cone, and vertex C, diagonally opposite A,
is on the base of the cone. Given that AR = 6, compute the radius of the cone.

Problem 3. Given that logb3 a2 + logb9 a4 + logb27 a8 + · · · = 1, compute logb a.

37
16 Tiebreaker Answers

Answer 1. 137

√
Answer 2. 6 6

1
Answer 3. (or 0.5)
2

38
17 Tiebreaker Solutions

Problem 1. Compute the least positive N such that there exists a quadratic polynomial f (x) with integer coeffi-
cients satisfying
f (f (1)) = f (f (5)) = f (f (7)) = f (f (11)) = N.

Solution 1. Because f (x) is a quadratic polynomial, the graph of y = f (x) is symmetric about some vertical line;
for each output, there are no more than two corresponding inputs. Thus if f (p) = f (q) = f (r) = f (s), then the
variables p, q, r, s can take on at most two distinct values. Hence the set {f (1), f (5), f (7), f (11)} contains at
most two distinct integers. However, the same logic shows that the equation f (p) = f (q) = f (r) has no solutions
when p, q, r are required to be distinct. Hence it is impossible that three of the values f (1), f (5), f (7), f (11) be
the same. Thus the set {f (1), f (5), f (7), f (11)} contains exactly two distinct integers, dividing the expressions
f (1), f (5), f (7), f (11) into two pairs of equal values. Because of the symmetry of the graph, there must be
a unique value of h such that f (x) = f (2h − x) for all x. It follows that f (1) = f (11) and f (5) = f (7),
with h = 6. (Any other pairing yields a contradiction: for example, if f (1) = f (7), then h = 4, but the
equality of the other pair f (5) = f (11) yields h = 8.) Hence f (x) is of the form f (x) = a(x − 6)2 + k, and
2
f (f (x)) = a a(x − 6)2 + k − 6 + k.

The following argument shows that N is minimal when a = 1. From the condition that f (f (1)) = f (f (5)) =
f (f (7)) = f (f (11)), it follows that

a(f (1) − 6)2 + k = a(f (5) − 6)2 + k = a(f (7) − 6)2 + k = a(f (11) − 6)2 + k.

Simplifying yields the following equation:

±(f (1) − 6) = ±(f (5) − 6) = ±(f (7) − 6) = ±(f (11) − 6).

Moreover, it follows from the above symmetry argument that f (5) = f (7) 6= f (1) = f (11). Thus f (1) − 6 =
−(f (5) − 6). Because f (1) = 25a + k, and f (5) = a + k, it follows that 25a + k − 6 = −a − k + 6, which is
equivalent to k = 6 − 13a. (Notice that the equation f (7) − 6 = −(f (11) − 6) adds no information and yields
an equivalent equation.) Now, computing N = f (f (5)),

f (f (5)) = f (a + k)
= a · (a + k − 6)2 + k
= a · (−12a)2 + (6 − 13a)
= 144a3 − 13a + 6.

So the problem reduces to finding the minimum positive value of the expression 144a3 − 13a + 6 when a is an
integer. First note that a 6= 0 because f is a quadratic polynomial. Next, if a ≤ −1, then a3 ≤ a, so

N = 144a3 − 13a + 6 ≤ 144a − 13a + 6 = 131a + 6 ≤ −131 + 6 < 0,

which violates the condition that N be positive. Because a is an integer, no values between −1 and 1 need be
considered. For a ≥ 1, a3 ≥ a, so

N = 144a3 − 13a + 6 ≥ 144a − 13a + 6 = 131a + 6,

with equality precisely when a = 1. Thus, with a = 1, N = 137, and for any other positive value of a, N will
be strictly larger. Hence the least positive value of N is 137.

39
Solution 2. The cone and cube are pictured in the diagram below.

A
s

L R P H

M J

C Q

First note that point C must be the center of the cone’s base because of the rotational symmetry of vertices
R, L, H about the axis of the cone. Thus AC is the height of the cone. Furthermore, plane RLH is parallel to
the cone’s base (again, because of the rotational symmetry of vertices R, L, H). Let P be√ the point where AC
intersects plane RLH, and let s = AR. Because
Ä √ √ ä 4RLH√
is equilateral with side length s 2, conclude that P is
2 s 2 s 6
the centroid of 4RLH, and P H = 3 2 · 3 = 3 . Use the Pythagorean Theorem on 4AP H to compute
√ −−→ √
AP = s 3 3 . Extend AH to meet the bottom of the cone at Q. Because AC = s 3 and 4AP H ∼ 4ACQ, it
√ √
follows that PCQ AC
H = AP = 3, hence CQ = s 6. Substitute s = 6 to obtain CQ = 6 6.

Problem 3. Given that logb3 a2 + logb9 a4 + logb27 a8 + · · · = 1, compute logb a.

Solution 3. Apply logarithm laws to the left-hand side of the given equation:
2 4 8
logb3 a2 + logb9 a4 + logb27 a8 + · · · = logb a + logb a + logb a + · · ·
3 Å 9 27 ã
2 4 8
= (logb a) + + + ··· ,
3 9 27
2
which is an infinite geometric series with first term 3 logb a and common ratio 23 . Hence the sum of the series
1
is 2 logb a = 1. Therefore logb a = .
2

40
ARML Competition 2018

George Reuter, Head Writer

Chris Jeuell, Lead Editor
Evan Chen
Paul Dreyer
Edward Early
Zuming Feng
Zachary Franco
Silas Johnson
Paul J. Karafiol
Winston Luo
Jason Mutford
Andy Niedermaier
Graham Rosby
Eric Wepsic

June 1–2, 2018

1 Team Problems

Problem 1. Compute the number of non-empty subsets S of {1, 2, 3, . . . , 20} for which |S| · max{S} = 18. (Note:
|S| is the number of elements of the set S.)

Problem 2. A class of 218 students takes a test. Each student’s score is an integer from 0 to 100, inclusive.
Compute the greatest possible difference between the mean and the median scores.

Problem 3. Regular hexagon RAN GES has side length 6. Pentagon RAN GE is revolved 360◦ about the line
containing RE to obtain a solid. The volume of the solid is k · π. Compute k.

Problem 4. A fair 12-sided die has faces numbered 1 through 12. The die is rolled twice, and the results of the
two rolls are x and y, respectively. Given that tan(2θ) = xy for some θ with 0 < θ < π2 , compute the probability
that tan θ is rational.

Problem 5. The absolute values of the five complex roots of z 5 − 5z 2 + 53 = 0 all lie between the positive integers
a and b, where a < b and b − a is minimal. Compute the ordered pair (a, b).

Problem 6. Let S be the set of points (x, y) whose coordinates satisfy the system of equations:
bxc · dye = 20
dxe · byc = 18.
Compute the least upper bound of the set of distances between points in S.

Problem 7. Compute the least integer

√ d > 0 for which there exist distinct lattice points A, B, and C in the
coordinate plane, each exactly d units from the origin, satisfying csc(∠ABC) > 2018.

Problem 8. Compute the number of unordered collections of three integer-area rectangles such that the three
rectangles can be assembled without overlap to form one 3 × 5 rectangle. (For example, one such collection
contains one 3 × 3 and two 1 × 3 rectangles, and another such collection contains one 3 × 3 and two 2 × 1.5
rectangles. The latter collection is equivalent to the collection of two 1.5×2 rectangles and one 3×3 rectangle.)

Problem 9. Let Γ be a circle with diameter XY and center O, and let γ be a circle with diameter OY . Circle
ω1 passes through Y and intersects Γ and γ again at A and B, respectively. Circle ω2 also passes through Y
and intersects Γ and γ again at D and C, respectively. Given that AB = 1, BC = 4, CD = 2, and AD = 7,
compute the sum of the areas of ω1 and ω2 .

Problem 10. In the number puzzle below, clues are given for the four rows, each of which contains a four-digit
number. Cells inside a region bounded by bold lines must all contain the same digit, and each of the eight
regions contains a different digit. The variables in the clues are all positive integers. Complete the number
puzzle.

1: 4a + 13b + 14c

2: 5p + 13q + 17r

3: 4x + 5y + 31z

4: the average of the other 3 rows

1
2 Answers to Team Problems

Answer 1. 19
5400 59
Ñ é

Answer 2. or 49
109 109
√
Answer 3. 342 3
1
Answer 4. (or 0.05)
18
Answer 5. (2, 3)
√
Answer 6. 2 73

Answer 7. 2,038,181

Answer 8. 132

Answer 9. 20π

5 1 9 7
5 1 0 7
Answer 10.
4 1 0 2
4 8 0 2

2
3 Solutions to Team Problems

Problem 1. Compute the number of non-empty subsets S of {1, 2, 3, . . . , 20} for which |S| · max{S} = 18. (Note:
|S| is the number of elements of the set S.)

Solution 1. Both |S| and max{S} are integers between 1 and 20, so factor 18 and consider the possible ways to
have |S| and max{S} equal each factor. Then count the number of possible sets S that satisfy that condition.

|S| max{S} Number of possible sets S

1 18 1: this can only be the set S = {18}.
2 9 8: this can be any set S = {a, 9}, where 1 ≤ a ≤ 8.
5 5

3 6 2 = 10: the greatest element must be 6, and there are 2 ways of selecting the
two other elements of S.
6 3 0: this is impossible because S cannot have 6 elements, the greatest of which is 3.
9 2 0: this is impossible because S cannot have 9 elements, the greatest of which is 2.
18 1 0: this is impossible because S cannot have 18 elements, the greatest of which is 1.

Thus the answer is 1 + 8 + 10 = 19.

Problem 2. A class of 218 students takes a test. Each student’s score is an integer from 0 to 100, inclusive.
Compute the greatest possible difference between the mean and the median scores.

Solution 2. Intuitively, the maximum difference can be attained by forcing the median to be 0 by selecting just
enough 0 scores, and then maximizing the mean by maximizing the remaining scores. Such a distribution would
have 110 scores of 0 and 108 scores of 100, giving a median score of 0 and a mean score of 108·100
218 = 5400
109 . This
5400
gives a difference of 109 .

To prove that this is maximal, assume that the median score is m. Then, by definition, at least 109 scores
must be at most m. Further, the remaining 109 scores are at most 100, so the mean µ satisfies µ ≤
1
218 (109m + 10900) =
m+100
2 . Therefore µ − m ≤ 100−m
2 .

If m = 0, then the maximum mean is found as above. If m ≥ 1, then µ − m ≤ 100−1 2 = 49.5 < 5400
109 ,
and the difference is greater than it is in the above case with m = 0. There only remains the possibility that
m = 21 . For this to be the case, 109 scores must be 0 and one score must be 1. To maximize the mean, the
remaining 108 scores must be 100. However, compared to the case of having 110 scores of 0 and 108 of 100,
this increases the median by 12 but increases the mean by only 1218, so the difference will certainly be smaller.
Thus the answer is 5400
109 .

Problem 3. Regular hexagon RAN GES has side length 6. Pentagon RAN GE is revolved 360◦ about the line
containing RE to obtain a solid. The volume of the solid is k · π. Compute k.

Solution 3. First note that because RA and EG are parallel, the solid obtained will consist of two conic frustums,
each with a base along the plane through N S and perpendicular to ER. It is easier to compute the volume
−−→ −−→
of one of these conic frustums at a time. Let O be the midpoint of ER, and let N A and ER intersect at P .
Then the conic frustum is a cone with radius ON and height OP , cut off by the plane containing AR and
perpendicular to ER.

3
P

R 6 A

S O N
6 3

E G
√ √ √
Note that ON = N S − OS = 12 − 3 = 9, and OP = OR + RP = 3 3 + 6 3 = 9 3. Then the volume of one
of the conic frustums is
1 Ä √ ä 1 Ä √ ä √
π 92 9 3 − π 62 6 3 = 171 3 π.
3 3
√ √
The volume of the entire solid is therefore 342 3 · π, so k = 342 3.

2 tan θ
Solution 4. Suppose tan 2θ = xy . Then the identity for the tangent of a sum gives 1−tan x
2 θ = y , which is equivalent
2 2
to x−x tan θ = 2y tan θ. Rearranging gives a quadratic equation in tan θ: x tan θ+2y tan θ−x = 0. Thus tan θ
is rational if and only if the discriminant is a perfect square. The discriminant is (2y)2 − 4 · x · (−x) = 4(x2 + y 2 ),
which is a perfect square if and only if x2 + y 2 is a perfect square. The only Pythagorean triples (a, b, c) with
a, b ≤ 12 are (3, 4, 5), (6, 8, 10), (9, 12, 15), and (5, 12, 13). Therefore the only (x, y) that give a rational value for
8 1
tan θ are (3, 4), (4, 3), (6, 8), (8, 6), (9, 12), (12, 9), (5, 12), (12, 5). The desired probability is therefore 12·12 = 18 .

Problem 5. The absolute values of the five complex roots of z 5 − 5z 2 + 53 = 0 all lie between the positive integers
a and b, where a < b and b − a is minimal. Compute the ordered pair (a, b).

Solution 5. First note that the question asks for the tightest possible integer bounds on the magnitude of a zero
of the complex polynomial z 5 − 5z 2 + 53, essentially describing the smallest complex annulus with integer radii
that contains the roots.

Now recall two important properties of the absolute value function: |ab| = |a||b| and |a + b| ≤ |a| + |b| for
all complex numbers a and b. The latter also implies that |a| − |b| ≤ |a − b|. In order to have z 5 − 5z 2 + 53 = 0,
it must be the case that |z 5 − 5z 2 | = 53. This is enough to restrict the possible magnitude of z.

Suppose that |z| > 3. Then |z 5 − 5z 2 | = |z|2 |z 3 − 5| > 9|z 3 − 5| ≥ 9(|z|3 − 5) > 9(27 − 5) = 198 > 53.
So |z| ≤ 3.

Suppose that |z| < 2. Then |z 5 − 5z 2 | = |z|2 |z 3 − 5| < 4|z 3 − 5| ≤ 4(|z|3 + |5|) < 4(8 + 5) = 51 < 53.
So |z| ≥ 2.

Therefore 2 ≤ |z| ≤ 3 for all roots z of z 5 − 5z 2 + 53. Because a < b, and a and b are integers, b − a is
at least 1, so this is the best possible bound. The answer is (2, 3).

4
Problem 6. Let S be the set of points (x, y) whose coordinates satisfy the system of equations:

bxc · dye = 20
dxe · byc = 18.

Compute the least upper bound of the set of distances between points in S.

Solution 6. First note that x and y cannot both be integers because otherwise, the given system would be incon-
sistent. Now suppose that exactly one of x or y is an integer. If x is an integer and byc = b, then dye = b + 1,
and multiplying the two given equations and simplifying results in the equation x2 · b(b + 1) = 360 (†). Because
1 ≤ x2 < 360, the only possible values of x2 are the perfect square factors of 360, namely 1, 4, 9, and 36.
Substituting in x2 = 1, 9, and 36 into (†) results in non-integral solutions for b. But substituting x2 = 4 into
(†) results in b2 + b − 90 = (b + 10)(b − 9) = 0, and b = −10 or b = 9. If x = 2, then byc = 9 and dye = 10,
and this yields the solutions (2, y), where 9 < y < 10. On the other hand, if x = −2, then byc = −9 and
dye = −10, which is impossible. By a similar argument, if x is not an integer and y is an integer, this results in
the solutions (x, −2), where −10 < x < −9. Thus if I1 is the open interval (9, 10) and I2 is the open interval
(−10, −9), then the solutions (x, y) to the given system where exactly one of x and y is an integer are the
ordered pairs belonging to the union of the two sets {2} × I1 ∗ and I2 × {−2}. Graphically, these represent
segments of unit length (one vertical, one horizontal) that do not include the endpoints. For later reference,
let Ix = I2 × {−2} and let Iy = {2} × I1 .

Now suppose that neither x nor y is an integer and that bxc = a and byc = b. Then dxe = a + 1, dye = b + 1,
and multiplying the two given equations and simplifying results in the equation a(a + 1)b(b + 1) = 360 (‡). By
noting the factorization 360 = 3 · 4 · 5 · 6, the following ordered pairs (a, b) satisfy (‡):

(3, 5); (−4, 5); (−4, −6); (3, −6); (5, 3); (5, −4); (−6, −4); (−6, 3).

However, only (a, b) = (5, 3) and (−4, −6) satisfy the given system:

a(b + 1) = 20
(a + 1)b = 18.

Let I3 and I4 denote the open intervals (5, 6) and (3, 4), respectively and let I5 and I6 denote the open intervals
(−4, −3) and (−6, −5), respectively. Then the solutions (x, y) to the given system in which neither x nor y is
an integer are given by the union of the two regions R1 and R2 , defined by

R1 = I3 × I4 and R2 = I5 × I6 .

Each of R1 and R2 is the interior of a unit square (R1 lies in the first quadrant and R2 lies in the third
quadrant). Also note that the solutions are symmetric about the line y = −x. A plot of the points of S is
shown.

∗ Here, × denotes the Cartesian product. I.e., if P and Q are two sets, then P × Q = {(p, q) | p ∈ P and q ∈ Q}.

5
y
10 (2, 10)
9 (2, 9)
8

5
(5, 4)
4

3
(6, 3)
2

−10 −9 −8 −7 −6 −5 −4 −3 −2 −1 1 2 3 4 5 6 7 8 9 10 x
−1
(−9, −2)
−2
(−10, −2)
−3

−4
(−4, −5)
−5

−6
(−3, −6)
−7

−8

−9

−10

The answer to the problem is max{`1 , `2 , `3 }, where:

• `1 is the least upper bound of the distances between a point in Ix and a point in Iy ;
• `2 is the least upper bound of the distances between a point in R1 and a point in R2 ;
• `3 is the least upper bound of the distances between a point in Ix ∪ Iy and a point in R1 ∪ R2 .

Compute `1 by taking the distance between

√ the extremal boundary points of Ix and Iy , namely (−10, −2) and
√
(2, 10), respectively. This distance is 2 · 122 = 288.

Compute `2 by taking the distance between

√ the extremal
√ boundary points of R1 and R2 , namely (6, 4) and
(−4, −6), respectively. This distance is 2 · 102 = 200.

Compute `3 by taking the distance between the extremal boundary points of Ix and √ R1 (or of √
Iy and R2 ,
owing to the symmetry), namely (−10, −2) and (6, 4), respectively. This distance is 162 + 62 = 292.
√ √ √ √ √
Thus the answer is max{ 288, 200, 292} = 292 = 2 73.

Problem 7. Compute the least integer

√ d > 0 for which there exist distinct lattice points A, B, and C in the
coordinate plane, each exactly d units from the origin, satisfying csc(∠ABC) > 2018.

√
Solution 7. The points A, B, and C lie on a circle with diameter 2 d. Consequently, by the Extended Law of
Sines, √
√ AC 2 d
2 d= → csc(∠ABC) = .
sin(∠ABC) AC

6
The solution now proceeds in three cases.
• It is impossible that AC = 1. Indeed, if A = (x, y), then without loss of generality, assume C = (x + 1, y).
Then d = x2 + y 2 = (x + 1)2 + y 2 , which implies 2x + 1 = 0, contradiction.
√
• Suppose AC = 2. If A = (x, y), then without loss of generality, assume C = (x + 1, y + 1), by reflecting
or rotating as necessary. Then d = x2 + y 2 = (x + 1)2 + (y + 1)2 , hence 2x + 1 + 2y + 1 = 0. It follows
that y = −(x + 1) and d must be of the form d = x2 + (x + 1)2 . In that case,
√
2 d » 2
csc(∠ABC) = √ = 2(x + (x + 1)2 )
2

and the least x for which this exceeds 2018 is x = 1009, meaning d = 10092 + 10102 .
√ √ »
• Finally assume AC ≥ 3. Then 2018 < 2ACd ≤ 43 d, which would mean d > 20182 · 34 > 10092 + 10102 .
Thus all values of d achieved in this case are greater than the value of d in the second case.

In conclusion, the least possible value of d is 10092 + 10102 = 2038181. This value of d is indeed possible, as
the points A = (1009, 1010), B = (−1009, −1010), and C = (1010, 1009) satisfy the conditions of the problem.

Solution 8. In order to enumerate the possible rectangles it is necessary to consider several cases. One possible
approach is presented below.

Consider the possible dissections of the given 3 × 5 rectangle. There are four shapes which can be achieved, as
shown below.

Case 1 Case 2
6= 2

Case 3 Case 4

Begin by counting the number of unordered collections that arise from each case.
• In Case 1, a collection corresponds to a multiset {a, b, c} satisfying a + b + c = 15 (as the three rectangles
can be reordered freely). Assume, without loss of generality, that a ≤ b ≤ c. To eliminate the inequality

7
constraints, let x = a − 1, y = b − a, z = c − b denote nonnegative integers. Accordingly, a = x + 1,
b = x + y + 1, z = x + y + z + 1, and the given equation now reads x + 2y + 3z = 12 or, equivalently,
2y + 3z ≤ 12. The number of solutions in nonnegative integers is then
4 Å
12 − 3z
X õ ûã
1+ = 7 + 5 + 4 + 2 + 1 = 19.
z=0
2

• Case 2 has 19 solutions by repeating the argument from Case 1 verbatim.

• In Case 3, denote by 15 − n the area of the top rectangle. The number of possible dissections of the lower
rectangle is then exactly bn/2c. Moreover, the top rectangle (because it has a side length of 5) is never
congruent to either of the lower rectangles. Thus the number of solutions in this case is
14 j k
X n
= 49.
n=1
2

• In Case 4, assume the leftmost region does not have width 2. This ensures that Case 2 and Case 4 will be
disjoint from each other, because the two rectangles on the right will not form a 3 × 3 square that could
be rotated to obtain a dissection already counted in Case 2.

In order to count the number of such configurations, denote by 15 − n the area of the left rectangle,
where 15 − n 6= 3 · 2 or, equivalently, n 6= 9. Then repeating the logic of Case 3 gives a count of
14 j k
n 9
X õ û
= 49 − = 45.
n=1
2 2
n6=9

It has already been checked that Cases 3 and 4 are disjoint, and the other pairs of cases are seen to be mutually
exclusive by comparing the number of sides of length 5 in the dissection. Consequently, the final answer is
19 + 19 + 49 + 45 = 132.

Solution 9. Note that A, B, C, and D are collinear, in that order, because AD = AB + BC + CD. Extend Y B
and Y C to meet Γ again at P and Q, respectively. Then BC is the √ midline of 4Y P Q,p
by homothety.√ By
power of a point, AB · BD = 1 · (4 + 2) = P B · BY = BY 2 , so BY = 6. Similarly, CY = 2 · (4 + 1) = 10.
In particular, 4BY C is right.

8
P X

A
B O

S Y Q
C

Now let S and T be the antipodes of B and C on ω1 and ω2 , respectively. It remains to evaluate BS and CT .
First note that m∠BY S = 90◦ and m∠CAS = m∠BAS = 90◦ . Because m∠BY C = 90◦ , it follows that points
C, Y , and S are collinear and that 4CY B ∼ 4CAS. Therefore
√
…
YB 6
AS = · AC = · 5 = 15,
YC 10
√
2 2
hence BS = AS + AB = 4. In the same fashion, with 4BY C ∼ 4BDT ,
√
…
YC 10
DT = · BD = · 6 = 60,
BY 6
√
hence CT = DT 2 + CD2 = 8. So the sum of the areas of ω1 and ω2 is π(22 + 42 ) = 20π.

1: 4a + 13b + 14c

2: 5p + 13q + 17r

3: 4x + 5y + 31z

4: the average of the other 3 rows

Solution 10. The following solution only relies on the clues for the third and fourth rows. Label the digits in the
array as follows.
A B C D
A B Y D
W B Y Z
W X Y Z

9
The solution proceeds in three steps.

Step 1. The final row is the average of the first three rows, which implies

3(1000W + 100X + 10Y + Z) = 1000(2A + W ) + 100(3B) + 10(C + 2Y ) + (2D + Z).

This rearranges to
0 = 2000(A − W ) + 300(B − X) + 10(C − Y ) + 2(D − Z).
By comparing the magnitudes of |2000(A − W )| to |300(B − X) + 10(C − Y ) + 2(D − Z)|, it is impossible
to have |A − W | ≥ 2, hence A − W = ±1. A similar argument gives B − X = ∓7, and then C − Y = ±9,
D − Z = ±5, where the signs correspond. In particular, {C, Y } = {0, 9}.

Step 2. Next, the number in the third row satisfies

4x + 5y + 31z ≡ 4x + 6 (mod 10)

and consequently either Z = 0 or Z = 2. However, one of C and Y must be 0, hence Z = 2 . This can
only occur if D = 7 , which determines all the signs in the previous step: it follows that A − W = −1,
B − X = −7, C − Y = 9, D − Z = 5. In particular, C = 9 and Y = 0 .

Step 3. Because B − X = −7 and the digits 0 and 2 have already been used, it follows that B = 1 and
X = 8 . The third clue now reads
4x + 5y + 31z = W 1 0 2.
The mod 10 calculation in the previous step implies that x is even and in particular, x > 1. Taking the
previous equation modulo 8 gives 5y + 31z ≡ 102 ≡ 6 (mod 8), which can only occur if y is odd and z
is even. This implies z = 2 (as z ≤ 2). Therefore 4x + 5y ≡ 141 (mod 1000) and because x ≤ 6 is even,
y ≤ 5, and this forces x = 2 and y ∈ {3, 5}. The value y = 3 would give W = 1, which is not permitted
because B = 1, so y = 5 and W = 4 . Finally, A = 5 .

The completed number puzzle reads as follows.

5 1 9 7
5 1 0 7
4 1 0 2
4 8 0 2

10
4 Power Question 2018: Partitions
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part.
To receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “com-
pute,” you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show
your work or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a
problem says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered
items may be used in solutions to later numbered items, but not vice versa. Pages submitted for credit should be
NUMBERED IN CONSECUTIVE ORDER AT THE TOP OF EACH PAGE in what your team considers to be
proper sequential order. PLEASE WRITE ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM
NUMBER (not the team name) on the cover sheet used as the first page of the papers submitted. Do not identify
the team in any other way.

Binary Partitions

A binary partition of a positive integer n is an ordered n-tuple of non-increasing integers, each of which is either 0
or a power of 2, whose sum is n. Each of the integers in the n-tuple is called a part of the partition. Each binary
partition of n has n parts. Let p2 (n) denote the number of binary partitions of n. For example, p2 (3) = 2 because
of the two ordered triples (2, 1, 0) and (1, 1, 1).

1. Compute p2 (n) for n = 4, 5, 6, and 7. [4 pts]

2. a. Show that p2 (n) ≤ p2 (n + 1) for all positive integers n. [3 pts]
b. Is the inequality strict for sufficiently large n? Justify your answer. [3 pts]
a. Prove that if n is even and n ≥ 4, then p2 (n) = p2 n2 + p2 (n − 2).

3. [2 pts]
b. Find the least n > 1 such that p2 (n) is odd, or prove that no such n exists. [2 pts]

Partial Orderings

A partial ordering on a set S is a relation, usually denoted , such that all of the following conditions are true:
• a a for all a ∈ S (reflexivity property),
• a b and b c implies a c for all a, b, c ∈ S (transitivity property), and
• a b and b a implies a = b for all a, b ∈ S (antisymmetry property).

The word “partial” refers to the possibility that some two elements, a and b, may be incomparable, i.e., neither
a b nor b a (these negative relations on are sometimes written as a b and b a, respectively). This Power
Question largely defines and explores a partial ordering on the set of binary partitions of n.

The notation a ≺ b means that a b and a 6= b. The symbols and may also be used, and they are defined in
the following way: a b means b ≺ a, and a b means b a.

If a and b are elements of S with a ≺ b, and if there is no element c ∈ S for which a ≺ c ≺ b, then it is said
that b covers a.

Let a denote the binary partition (a1 , a2 , . . . , an ). Similarly, let b denote the binary partition (b1 , b2 , . . . , bn ).
In this Power Question, define a ≺ b if it is possible to obtain a from b by a sequence of replacing one 2k by
two 2k−1 s (and deleting a 0). For example, (4, 1, 1, 1, 1, 0, 0, 0) ≺ (4, 4, 0, 0, 0, 0, 0, 0) because (4, 1, 1, 1, 1, 0, 0, 0) ≺
(4, 2, 1, 1, 0, 0, 0, 0) ≺ (4, 2, 2, 0, 0, 0, 0, 0) ≺ (4, 4, 0, 0, 0, 0, 0, 0). This Power Question will use this partial ordering on
binary partitions.

11
4. a. Show that the binary partitions of 5 are totally ordered; i.e., if p and p0 are two different binary partitions
of 5, then either p ≺ p0 or p0 ≺ p. [2 pts]
b. Show that the binary partitions of 8 are not totally ordered, i.e., find two binary partitions of 8 – call
them q and q 0 – such that q ⊀ q 0 and q 0 ⊀ q. [3 pts]
5. a. Find the smallest binary partition of n, using this partial ordering. That is, find the binary partition p
such that for all other binary partitions p0 , p ≺ p0 . [2 pts]
b. Find the largest binary partition of n, using this partial ordering. That is, find the binary partition P
such that for all other binary partitions P 0 , P P 0 . [3 pts]

Hasse Diagrams

Suppose that a set S has a partial ordering . Then a Hasse diagram can be used to display the covering relation
in a graphical way. The Hasse diagram is a graph whose vertices are the elements of S and where edges are drawn
between two elements x and y if x ≺ y or y ≺ x and if there is no element z for which x ≺ z and z ≺ y or for which
y ≺ z and z ≺ x. Also, y appears “above” x if x ≺ y. Note that it is possible for more than one element to appear on
the same level of a Hasse diagram. For example, the partially ordered set of divisors of 12, ordered by divisibility, is
shown in Figure 1. In this diagram, the number 1 is said to be at Level 0 because it is the least divisor in the partial
ordering shown. The numbers 2 and 3 are said to be on Level 1 because 2 1 and 3 1 and there is no number n
such that 2 n and n 1 or 3 n and n 1. Other Levels are similarly defined. The number 12 is on Level 3.
12

4 6

2 3

1
Figure 1

Hasse diagrams can be drawn to show the ordering of partitions such as the ones from Problems 4 and 5. For
convenience, rather than labeling the vertices in the Hasse diagram with the partition itself, like (8, 0, 0, 0, 0, 0, 0, 0),
it is common to label the vertex with its nonzero parts only, using exponents to indicate parts within the partition
with multiplicity greater than 1. For example, the partition (8, 0, 0, 0, 0, 0, 0, 0) would be labeled 8, the partition
(4, 2, 2, 0, 0, 0, 0, 0) would be labeled 422 , and the partition (2, 2, 1, 1, 1, 1, 0, 0) would be labeled 22 14 .

6. a. Draw the Hasse diagram for the binary partitions of 8. Label each vertex. [1 pt]
b. List a path through the Hasse diagram for the binary partitions of 8 that begins at the bottom vertex
(that is, the vertex at Level 0) and, traveling only along edges, passes through every other vertex exactly
once. Such a path is called a Hamiltonian path. [1 pt]
c. Let n be an even integer with n > 4. Let S be the set of binary partitions of n. Let S1 be the set of binary
partitions of n2 . Let S2 be the set of binary partitions of n − 2. Prove that there is a bijection B (i.e., a
one-to-one correspondence) from the set S to the set S1 ∪ S2 . Prove that this bijection B preserves order;
that is, given that p ≺ p0 for binary partitions p, p0 ∈ S, then either B(p) ≺ B(p0 ) or B(p) and B(p0 ) are
incomparable. [3 pts]
d. Prove that for each positive integer n, the Hasse diagram of the binary partitions of n has a Hamiltonian
path that begins with the vertex at Level 0. [5 pts]

Let fL (n) represent the number of elements at level L in the Hasse diagram of the binary partitions of n.

7. Prove that the value of fL (n) is the number of binary partitions of n that have n − L nonzero parts. [3 pts]

12
8. Not all partitions are binary partitions. Some partitions have parts of the form 2j − 1 where j is a nonnegative
integer. Such partitions will be called s-partitions, and their parts are written in nonincreasing order. Two
s-partitions of 5 are (3, 1, 1, 0, 0) and (1, 1, 1, 1, 1).
a. List two partitions of 7, one that is an s-partition and one that is neither an s-partition nor a binary
partition. Make sure to identify which is which. [2 pts]
b. Prove that if n ≥ 2L, the value of fL (n) is equal to the number of s-partitions of L. [3 pts]

Trinary Partitions

A trinary partition of a positive integer n is an ordered n-tuple of non-increasing integers, each of which is either 0
or a power of 3, whose sum is n. Let p3 (n) denote the number of trinary partitions of n. For example, p3 (4) = 2
because of the two ordered quadruples (3, 1, 0, 0) and (1, 1, 1, 1).

As with binary partitions, one can define partial orderings for trinary partitions. Let c denote the trinary par-
tition (c1 , c2 , . . . , cn ). Similarly, let d denote the trinary partition (d1 , d2 , . . . , dn ). Define c ≺ d if it is possible to
obtain c from d by a sequence of replacing one 3k by three 3k−1 s (and deleting two 0s).

9. Draw the Hasse diagram for the trinary partitions of 12. Label each vertex. [2 pts]
10. State a value of n less than 23 for which the Hasse diagram of the trinary partitions of n does not contain a
Hamiltonian path. Prove your claim. (Recall that a Hamiltonian path is defined in Problem 6b.) [6 pts]

13
5 Solutions to Power Question
1. The values are as follows.
The value of p2 (4) is 4 from (4, 0, 0, 0), (2, 2, 0, 0), (2, 1, 1, 0), and (1, 1, 1, 1).
The value of p2 (5) is 4 from (4, 1, 0, 0, 0), (2, 2, 1, 0, 0), (2, 1, 1, 1, 0), and (1, 1, 1, 1, 1).
The value of p2 (6) is 6 from (4, 2, 0, 0, 0, 0), (4, 1, 1, 0, 0, 0), (2, 2, 2, 0, 0, 0), (2, 2, 1, 1, 0, 0), (2, 1, 1, 1, 1, 0), and
(1, 1, 1, 1, 1, 1).
The value of p2 (7) is 6 from (4, 2, 1, 0, 0, 0, 0), (4, 1, 1, 1, 0, 0, 0), (2, 2, 2, 1, 0, 0, 0), (2, 2, 1, 1, 1, 0, 0),
(2, 1, 1, 1, 1, 1, 0), and (1, 1, 1, 1, 1, 1, 1).
2. a. A binary partition for n can be changed into a binary partition for n + 1 by inserting a 1 immediately
following the rightmost nonzero entry. For example, (2, 2, 1, 1, 0, 0) is a binary partition of 6, and this
can be changed into (2, 2, 1, 1, 1, 0, 0), which is a binary partition of 7. Similarly, (1, 1, 1, 1) becomes
(1, 1, 1, 1, 1), and these are binary partitions of 4 and 5, respectively. Different binary partitions of n
become different binary partitions of n + 1 in this way. Because every binary partition of n maps to a
binary partition of n + 1, there are at least as many binary partitions of n + 1 as of n.
b. The answer to the question is no. If n is even, then n + 1 is odd. Because a binary partition of an
odd number must contain at least one 1, the correspondence described in the solution to Problem 2a is
one-to-one. To establish this, given a binary partition of n + 1, delete one 1 to obtain a binary partition
of n. Thus p2 (n + 1) = p2 (n) for any even n. However, note that if n is odd, then p2 (n + 1) is strictly
greater than p2 (n) because in addition to the binary partitions of n + 1 that can be obtained by inserting
a 1 into a binary partition of n, there are binary partitions of n + 1 consisting of all even numbers.
3. a. Let n be an even integer. Given a binary partition of n, either all of its parts are even or there is at least
one 1. If all parts are even, then all of the positive parts must occur in the first n2 elements of the n-tuple
(otherwise, their sum would exceed n). Dividing all the parts by 2 yields an n-tuple of powers of 2 and 0s
n n
that sum
to 2 . Note that the last 2 elements of this n-tuple must all be 0s and truncating them results in
an 2 -tuple which is a binary partition of n2 . Conversely, for any binary partition of n2 , by doubling each
n
n
element
and appending 2 zeros at the end results in an n-tuple which is a binary partition of n. Hence
n
p2 2 counts the number of binary partitions of n in which all the parts are even. If there is at least one
1, then the assumption that n is even means that there are at least two 1s, and deleting them produces a
binary partition of n − 2. Conversely, for each binary partition of n − 2, appending two 1s (right before
the 0s, if there are any, otherwise, append them to the end of the (n − 2)-tuple) produces an n-tuple that
is a binary partition of n. Hence p2 (n − 2) counts the number of binary partitions of n with at least one
1, and thus the desired recursion follows.
b. There is no such n. Proceed by cases. If n is odd, then p2 (n) = p2 (n − 1) because all binary partitions
of an odd number include at least one 1, and removing that 1 gives a bijection to partitions of n − 1. If
n is even, then p2 (n) = p2 n2 + p2 (n − 2) by Problem 3a. Because p2 (2) = 2 and p2 (4) = 4, it follows
inductively that p2 (n) is even for all n > 1.
4. a. The four binary partitions of 5 are ordered in the following way: (1, 1, 1, 1, 1) ≺ (2, 1, 1, 1, 0) ≺ (2, 2, 1, 0, 0) ≺
(4, 1, 0, 0, 0). This is a total ordering.
b. Consider these two binary partitions of 8: (4, 1, 1, 1, 1, 0, 0, 0) and (2, 2, 2, 2, 0, 0, 0, 0). The 4 can break
down into two 2s, but that is the only way to generate 2s, and that does not generate four 2s, so it is not
the case that (2, 2, 2, 2, 0, 0, 0, 0) ≺ (4, 1, 1, 1, 1, 0, 0, 0). Likewise, it is not possible for a partition with no 4s
to be related by ≺ to a partition with 4s, so it is not the case that (4, 1, 1, 1, 1, 0, 0, 0) ≺ (2, 2, 2, 2, 0, 0, 0, 0).

Similarly, the following two pairs of partitions are not comparable:

(4, 1, 1, 1, 1, 0, 0, 0) and (2, 2, 2, 1, 1, 0, 0, 0), (4, 2, 1, 1, 0, 0, 0, 0) and (2, 2, 2, 2, 0, 0, 0, 0).

5. a. The smallest binary partition of n is the partition with n 1s. That is, it is the partition (1, 1, . . . , 1) with
n copies of 1 in the partition. If a partition has any part that is not equal to 0 or 1, then that part is a
positive power of 2 that may be replaced by lesser powers of 2. This process can continue until all of the
powers of 2 are 1s, but the process cannot continue beyond that because a partition must consist only

14
of integers. Thus every binary partition is greater than the one consisting of all 1s. Also note that the
partition (1, 1, . . . , 1) is not larger than or incomparable with any other partition. Thus (1, 1, . . . , 1) is the
smallest binary partition.
b. The largest binary partition of n is the one that essentially gives the base-2 equivalent of n. That is, if
the base-2 representation of n has 1s in the 2a , 2b , 2c places, and so on, with a > b > c > · · · , then
the largest binary partition of n is (2a , 2b , 2c , . . . , 0, 0, 0) (with enough 0s to finish the partition). If a
partition contains two copies of a number other than 0, they may be combined to form a greater power
of 2. Thus, starting with any partition, continue combining powers of 2 until the partition contains no
repeated numbers. Because no further combinations are allowed at this point, by the definition, there can
be no partition larger than a partition with no repeated numbers. A priori there could be other partitions
that are not comparable to such a partition, however. But because such a partition expresses n as a sum
of powers of 2, with no repeated nonzero parts, it is essentially the base-2 representation of n, which is
unique. So because any partition may undergo the process of combining like powers of 2 until there are
no repeats, every binary partition is less than the unique partition which expresses n in binary.
6. a. A Hasse diagram for the binary partitions of 8 is shown below.

b. The following sequence of vertices is a Hamiltonian path: 18 , 216 , 22 14 , 414 , 4212 , 23 12 , 24 , 422 , 42 , 8. This
path begins at the vertex at Level 0, then visits each vertex exactly once.
c. The bijection was effectively established in the solution to 3a. Now consider whether B(p) ≺ B(p0 ) for p
and p0 which are different binary partitions of n and for which p ≺ p0 . Suppose first that both p and p0
have no 1s at all. Then dividing all parts by 2 preserves the covering relation. This is because p can be
produced by swapping out powers of 2 in p0 for lesser powers of 2, and dividing all parts by 2 does not
change this covering. Now suppose that p and p0 both have at least two 1s. Then removing two 1s from
each partition preserves the covering relation. This is because removing two 1s from the partition in no
way changes the “swap-outs” of powers of 2 for lesser powers of 2. Now suppose that one of p or p0 has
two 1s and the other doesn’t. Then B(p) is incomparable with B(p0 ), because B(p) ∈ S1 and B(p0 ) ∈ S2
or vice versa, and these are sets of partitions of different numbers because n − 2 > n2 when n > 4. The
relation ≺ is not defined between partitions of different numbers, so B(p) and B(p0 ) are not comparable.
d. As in the solutions to Problems 2 and 3, it is only necessary to consider the problem for even n. This
is because a Hasse diagram for partitions of 2r + 1 is the same as that for the partitions of 2r with an
extra 1 in every partition. So let n = 2r = a · 2k , where a is an odd integer and k is positive. The proof
proceeds by strong induction to prove that there is a Hamiltonian path through the Hasse diagram for

15
partitions of n starting at 1n and ending at (2k )a . To get started, the path 12 → 2 is clearly a Hamiltonian
path through the vertices of the Hasse diagram for partitions of 2, and the path 14 → 212 → 22 → 4 is a
Hamiltonian path through the vertices of the Hasse diagram for partitions of 4. These establish the base
case.

The idea of the proof will be to use the bijection from part c. In essence, the partitions of n will be
split into those that are matched with partitions of n − 2 and those that are partitions of n/2, and the
Hamiltonian paths through each of these sets of smaller partitions will be joined into a single path through
the partitions of n.

The path starts at 1n . Consider this as 1n−2 12 . That is, group the first n − 2 1s together and the
last two 1s together. From the induction hypothesis, there is a Hamiltonian path through the partitions
of n − 2 starting with 1n−2 . By adjoining two 1s to each of these, there is a Hamiltonian path through the
partitions of n that have two or more 1s in them. This path ends at (2` )c 12 , where n − 2 = 2(r − 1) = c · 2` ,
c is an odd integer, and ` is positive. The manner in which the path will be continued through the parti-
tions of n with no 1s in them is dependent on the parity of r.

First consider the case in which r is even. Then r − 1 is odd. In this case, c = r − 1 and ` = 1.
Thus the path in the previous paragraph ended at 2c 12 . The next step in the path will be to 2c+1 = 2r .
Now by induction, there is a Hamiltonian path through the partitions of n/2 = r starting at 1r and ending
at (2k−1 )a . Double every entry of each of these partitions. This gives a Hamiltonian path through the
partitions of n that have no 1s in them, starting at 2r and ending at (2k )a . Attaching this to the end of
the path from 1n to 2r already obtained results in the desired Hamiltonian path from 1n to (2k )a .

In the case where r is odd, r − 1 is even. In this case, a = r and k = 1, while n − 2 = 2(r − 1)
must be a multiple of 4. So the path through the partitions of n containing 1s ended at (2` )c 12 for some
` ≥ 2. The next step will be to the partition (2` )c 2. Now r = n/2 is odd, so every binary partition of r
must contain at least one 1. Note that these are the binary partitions of r − 1 with a 1 adjoined, and by
induction, there is a Hamiltonian path through them starting at 1r and ending at (2`−1 )c 1. Double each
entry of each of these, and run through this path backward to obtain a path starting at (2` )c 2 and ending
at 2r . Concatenating this with the path from 1n to (2` )c 2 previously obtained will yield the desired path
from 1n to 2r = (2k )a .
7. Proceed by induction on L. The only element for which L = 0 is the smallest binary partition of n, which
indeed has n 1s by the solution to Problem 5, and so this partition has n − L = n − 0 parts. Moving up via a
covering relation, when two 2k s are replaced by one 2k+1 , it is true that one nonzero part is lost. Assuming an
element at level L has n − L nonzero parts, an element at level L + 1 will have n − L − 1 = n − (L + 1) parts.
This establishes the inductive step.
8. a. There are several s-partitions of 7. They include (7, 0, 0, 0, 0, 0, 0), (3, 3, 1, 0, 0, 0, 0), (3, 1, 1, 1, 1, 0, 0), and
(1, 1, 1, 1, 1, 1, 1). Other partitions of 7 are not s-partitions; some of those are also not binary. They
include (6, 1, 0, 0, 0, 0, 0), (5, 2, 0, 0, 0, 0, 0), (5, 1, 1, 0, 0, 0, 0), (3, 2, 2, 0, 0, 0, 0), and (3, 2, 1, 1, 0, 0, 0).
b. Construct a bijection. Given a binary partition of n at level L, subtract 1 from each nonzero part. Hence
each part 2j is replaced by 2j − 1, leaving an s-partition of n − (n − L) = L.

For the inverse map, consider an s-partition of L, with parts of the form 2j − 1 for some positive integer j.
This cannot have more than L parts, and that is achievable only with all parts equal to 1. Consider the
partition to end in enough 0s to bring the total number of parts up to n − L; this is possible because
n − L = n − 2L + L ≥ 2L − 2L + L = L. Now add 1 to each part. The result is a binary partition of
L + n − L = n with n − L parts, which puts it at level L by Problem 7.
9. A Hasse diagram for the trinary partitions of 12 is shown below.

16
10. The three possible answers are 18 or 19 or 20. A Hasse diagram for the trinary partitions of 18 is shown
below. The answer will be established when it is shown that there is no Hamiltonian path through the Hasse
diagram. Note that all Hasse diagrams for the trinary partitions of 18 are isomorphic to the one in the figure.
Note also that the Hasse diagram for the trinary partitions of 19 is isomorphic to the Hasse diagram for the
trinary partitions of 18; a bijection can be established by adding a 1 after the rightmost nonzero part of any
trinary partition of 18. Similarly, the Hasse diagram for the trinary partitions of 20 is isomorphic to the Hasse
diagram for the trinary partitions of 18.

Suppose there is a Hamiltonian path through the Hasse diagram. Because the Hasse diagram has only one
edge coming from 118 and only one edge coming from 92 , the Hamiltonian path must begin and end at those
vertices in some order. Without loss of generality, assume the path begins at 118 and ends at 92 .

The path must be of the form 118 , 3115 , 32 112 , 33 19 , . . . , 92 . The first choice comes when deciding if the fifth
vertex in the path is 919 or 34 16 . If the choice is to make 34 16 the fifth vertex in the path, then it is not
possible to enter and leave 919 because the edge {33 19 , 919 } cannot be used for entry or for exit; both would
require revisiting 33 19 . So the fifth vertex in the path must be 919 , and thus the sixth vertex in the path is 9316 .

The next choice comes when considering the seventh vertex in the path. If the seventh vertex is 932 13 ,
then it will be impossible to visit 34 16 and also 92 . If the seventh vertex is 34 16 , then the eighth vertex in
the path is 35 13 , and either choice from that vertex (932 13 or 36 ) leads to a contradiction because to visit the

17
other requires a path that prevents visiting 92 . This completes the proof that the Hasse diagram of the trinary
partitions of 18 does not contain a Hamiltonian path.

Checking other possible values of n that are multiples of 3, there is a Hamiltonian path through the Hasse
diagrams for each one, as the following argument demonstrates.
• A Hamiltonian path through the Hasse diagram for n = 3 is 13 , 3.
• A Hamiltonian path through the Hasse diagram for n = 6 is 16 , 313 , 32 .
• A Hamiltonian path through the Hasse diagram for n = 9 is 19 , 316 , 32 13 , 33 , 9.
• A Hamiltonian path through the Hasse diagram for n = 12 is 112 , 319 , 32 16 , 33 13 , 913 , 93, 34 .
• A Hamiltonian path through the Hasse diagram for n = 15 is 115 , 3112 , 32 19 , 33 16 , 916 , 9313 , 932 , 35 , 34 13 .
• A Hamiltonian path through the Hasse diagram for n = 21 is 121 , 3118 , 32 115 , 33 112 , 9112 , 9319 , 34 19 , 35 16 ,
932 16 , 933 13 , 934 , 92 3, 92 13 .
Note that there are Hamiltonian paths through the Hasse diagram for both n = 1 (the path is just the vertex
1) and n = 2 (the path is 12 , 2). Note also that if k is a positive integer, the Hasse diagrams for the trinary
partitions of n = 3k, n = 3k + 1, and n = 3k + 2 are all isomorphic, in the same way that the Hasse diagrams
for the trinary partitions of 18, 19, and 20 are isomorphic.

Thus the only values of n less than 23 for which the Hasse diagram of trinary partitions does not contain
a Hamiltonian path are 18 or 19 or 20.

18
6 Individual Problems

Problem 1. Compute the greatest prime factor of N , where

N = 2018 · 517 + 517 · 2812 + 2812 · 666 + 666 · 2018.

Problem 2. Compute the number of ordered pairs of positive integers (a, b) that satisfy
a2 b3 = 2018 .

Problem 3. Compute the number of positive three-digit multiples of 3 whose digits are distinct and nonzero.

2 2
Problem 4. In 4ABC, cos(A) = and cos(B) = . Given that the perimeter of 4ABC is 24, compute the area
3 7
of 4ABC.

Problem 5. Compute the product of all positive values of x that satisfy

bx + 1c2x − 19bx + 1cx + 48 = 0.

Problem 6. Triangle ABC is inscribed in circle ω. The line containing the median from A meets ω again at M ,
the line containing the angle bisector of ∠B meets ω again at R, and the line containing the altitude from C
meets ω again at L. Given that quadrilateral ARM L is a rectangle, compute the degree measure of ∠BAC.

Problem 7. Let S be the set of points (x, y) whose coordinates satisfy x2 + y 2 ≤ 36 and (max{x, y})2 ≤ 27.
Compute the perimeter of S.

Problem 8. Rectangle P QRS is drawn inside square ABCD, as shown. Given that [AP S] = 20 and [CQR] = 18,
compute [ABCD].

Problem 9. Compute the least positive four-digit integer N for which N and N +2018 contain a total of 8 distinct
digits.

Problem 10. Compute the least positive value of t such that

Arcsin(sin(α)), Arcsin(sin(2α)), Arcsin(sin(7α)), Arcsin(sin(tα))
π
is a geometric progression for some α with 0 < α < 2.

19
7 Answers to Individual Problems

Answer 1. 23

Answer 2. 28

Answer 3. 180
√
Answer 4. 12 5

Answer 5. 4

Answer 6. 36 (or 36◦)

Answer 7. 8π + 12 (or 12 + 8π)

Answer 8. 243

Answer 9. 1489
√ √
Answer 10. 9 − 4 5 (or −4 5 + 9)

20
8 Solutions to Individual Problems

Problem 1. Compute the greatest prime factor of N , where

N = 2018 · 517 + 517 · 2812 + 2812 · 666 + 666 · 2018.

Solution 1. Factor N as follows:

N = 517 · (2018 + 2812) + 666 · (2812 + 2018)

= (517 + 666) · (2018 + 2812)
= 1183 · 4830
= (7 · 169) · (2 · 5 · 3 · 161)
= (7 · 132 ) · (2 · 3 · 5 · 7 · 23).

Each of the integers 2, 3, 5, 7, 13, and 23 is prime. Thus the largest prime factor of N is 23.

Problem 2. Compute the number of ordered pairs of positive integers (a, b) that satisfy

a2 b3 = 2018 .

Solution 2. The prime factorization of 2018 is 236 · 518 , so a and b may only have prime factors of 2 and 5. Let
36 18
a = 2x · 5y . Then b3 = 2 a·52 = 236−2x · 518−2y . The value of b will be an integer if and only if both 36 − 2x
and 18 − 2y are nonnegative multiples of 3. Therefore x and y must both be nonnegative multiples of 3. This
means that x ∈ {0, 3, 6, 9, 12, 15, 18} and y ∈ {0, 3, 6, 9}, so there are a total of 7 · 4 = 28 solution pairs.

Problem 3. Compute the number of positive three-digit multiples of 3 whose digits are distinct and nonzero.

Solution 3. If a number is divisible by 3, then the sum of its digits is divisible by 3. The digits must be equivalent
to one of the following (modulo 3): 0, 0, 0; 1, 1, 1; 2, 2, 2; or 0, 1, 2. Note that there are 3 possible nonzero
digits congruent to each of 0, 1, and 2 (modulo 3). Therefore the number of ways to choose the three digits
(without regard to their order) is
Ç å Ç å Ç å Ç å
3 3 3 3
3· + · · = 3 · 1 + 3 · 3 · 3 = 30.
3 1 1 1

Having chosen the three digits, consider the number of ways the three digits can be ordered. There are 3! = 6
ways to order the three digits, and this gives a total of 30 · 6 = 180 three-digit numbers.

2 2
Problem 4. In 4ABC, cos(A) = and cos(B) = . Given that the perimeter of 4ABC is 24, compute the area
3 7
of 4ABC.

√
sin A 5/3
Solution 4. Let D be the foot of the altitude from C. By the Law of Sines, a
b = sin B = √
3 5/7
= 79 . Thus b = 9a
7 ,
AD = b cos A = 6a 2a
24, solve a+ 9a
7 , and BD = a cos B = 7 . Because the perimeter of 4ABC is√
6a
7 +√
2a
7 + 7 = 24
to obtain a = 7, which implies b = 9 and c = 8. By Heron’s Formula, [ABC] = 12 · 5 · 3 · 4 = 12 5.

21
Alternatively, drop altitude CD as before, and then apply the definition of cosine for ∠CAD in 4CAD and
for ∠CBD in 4CBD. Thus AD = 2x and AC = 3x for some real x and DB = 2y and CB = 7y for some
real y. By the Pythagorean Theorem, (CD)2 = 9x2 − 4x2 = 5x2 and (CD)2 = 49y 2 − 4y 2 = 45y 2 . This
implies 5x2 = 45y 2 so x = 3y. Substituting, AD = p6y and AC√ = 9y. √ 4ABC is√24,
Because the perimeter of √
9y+7y+6y+2y = 24 → y = 1, so AB = 8 and CD = 45y 2 = 45 = 3 5. Thus [ABC] = 21 ·8·(3 5) = 12 5.

Problem 5. Compute the product of all positive values of x that satisfy

bx + 1c2x − 19bx + 1cx + 48 = 0.

Solution 5. Introduce the substitution y = bx + 1cx . Then the given equation becomes y 2 − 19y + 48 = 0, and
the left-hand side of the equation in y factors as (y − 3)(y − 16) = 0. Therefore bx + 1cx = 3 or bx + 1cx = 16.

Case 1: y = bx + 1cx = 3. Note that if 0 ≤ x ≤ 1, then bx + 1cx ≤ 2 and if x ≥ 2, then bx + 1cx ≥ 9. Therefore
1 < x < 2, and hence bx + 1c = 2. The equation 2x = 3 has the solution x = log2 3, and indeed, 1 < log2 3 < 2,
which checks.

Case 2: y = bx + 1cx = 16. Note that if 0 ≤ x ≤ 2, then bx + 1cx ≤ 9 and if x ≥ 3, then bx + 1cx ≥ 64.
Therefore 2 < x < 3, and hence bx + 1c = 3. The equation 3x = 16 has the solution x = log3 16, and indeed,
2 < log3 16 < 3, which checks.

Use the change of base rule for logarithms to obtain the simplified product of the solutions:
log 3 log 16 log 16
(log2 3)(log3 16) = · = = log2 16 = 4.
log 2 log 3 log 2

Solution 6. Because ARM L is a rectangle inscribed in a circle, its diagonal AM is a diameter of ω. The point B
←−→
is on the circle and in one half-plane bounded by AM . Because AM passes through the midpoint of side BC,
←−→
the point C is on the circle and in the opposite half-plane bounded by AM as B. Also, C must be the same
distance from AM as B. Therefore either C is the reflection of B across line AM or C is the reflection of B
across O, the center of ω, as shown below.

22
M M

B C B

O O

A A

For ARM L to be a rectangle, it is sufficient that LR be a diameter of ω or, equivalently, that m∠LM R = 90◦ .
So consider the two cases for C, above, and determine if in either case it is possible to have m∠LM R = 90◦ .

In the first case, 4ABC is isosceles with AB = AC. Let β = m∠ABC = m∠ACB. Then m∠AM R =
m∠ABR = β2 and m∠AM L = m∠ACL = 180◦ − 90◦ − m∠BAC = 2β − 90◦ . Setting β2 + 2β − 90◦ = 90◦ gives
β = 72◦ . Then m∠BAC = 180◦ − 2β = 36◦ .
M
B C

L
O

In the second case, BC is a diameter of ω, so m∠BAC = 90◦ . This means that L is coincident with A, which
leads to a degenerate result. Thus the only possible value for m∠BAC is 36◦ .

Alternate Solution: Place the triangle in the complex plane so that ω coincides with the unit circle, and
a = 1. (Here, a lowercase letter represents the complex coordinates of the point denoted by the corresponding
uppercase letter.) Then m = −1 because diagonal AM of rectangle ARM L is a diameter of ω. Because the
midpoint of BC has coordinates b+c 2 and because it lies on AM (which coincides with the real axis), it follows
that b + c ∈ R → b + c = 1b + 1c → bc(b + c) = b + c, and thus either bc = 1 or b + c = 0. The latter cannot
be true, as it implies that ∠BAC is a right angle, and thus ` = a, r = m = −1, and then c = a, which
results in a degenerate triangle. Hence bc = 1. Now, using some identities about the unit circle in the complex
plane (which are straightforward to prove), conclude that r2 = ac (due to arc bisection) and c` = −ab (from
perpendicular chords). Expressing everything in terms of r:

r2 1 ab
c= = r2 , b= = r−2 , and ` = − = −r−4 .
a c c
Because ARM L is a rectangle, its other diagonal, RL, must also be a diameter of ω. Then r = −` = r−4 →
r5 = 1, and so r is a fifth root of unity. It is impossible for r to be 1, as this would cause triangle ABC to
be degenerate. Also, by replacing r with r, note that the entire diagram is reflected across the real axis, and
2πi 4πi
the measure of ∠BAC does not change. Thus there are only two cases to consider: r = e 5 and r = e 5 .
4πi 8πi 2πi
Suppose that r = e 5 . Then c = e 5 and b = e 5 . But then B and L would lie on the same side of AC, and
2πi 6πi 4πi
BL would be the external angle bisector of ∠B. Thus r = e 5 , which implies b = e 5 and c = e 5 . Then
minor arc BC measures 2π ◦
5 radians, which is 72 , and so the degree measure of ∠BAC is half of that, or 36 .
◦

Problem 7. Let S be the set of points (x, y) whose coordinates satisfy x2 + y 2 ≤ 36 and (max{x, y})2 ≤ 27.
Compute the perimeter of S.

23
2
Solution 7. Let C be the region defined by x2 + y 2 ≤ 36, and let R be the
√ region defined by (max{x, y}) ≤ 27,
so S = C ∩ R. If (x, y)
√ ∈ R, then
√ neither x nor y can be greater than 3 3, and at least one of the coordinates
must be between −3 3 and 3 3. Thus if
¶ √ √ ©
R1 = (x, y) : x ≤ 3 3, |y| ≤ 3 3 and
¶ √ √ ©
R2 = (x, y) : |x| ≤ 3 3, y ≤ 3 3 ,

then R = R1 ∪ R2 . Consider the diagram below.

Most of√circle C lies√inside the central square R1 ∩ R2 , except for the segment cut off by the chord from
A(−3, 3 3) to B(3, 3 3), and three other identical segments, none of which overlap. The segment on the left
side of C lies inside R1 , and the bottom segment lies inside R2 . Thus S consists of the circle C with two such
segments removed. If O is the origin, then 4ABO is equilateral with side length 6, so the arc of the circle cut
off by chord AB measures 60◦ . Thus removing each segment replaces a 60◦ arc of the circle with a chord of
length 6, so the total perimeter of S, the region remaining, is
1
2π · 6 − 2 · (2π · 6) + 2 · 6 = 8π + 12.
6

Problem 8. Rectangle P QRS is drawn inside square ABCD, as shown. Given that [AP S] = 20 and [CQR] = 18,
compute [ABCD].

24
Solution 8. Notice first that triangles ABQ, QCR, SP A, and RDS are all similar. Let P S = x and m∠ASP = α,
as shown below.

2 2
Then [AP S] = P S·AP
2 = x·x tan
2
α
= x tan
2
α
and [CQR] = CQ·CR
2 = x cos α·x cos
2
α tan α
= x tan
2
α
· cos2 α. Hence
18 2 3 1 y 2y 2y
20 = cos α → cos α =
√ . Thus tan α = . Now let AB = y. Then BQ = , so CQ =
10 3 3 3 and CR = 9 .
2y 2
Thus [CQR] = 27 = 18, and so [ABCD] = y 2 = 243.

AS 2 20
Alternate Solution: Because [AP S] = 20, [RCQ] = 18, and 4AP S ∼ 4RCQ, conclude that = , so
RQ2 18
√ √ Ç√ å2
10 10 2 10 x2
AS = · RQ. Let RQ = x. Then RQ = P S = x, thus AS = x, giving AP = x − x2 = ,
3 3 3 9
x 1 x √ 3 √ √
so AP = . This implies [AP S] = · · x = 20, so x = 2 30. Then QC = √ · 2 30 = 6 3. Now
3 2 3 10
y
let AB = y. Because 4ABQ ∼ 4SP A, it follows that BQ = . From y = BQ + QC, conclude that
√ √ √ 3
y
y = + 6 3 → y = 9 3. Thus the area of the square is (9 3)2 = 243.
3

A S D

x
P

R
x

B C
Q

Problem 9. Compute the least positive four-digit integer N for which N and N +2018 contain a total of 8 distinct
digits.

Solution 9. Let N = A B C D and N + 2018 = E F G H, where A, B, C, D, E, F, G, H represent distinct digits.

To minimize N , first consider A = 1 and B 6= 9. Then 3018 < N + 2018 < 3918, so E = 3. Because B 6= F ,
there must be a carry from the tens place to the hundreds place, and it follows that C = 8 or C = 9, and
B + 1 = F . The least possible value for B is now 4, so assume B = 4 and F = 5. If there is no carry from the
ones place to the tens place, then the only possibility for the ones digits is D = 0 and H = 8, but this requires
C = 9 and G = 0 to achieve the carry into the hundreds place, which is a contradiction. Therefore there is a
carry from the ones to the tens place. Because G 6= 1, it follows that C = 8 and G = 0. Finally, because D + 8

25
is the two-digit number 1 H, it must be that D = H + 2, and because H 6= 0, 1, 3, 4, 5 and D 6= 4, 8, the only
possible values remaining are D = 9 and H = 7. Thus the least possible value of N is 1489.

Problem 10. Compute the least positive value of t such that

Arcsin(sin(α)), Arcsin(sin(2α)), Arcsin(sin(7α)), Arcsin(sin(tα))
π
is a geometric progression for some α with 0 < α < 2.

Solution 10. Let the ratio of consecutive terms in the sequence be r. If there exists an α for which the first three
terms form a geometric progression with ratio r, the least t for which Arcsin sin(tα) continues the sequence will
be t = r3 . This solution will focus on finding the least viable ratio r.

Note that as 2α < π, the first two terms of the progression are both positive, so the ratio r must also be
positive. The graphs of y = Arcsin(sin 2x) and y = Arcsin(sin 7x) are both piecewise linear, the latter of which
is positive only for (0, π7 ) ∪ ( 2π 3π
7 , 7 ), as seen below.

Thus consider only values of α in (0, π7 ) ∪ ( 2π 3π

7 , 7 ), because all terms in the progression must be positive. The
function Arcsin(sin 7x) is piecewise composed of four line segments in that domain. So, search along these
segments to see if there exists an α for which a geometric progression results.

2α 7α
Along the first segment, no α exists, because the ratios between the first three terms are α = 2 6= 2α .

Along the second segment, Arcsin(sin 7α) = π − 7α, and the ratio equality becomes 2α π−7α π
α = 2α → α = 11 . In
this case, r = 2, and because r3 α = 8π
11 > π
2 , the geometric progression gets too large to be within the range of
the Arcsin function.

π−2α
Along the third segment, the three functions are equal to x, π − 2x, 7x − 2π, so the ratio equality is α =
7α−2π π 3
π−2α → α = 3 . In this case, r = 1 and the least t is r = 1.

Finally, along the fourth segment,√

the three functions are equal to x, π − 2x, 3π − 7x, and the ratio equal-
π−2α 3π−7α (7± 5)π
ity is α = π−2α → α = 22 . One can check that the smaller value of α is extraneous while the larger,
√ √ √
(7+ 5)π
α= π−2α
, is not, and so r = α = α π
− 2 = 3−2 5 , which gives r3 = 9 − 4 5, which is positive because
√ 22 √ √ √
9 − 4 5 = 81 − 80 > 0. This is the least value of t among all cases, so t = 9 − 4 5.

26
9 Relay Problems

Relay 1-1. The number N = 327763 can be expressed as

N = 513 + 583 = 303 + 673 .

Compute the least prime factor of N .

Relay 1-2. Let T = T N Y W R. Rectangle ABCD has area T + 5. Point M is the midpoint of AB and point N is
a trisection point on CD. The segment M N divides rectangle ABCD into two trapezoids. Compute the area
of the larger of these two trapezoids.

√
4
√
Relay 1-3. Let T = T N Y W R. Define the sequence a1 , a2 , a3 , . . . by a1 = 2, a2 = 2, and for n ≥ 3, an =
bT c
an−1 an−2 . Compute the least value of k such that ak is an integer multiple of 2 .

Relay 2-1. Given that A and B are integers such that 3A + 2B = 218 and |A − B| is as small as possible, compute
A + B.

Relay 2-2. Let T = T N Y W R. Compute the number of lattice points in the interior of the triangle with vertices
(0, T ), (T, 0), and (0, 0).

Relay 2-3. Let T = T N Y W R. The integers a and b are chosen from the set {1, 2, 3, . . . , T }, with replacement.
Compute the number of ordered pairs (a, b) for which

a + b = 4036.

27
10 Relay Answers

Answer 1-1. 31

Answer 1-2. 21

Answer 1-3. 11

Answer 2-1. 87

Answer 2-2. 3655

Answer 2-3. 3275

28
11 Relay Solutions

Relay 1-1. The number N = 327763 can be expressed as

N = 513 + 583 = 303 + 673 .

Compute the least prime factor of N .

Solution 1-1. Note that the factorization a3 + b3 = (a + b)(a2 − ab + b2 ) implies that a + b is a divisor of a3 + b3 .
Thus 51 + 58 = 109 and 30 + 67 = 97 are each factors of N . From long division, N = 31 · 97 · 109, and each of
the three factors in the product is prime, hence the answer is 31.

Alternate Solution: Note that N = 327 · 1000 + 763 and that 327 = 3 · 109 and 763 = 7 · 109. Thus
N = 109(3 · 1000 + 7). As in the first solution, conclude that 97 must be a factor of 3007 and, using long
division, the other factor is 31.

Note: Numbers that can be represented as the sum of two positive cubes in two different ways are exam-
ples of taxicab numbers. The least taxicab number is 1729 (1729 = 103 + 93 = 123 + 13 ). The origin of the
name “taxicab number” comes from an anecdote involving a visit between mathematicians G. H. Hardy and S.
Ramanujan. Hardy related that the number of the taxi he took to visit Ramanujan was 1729, and he remarked
that the number seemed to be rather dull. Ramanujan countered by stating the fact discussed above.

Solution 1-2. Let AB = CD = 6x and BC = AD = y. The smaller trapezoid has bases of lengths 2x and 3x, so
its area is 52 xy. The larger trapezoid has bases of lengths 3x and 4x, so its area is 27 xy. Because 6xy = T + 5,
7
the area of the larger trapezoid is 12 (T + 5). With T = 31, the desired area is 21.

Solution 1-3. Make a table consisting of the first few values of the sequence a1 , a2 , a3 , . . . .

n 1 2 3 4 5 6
an 21/4 22/4 23/4 25/4 28/4 213/4

Note that an is of the general form 2Fn+1 /4 , where Fn denotes the nth Fibonacci number. The desired value
of k has the property that Fk+1 is a multiple of 4 and is the least such value satisfying Fk+1 ≥ 4bT c. With
T = 21, the least Fibonacci number that is both a multiple of 4 and which is greater than or equal to 4 · 21 = 84
is F12 = 144, hence k = 11.

Relay 2-1. Given that A and B are integers such that 3A + 2B = 218 and |A − B| is as small as possible, compute
A + B.

Solution 2-1. Note that |A − B| 6= 0 because with A = B, 3A + 2B is a multiple of 5, but 218 is not a multiple of
5. Suppose there exist integers A and B such that |A − B| = 1. If B = A + 1, then the given equation becomes
5A + 2 = 218, but this does not result in an integral value for A. On the other hand, if A = B + 1, then the
given equation becomes 5B + 3 = 218, and this yields the solution (A, B) = (44, 43), hence A + B = 87.

29
Relay 2-2. Let T = T N Y W R. Compute the number of lattice points in the interior of the triangle with vertices
(0, T ), (T, 0), and (0, 0).

Solution 2-2. The following solution assumes that T is a positive integer. (As an exercise, the reader might want
to consider how the solution would change if T were a negative integer or if T were not an integer.) It is easy
to verify that if T = 1 or T = 2, then the answer is 0. So suppose that T ≥ 3. For a fixed value of N with
1 ≤ N ≤ T − 2, the points (1, N ), (2, N ), . . . , (T − N − 1, N ) all lie in the interior of the triangle. This is a total
of T − N − 1 points. Summing over the possible values of N yields (T − 2) + (T − 3) + · · · + 1 = (T −2)(T 2
−1)
.
With T = 87, the answer is 3655.

Alternatively, one can apply Pick’s Theorem, which states that for a polygon whose vertices are all lattice points,
K = I + B2 − 1, where K is the area of the polygon, I is the number of lattice points in its interior, and B is the
number of lattice points on its boundary. For the given triangle, K = 21 T 2 and B = (T + 1) + T + (T − 1) = 3T .
2
Hence I = T −3T 2
+2
, as established above.

Relay 2-3. Let T = T N Y W R. The integers a and b are chosen from the set {1, 2, 3, . . . , T }, with replacement.
Compute the number of ordered pairs (a, b) for which

a + b = 4036.

Solution 2-3. The answer is the number of ordered pairs of integers (x, 4036 − x) for which x and 4036 − x are both
in the set {1, 2, 3, . . . , T }. If T < 4036
2 = 2018, then the answer is 0 because the greatest possible sum would be
achieved with the ordered pair (T, T ), but 2T < 4036. If 2018 ≤ T < 4036, the ordered pairs are (4036 − T, T ),
(4037 − T, T − 1), . . . , (T, 4036 − T ); there are T − (4036 − T ) + 1 = 2T − 4035 of these ordered pairs. Finally,
if T ≥ 4036, then the answer is 4035 because each of the ordered pairs (1, 4035), (2, 4034), . . . , (4035, 1) has
the sum of its coordinates equal to 4036. Because T = 3655, the desired number of ordered pairs is therefore
2 · 3655 − 4035 = 3275.

30
12 Super Relay
1. Given that a, b, and c are positive integers such that a + bc = 20 and a + b = 18, compute the least possible
value of abc.
2. Let T = T N Y W R. In convex quadrilateral LEOS, LE = EO, LS = SO, and ES and LO intersect in point
J. Given that EJ = 24, JS = 60, and EO = T , compute LS.
3. Let T = T N Y W R. For each integer n, let f (n) be the remainder when n2 − 1 is divided by 8. Compute
f (T ) + f (T + 1) + f (T + 2) + f (T + 3).
4. Let T = T N Y W R. Compute the shortest distance between the lines y = x + T and y = x − T .
√ √
5. Let T = T N Y W R. Compute the value of x for which x + x + 80 = T .
6. Let T = T N Y W R. Jane rolls a fair T -sided die whose faces are numbered from 1 to T inclusive. The
probability that she rolls a multiple of either 2 or 3 is the same as the probability that she rolls a multiple of
either 4, 5, 7, p, q, or r, where p, q, and r are prime numbers and 7 < p < q < r. Compute r.
7. Let T = T N Y W R, and let K = T + 3. Compute log4 21 + log4 22 + log4 23 + · · · + log4 2K−1 + log4 2K .

15. Compute 20183 − 3 · 20182 · 2017 + 3 · 2018 · 20172 − 20173 .

1 1
14. Let T = T N Y W R. In 4DES, sin D = and sin E = . Given that DS = 18, compute ES.
T 20
13. Let T = T N Y W R. A group of T + 60 students compete at an ARML site. At that site, a total of 50 students
got at least one of individual questions 9 and 10 correct, 48 students got question 9 correct, and an astonishing
22 students got question 10 correct! Compute the probability that a randomly selected student from this group
got both questions 9 and 10 correct.
1
12. Let T = T N Y W R, and let K = . The complex numbers a and b are the solutions to the equation
T
x + 20x − K = 0, and the complex numbers c and d are the solutions to the equation x2 − 18x + K − 4 = 0.
2

Compute a2 + b2 + c2 + d2 .
11. Let T = T N Y W R, and let P be the greatest prime factor of T . Given that a sphere has a surface area of
(P + 3)π, compute the radius of the sphere.
10. Let T = T N Y W R. Given that f (x) = −x2 + T x + c for some real number c and f (20) = 18, compute the
maximum value of f (x), where x ranges over the real numbers.
9. Let T = T N Y W R. In the increasing arithmetic sequence a1 , a2 , a3 , . . ., the common difference between
consecutive terms is 4, and a1 = T . Compute the value of n for which an = 2018.

8. Let A be the number you will receive from position 7 and let B be the number you will receive from position 9.
A circular track has a perimeter of A meters. Jasmine and Richard start running clockwise around the track at
the same time and from the same starting position on the track. Jasmine runs at a constant speed and takes
16 seconds to run A meters. Richard also runs at a constant speed and takes 160 seconds to run B meters.
When Jasmine and Richard first meet again at the starting point, Jasmine will have run m laps and Richard
will have run n laps. Compute 20m + 18n.

31
13 Super Relay Answers

1. 51
2. 75
3. 10
√
4. 10 2
5. 18
6. 17
7. 105

15. 1
14. 360
1
13.
21
12. 732
11. 4
10. 342
9. 420

8. 136

32
14 Super Relay Solutions

Problem 1. Given that a, b, and c are positive integers such that a + bc = 20 and a + b = 18, compute the least
possible value of abc.

Solution 1. Subtract the second equation from the first equation and factor to obtain b(c − 1) = 2. Hence either
(b, c) = (2, 2) or (b, c) = (1, 3). In the former case, a = 18 − 2 = 16, and in the latter case, a = 18 − 1 = 17.
The value of abc is thus minimized when a = 17, b = 1, and c = 3, and abc = 51.

Problem 2. Let T = T N Y W R. In convex quadrilateral LEOS, LE = EO, LS = SO, and ES and LO intersect
in point J. Given that EJ = 24, JS = 60, and EO = T , compute LS.

Solution 2. Note that quadrilateral LEOS is a kite, hence ES ⊥ LO. By the Pythagorean
√ Theorem, JO2 =
2 2 2 2 2 2 2 2 2 2 2
T − 24 and LS = OS = JO + JS = T − 24 + 60 = T + 3024, hence LS = T + 3024. With T = 51,
LS = 75. (Note: The calculation can be simplified by noting that once a value of T is received, 4EJO is
similar to an 8−15−17 triangle, and 4OJS is similar to a 3−4−5 triangle.)

Problem 3. Let T = T N Y W R. For each integer n, let f (n) be the remainder when n2 − 1 is divided by 8.
Compute f (T ) + f (T + 1) + f (T + 2) + f (T + 3).

Solution 3. Note that exactly two elements of {T, T + 1, T + 2, T + 3} are even, and the other two elements of the
set are odd. Similarly, exactly two elements of {T 2 − 1, (T + 1)2 − 1, (T + 2)2 − 1, (T + 3)2 − 1} are even, and
the other two elements of the set are odd. Without loss of generality, assume that T = 4k, for some integer k.
Then f (T ) = 16k 2 − 1 ≡ 7 (mod 8), f (T + 1) = (4k)(4k + 2) ≡ 0 (mod 8), f (T + 2) = 16k 2 + 8k + 3 ≡ 3 (mod 8),
and f (T + 3) = (4k + 2)(4k + 4) ≡ 0 (mod 8). Hence the answer is 7 + 0 + 3 + 0 = 10 (independent of T ).

Problem 4. Let T = T N Y W R. Compute the shortest distance between the lines y = x + T and y = x − T .

Solution 4. Let A = (0, T ), B = (0, −T ), and let C be the foot of the perpendicular from A to the line y = x − T .
Note that the two lines are parallel and, because their slopes are 1, they make an angle of inclination of 45◦
with the x-axis. Hence
√ m∠ABC = m∠BAC = 45◦ . Because AB √ = |2T |, the distance between the two lines,
AC, is equal to |T 2|. With T = 10, the answer is therefore 10 2.

√ √
Problem 5. Let T = T N Y W R. Compute the value of x for which x+ x + 80 = T .

√ √ √
Solution 5. Rewrite the given equation as xä+ 80 = T − x and square each side to obtain x+80 = T 2 −2T x+x.
√ Ä 2
−80
2 √
Thus 2T x = T 2 − 80, hence x = T 2T . With T = 10 2, the value of x is 18, which checks.

Problem 6. Let T = T N Y W R. Jane rolls a fair T -sided die whose faces are numbered from 1 to T inclusive.
The probability that she rolls a multiple of either 2 or 3 is the same as the probability that she rolls a multiple
of either 4, 5, 7, p, q, or r, where p, q, and r are prime numbers and 7 < p < q < r. Compute r.

Solution 6. Equivalently, the number of outcomes that are multiples of 2 or 3 should equal the number of outcomes
that are multiples of 4, 5, 7, p, q, or r. There are bT /2c multiples of 2 and bT /3c multiples of 3. Each of
these counts multiples of 2 and 3, so by the Inclusion-Exclusion Principle, the number of multiples of 2 or 3 is
bT /2c + bT /3c − bT /6c. Similarly, the number of outcomes that are multiples of 4, 5, 7, p, q, or r is bounded
above by bT /4c + bT /5c + bT /7c + bT /pc + bT /qc + bT /rc. Depending on the value of T , this count may need

33
to be adjusted in the event that an outcome is a multiple of at least two of the numbers 4, 5, 7, p, q, and r.
However, if T < 20, this will not happen. With T = 18, the number of multiples of 2 or 3 is 9 + 6 − 3 = 12,
and the number of multiples of 4, 5, 7, p, q, or r is 4 + 3 + 2 + 1 + 1 + 1 = 12. The only possible triple (p, q, r)
is (11, 13, 17), hence r = 17.

Problem 7. Let T = T N Y W R, and let K = T +3. Compute log4 21 +log4 22 +log4 23 +· · ·+log4 2K−1 +log4 2K .

Solution 7. The given expression is equal to

Ä ä
log4 21 · 22 · . . . · 2K−1 · 2K = log4 21+2+···+(K−1)+K

ÅÄ äK(K+1)/2 ã
1/2
= log4 4

= K(K + 1)/4
= (T + 3)(T + 4)/4.

With T = 17, the answer is 105.

Problem 15. Compute 20183 − 3 · 20182 · 2017 + 3 · 2018 · 20172 − 20173 .

Solution 15. Let a = 2018 and b = 2017. Then the given expression equals a3 − 3a2 b + 3ab2 − b3 = (a − b)3 .
Because a − b = 1, the answer is therefore 13 = 1.

1 1
Problem 14. Let T = T N Y W R. In 4DES, sin D = and sin E = . Given that DS = 18, compute ES.
T 20

ES DS sin D 20 360
Solution 14. By the Law of Sines, = , so ES = DS · = 18 · = . Because T = 1, the answer
sin D sin E sin E T T
is 360 (and ∠D is a right angle).

Problem 13. Let T = T N Y W R. A group of T + 60 students compete at an ARML site. At that site, a total of
50 students got at least one of individual questions 9 and 10 correct, 48 students got question 9 correct, and
an astonishing 22 students got question 10 correct! Compute the probability that a randomly selected student
from this group got both questions 9 and 10 correct.

Solution 13. Let N be the number of students who got both questions 9 and 10 correct. By the Inclusion-
20
Exclusion Principle, it follows that 50 = 48 + 22 − N , so N = 20. Thus the desired probability is . With
T + 60
1
T = 360, this gives an answer of 21 .

1
Problem 12. Let T = T N Y W R, and let K = . The complex numbers a and b are the solutions to the equation
T
x2 + 20x − K = 0, and the complex numbers c and d are the solutions to the equation x2 − 18x + K − 4 = 0.
Compute a2 + b2 + c2 + d2 .

Solution 12. By Vieta’s Formulas, a+b = −20, ab = −K, c+d = 18, cd = K −4. Note also that a2 +b2 +c2 +d2 =
(a+b)2 +(c+d)2 −2ab−2cd = (−20)2 +182 −2(−K)−2(K −4) = 400+324+2K −2K +8 = 732 (independent
of K and T ).

34
Problem 11. Let T = T N Y W R, and let P be the greatest prime factor of T . Given that a sphere has a surface
area of (P + 3)π, compute the radius of the sphere.

√
P +3
Solution 11. Let r be the radius of the sphere. Then 4πr2 = (P + 3)π, so r = . With T = 732 = 2 · 32 · 61,
√ 2
64
P = 61, and the answer is 2 = 4.

Problem 10. Let T = T N Y W R. Given that f (x) = −x2 + T x + c for some real number c and f (20) = 18,
compute the maximum value of f (x), where x ranges over the real numbers.

Solution 10. Plug in x = 20 to obtain 18 = f (20) = −400 + 20T + c, hence c = 418 − 20T . Complete the square in
2 2
the quadratic polynomial to obtain f (x) = − x − T2 + T4 + 418 − 20T . When x ranges over the real numbers,
2
f will be maximized when x = T2 because − x − T2 is negative except when x = T2 . Thus the maximum value
2
of f is T4 + 418 − 20T . With T = 4, this yields the answer of 342.

Problem 9. Let T = T N Y W R. In the increasing arithmetic sequence a1 , a2 , a3 , . . ., the common difference

between consecutive terms is 4, and a1 = T . Compute the value of n for which an = 2018.

2022 − T
Solution 9. The desired value of n satisfies the equation T + 4(n − 1) = 2018, thus n = . With T = 342,
4
this yields n = 420.

Problem 8. Let A be the number you will receive from position 7 and let B be the number you will receive from
position 9. A circular track has a perimeter of A meters. Jasmine and Richard start running clockwise around
the track at the same time and from the same starting position on the track. Jasmine runs at a constant speed
and takes 16 seconds to run A meters. Richard also runs at a constant speed and takes 160 seconds to run B
meters. When Jasmine and Richard first meet again at the starting point, Jasmine will have run m laps and
Richard will have run n laps. Compute 20m + 18n.

Solution 8. Note that Jasmine takes 16 seconds to run a lap around the track, whereas Richard takes 160 B ·A
seconds to run a lap around the track. Let ` = lcm(16, 160A/B). Then Jasmine and Richard will first meet
`
again at the starting point after ` seconds have elapsed. At that time, Jasmine will have run a total of 16 laps
`
and Richard will have run a total of 160A/B laps. With A = 105 and B = 420, it follows that ` = 80 and
therefore m = 80 80
16 = 5 and n = 40 = 2. Therefore 20m + 18n = 100 + 36 = 136.

35
15 Tiebreaker Problems

Problem 1. The increasing infinite arithmetic sequence of integers x1 , x2 , x3 , . . . contains the terms 17! and 18!.
Compute the greatest integer X for which X! must also appear in the sequence.

Problem 2. Let S1 = {1, 2, 3, 4}, S2 = {3, 4, 5, 6}, and S3 = {6, 7, 8}. Compute the number of sets H that satisfy
both of the following properties:
• H ⊆ {1, 2, 3, 4, 5, 6, 7, 8};
• each of H ∩ S1 , H ∩ S2 , and H ∩ S3 is non-empty.

Problem 3. Tom writes down the integers from 1 to 100, inclusive, on a chalkboard and then erases every number
that contains a prime digit. Compute the sum of the digits that remain on the chalkboard.

36
16 Tiebreaker Answers

Answer 1. 32

Answer 2. 201

Answer 3. 337

37
17 Tiebreaker Solutions

Problem 1. The increasing infinite arithmetic sequence of integers x1 , x2 , x3 , . . . contains the terms 17! and 18!.
Compute the greatest integer X for which X! must also appear in the sequence.

Solution 1. Because the sequence contains 17! and 18!, its common difference must divide M = 18! − 17! =
17 · 17!. Note that the set of integers appearing in all such sequences forms an arithmetic sequence with
first term 17! and common difference M . The problem is thus equivalent to finding the greatest integer X
such that X! = 18! + kM for some integer k. Divide both sides of the previous equation by 17!, to obtain
X(X − 1)(X − 2) · . . . · 19 · 18 = 18 + 17k, or in other words,

X(X − 1)(X − 2) · . . . · 19 · 18 ≡ 18 (mod 17). (∗)

Note that modulo 17, the left-hand side of (∗) is equal to (X − 17)!, hence the problem reduces to finding the
greatest solution to (X − 17)! ≡ 1 (mod 17). There is no such solution with X > 33, because then (X − 17)! is
divisible by 17. Also, if X = 33, then by Wilson’s Theorem, 16! ≡ −1 (mod 17) because 17 is prime. However,
16! = 15! · 16 ≡ 15! · (−1) ≡ −1 (mod 17), hence 15! ≡ 1 (mod 17). Thus the desired value of X is 32.

Solution 2. There are a total of 28 subsets of {1, 2, 3, 4, 5, 6, 7, 8}. Proceed indirectly and consider counting the
sets H such that at least one of H ∩ S1 , H ∩ S2 , and H ∩ S3 is empty. Let Na be the number of subsets H of
{1, 2, 3, 4, 5, 6, 7, 8} in which H ∩Sa is empty. Define Na,b and Na,b,c similarly. Then, by the Inclusion-Exclusion
Principle, the answer is
28 − N1 + N2 + N3 − (N1,2 + N1,3 + N2,3 ) + N1,2,3 .

Because |S1 | = 4, it follows that N1 = 28−4 = 16. Similarly, N2 = 16 and N3 = 32. There are 6 elements in
S1 ∪ S2 , thus N1,2 = 28−6 = 4. Similarly, N1,3 = 2 and N2,3 = 4. Finally, N1,2,3 = 1 because ∅ is the only
subset of {1, 2, 3, 4, 5, 6, 7, 8} that hasan empty intersection with each of S1 , S2 , and S3 . Thus the answer is
256 − 16 + 16 + 32 − (4 + 2 + 4) + 1 = 201.

Alternate Solution: Define a set H with the desired properties to be a hitting set. Count the number of
hitting sets H according to the possible sizes of the set H.

38
|H| Number of hitting sets H
0, 1 0: A set H with |H| ≤ 1 cannot have a non-empty intersection with each of
the three sets S1 , S2 , and S3 because there is no element common to all three
of these sets.
2 8: Note that |H ∩ S3 | = 1. If 6 ∈ H, then because H ∩ S1 6= ∅, There are
4 possible elements of S1 that could belong to H. If 6 ∈ / H, then there are
2 choices of the element of S3 that could belong to H (7 or 8) and there are
two choices of the element of S1 ∩ S2 that could belong to H (3 or 4). Thus if
6∈ / H, then there are 2 · 2 = 4 possible hitting sets H, for a total of 4 + 4 = 8
hitting sets of size 2.
8

3 3 − 18 = 38: Note that every subset H of {1, 2, 3, 4, 5, 6, 7, 8} with |H| = 3
will have a non-empty intersection with at least two of the sets S1 , S2 , and
S3 . Thereare 4 sets H, each of which has an empty intersection with S1 ; these
are the 43 subsets of {5, 6, 7, 8} of size 3. There are 4 sets H, each of which
has an empty intersection with S2 ; these are the 43 subsets of {1, 2, 7, 8} of

size 3. Finally, there are 10 sets H, each of which has an empty intersection
with S3 ; these are the 53 subsets of {1, 2, 3, 4, 5} of size 3. Thus there are
8

3 − (4 + 4 + 10) = 38 hitting sets of size 3.
8
4 4 − 7 = 63: Note that every subset H of {1, 2, 3, 4, 5, 6, 7, 8} with |H| = 4 will
have a non-empty intersection with at least two of the sets S1 , S2 , and S3 . There
is 1 set H that has an empty intersection with S1 : {5, 6, 7, 8}. There is 1 set
H that has an empty intersection with S2 : {1, 2, 7, 8}. Finally, there are 5 sets
H, each of which has an empty intersection with S3 : {1, 2, 3, 4}, {1, 2, 3, 5},
{1, 2, 4, 5}, {1, 3, 4, 5}, and {2, 3, 4, 5}. Thus there are 84 − (1 + 1 + 5) = 63
hitting sets of size 4.
8

5 5 − 1 = 55: Any set H with |H| = 5 will intersect each of S1 and S2 in at least
one element. The only set of size 5 that does not intersect S3 is {1, 2, 3, 4, 5},
hence there are 85 − 1 = 55 hitting sets of size 5.
8

6 6 = 28: Because each of S1 , S2 , S3 has at least 3 elements, a 6-element hitting
set will have a non-empty intersection with each of S1 , S2 , and S3 .
8

7 7 = 8: Because each of S1 , S2 , S3 has at least 3 elements, a 7-element hitting
set will have a non-empty intersection with each of S1 , S2 , and S3 .
8 1: The only possible set is {1, 2, 3, 4, 5, 6, 7, 8}, which clearly works.

From the above analysis, the total number of hitting sets is 0 + 0 + 8 + 38 + 63 + 55 + 28 + 8 + 1 = 201.

Note: As in the alternate solution, a set H with the property that it has a non-empty intersection with
each set in a collection of sets S1 , S2 , . . . , Sn is known as a hitting set. Intuitively, this name makes sense
because H “hits” (i.e., intersects) each set Si such that |H ∩ Si | ≥ 1. The following general problem is believed
to be computationally hard in the sense that there are no known efficient algorithms for answering it.
Given a collection of n sets S1 , S2 , . . . , Sn and a positive integer k, does there exist a hitting set H
with |H| ≤ k?

Solution 3. Every one- or two-digit integer can be represented as T U where 0 ≤ T, U ≤ 9 (and for one-digit
integers, T = 0). The prime digits are 2, 3, 5, and 7, so Tom will erase the 40 numbers of the form 2 U , 3 U ,
5 U , and 7 U . After that, consider a fixed value of T , not equal to 2, 3, 5, or 7. Tom will erase the numbers

39
T 2, T 3, T 5, and T 7. Among the six numbers remaining with tens digit T , the sum of their units digits is
0 + 1 + 4 + 6 + 8 + 9 = 28, and the sum of their tens digits is 6T . There are six such values of T , so the sum of
all the digits of the remaining one- and two-digit integers is 6 · 28 + 6(0 + 1 + 4 + 6 + 8 + 9) = 12 · 28 = 336. This
sum includes the “number” 00, but the zeros do not affect the sum. Finally, because the number 100 remains,
the answer is 336 + 1 = 337.

40
ARML Competition 2019

George Reuter, Head Writer

Chris Jeuell, Lead Editor
Evan Chen
Paul Dreyer
Edward Early
Zuming Feng
Zachary Franco
Silas Johnson
Winston Luo
Jason Mutford
Andy Niedermaier
Graham Rosby
Eric Wepsic

May 31–June 1, 2019

Problem 3. Compute the least positive value of t such that

Arcsin sin(t) , Arccos cos(t) , Arctan tan(t)

form (in some order) a three-term arithmetic progression with a nonzero common difference.

Problem 4. In non-right triangle ABC, distinct points P , Q, R, and S lie on BC in that order such that
∠BAP ∼= ∠P AQ ∼ = ∠QAR ∼ = ∠RAS ∼ = ∠SAC. Given that the angles of 4ABC are congruent to the angles
of 4AP Q in some order of correspondence, compute m∠B in degrees.

Problem 5. Consider the system of equations

log4 x + log8 (yz) = 2

log4 y + log8 (xz) = 4
log4 z + log8 (xy) = 5.

Given that xyz can be expressed in the form 2k , compute k.

Problem 6. A complex number z is selected uniformly at random such that |z| = 1. Compute the probability
that z and z 2019 both lie in Quadrant II in the complex plane.

Problem 7. Compute the least positive integer n such that the sum of the digits of n is five times the sum of the
digits of (n + 2019).

Problem 8. Compute the greatest real number K for which the graphs of

(|x| − 5)2 + (|y| − 5)2 = K and (x − 1)2 + (y + 1)2 = 37

have exactly two intersection points.

Problem 9. To morph a sequence means to replace two terms a and b with a+1 and b−1 if and only if a+1 < b−1,
and such an operation is referred to as a morph. Compute the least number of morphs needed to transform
the sequence 12 , 22 , 32 , . . . , 102 into an arithmetic progression.

Problem 10. Triangle ABC is inscribed in circle ω. The tangents to ω at B and C meet at point T . The tangent
to ω at A intersects the perpendicular bisector of AT at point P . Given that AB = 14, AC = 30, and BC = 40,
compute [P BC].

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 1
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
2 Answers to Team Problems

Answer 1. 216

Answer 2. 12
3π
Answer 3.
4
45
Answer 4. (or 22 21 or 22.5)
2
66
Answer 5. (or 9 73 )
7
505
Answer 6.
8076
Answer 7. 7986

Answer 8. 29

Answer 9. 56
800
Answer 10. (or 266 23 or 266.6)
3

2 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
3 Solutions to Team Problems

Problem 1. The points (1, 2, 3) and (3, 3, 2) are vertices of a cube. Compute the product of all possible distinct
volumes of the cube.

p √
Solution 1. The distance between points A(1, 2, 3) and B(3, 3, 2) is AB = (3 − 1)2 + (3 − 2)2 + (2 − 3)2 = 6.
Denote by s the side length of the cube. Consider three possibilities.
√
• If AB is an edge of the cube, then AB = s, so one possibility is s1 = 6.
√ √
• If AB is a face diagonal of the cube, then AB = s 2, so another possibility is s2 = 3.
√ √
• If AB is a space diagonal of the cube, then AB = s 3, so the last possibility is s3 = 2.

The answer is then s31 s32 s33 = (s1 s2 s3 )3 = 63 = 216.

Solution 2. The answer is 12. To give an example in which 12 selfies is possible, consider regular octagon
P1 P2 P3 P4 P5 P6 P7 P8 . Each vertex of the octagon represents a student and each of the diagonals and sides of
the octagon represents a pair of students. Construct eight triangles P1 P2 P4 , P2 P3 P5 , P3 P4 P6 , . . . , P8 P1 P3 .
Each of the segments in the forms of Pi Pi+1 , Pi Pi+2 , Pi Pi+3 appears exactly once in these eight triangles. Tak-
ing 8 three-person selfies (namely {P1 , P2 , P4 }, {P2 , P3 , P5 }, . . . , {P8 , P1 , P3 }) and 4 two-person selfies (namely
{P1 , P5 }, {P2 , P6 }, {P3 , P7 }, {P4 , P8 }) gives a total of 12 selfies, completing the desired task.

A diagram of this construction is shown below. Each of the eight triangles is a different color, and each
of the two-person selfies is represented by a dotted diameter.

P1 P8

P2 P7

P3 P6

P4 P5

It remains to show fewer than 12 selfies is impossible. Assume that the students took x three-person selfies
and y two-person selfies. Each three-person selfie counts 3 pairs of student appearances (in a selfie), and each
two-person selfie counts 1 pair of student appearances (in a selfie). Together, these selfies count 3x + y pairs
of student appearances. There are 82 = 28 pairs of student appearances. Hence 3x + y = 28. The number of

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 3
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
selfies is x + y = 28 − 2x, so it is enough to show that x ≤ 8.

Assume for contradiction there are x ≥ 9 three-person selfies; then there are at least 3 · 9 = 27 (individ-
ual) student appearances on these selfies. Because there are 8 students, some student s1 had at least d27/8e
appearances; that is, s1 appeared in at least 4 of these three-person selfies. There are 2 · 4 = 8 (individual)
student appearances other than s1 on these 4 selfies. Because there are only 7 students besides s1 , some other
student s2 had at least d8/7e (individual) appearances on these 4 selfies; that is, s2 appeared (with s1 ) in at
least 2 of these 4 three-person selfies, violating the condition that each pair of the students appears in exactly
one selfie. Thus the answer is 12.

Problem 3. Compute the least positive value of t such that

Arcsin sin(t) , Arccos cos(t) , Arctan tan(t)
form (in some order) a three-term arithmetic progression with a nonzero common difference.

Solution 3. For 0 ≤ t < π/2, all three values are t, so the desired t does not lie in this interval.

For π/2 < t < π,

Arcsin(sin(t)) = π − t ∈ (0, π/2)
Arccos(cos(t)) = t ∈ (π/2, π)
Arctan(tan(t)) = t − π ∈ (−π/2, 0) .
A graph of all three functions is shown below.

π
Arccos(cos(t))
3
4
π ( 43 π, 43 π)

1
2
π

1 ( 43 π, 14 π)
4
π
Arcsin(sin(t))

0 1
π 1
π 3
π π
4 2 4

( 43 π, − 41 π)

Arctan(tan(t))

Thus if the three numbers are to form an arithmetic progression, they should satisfy
t − π < π − t < t.
3π
The three numbers will be in arithmetic progression if and only if t + (t − π) = 2(π − t), which implies t = .
4
3π π π 3π
Note that if t = , the arithmetic progression is − , , , as required.
4 4 4 4

4 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Problem 4. In non-right triangle ABC, distinct points P , Q, R, and S lie on BC in that order such that
∠BAP ∼= ∠P AQ ∼ = ∠QAR ∼ = ∠RAS ∼ = ∠SAC. Given that the angles of 4ABC are congruent to the angles
of 4AP Q in some order of correspondence, compute m∠B in degrees.

Solution 4. Let θ = 51 m∠A. Because m∠P AQ = θ < 5θ = m∠A, it follows that either m∠B = θ or m∠C = θ.
Thus there are two cases to consider.

If m∠C = θ, then it follows that m∠AQP = m∠QAC + m∠ACB = 4θ, and hence m∠B = 4θ. So 4ABC has
angles of measures 5θ, 4θ, θ, and thus θ = 18◦ . However, this implies m∠A = 5θ = 90◦ , which is not the case.
A
θ
θ
θθθ

4θ 4θ θ
B C
P Q R S

If instead m∠B = θ, then it follows that m∠AP Q = m∠BAP + m∠ABP = 2θ, and hence m∠C = 2θ. So
4ABC has angles of measures 5θ, 2θ, θ, and thus θ = 22.5◦ . Hence m∠B = θ = 22.5◦ .
A
θ
θ θ θ θ

B θ 2θ 2θ C
P Q R S

Problem 5. Consider the system of equations

log4 x + log8 (yz) = 2
log4 y + log8 (xz) = 4
log4 z + log8 (xy) = 5.
Given that xyz can be expressed in the form 2k , compute k.

Solution 5. Note that for n > 0, log4 n = log64 n3 and log8 n = log64 n2 . Adding together the three given
equations and using both the preceding facts and properties of logarithms yields
log4 (xyz) + log8 (x2 y 2 z 2 ) = 11
=⇒ log64 (xyz)3 + log64 (xyz)4 = 11
=⇒ log64 (xyz)7 = 11
=⇒ 7 log64 (xyz) = 11.
66
The last equation is equivalent to xyz = 6411/7 = 266/7 , hence the desired value of k is .
7

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 5
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Problem 6. A complex number z is selected uniformly at random such that |z| = 1. Compute the probability
that z and z 2019 both lie in Quadrant II in the complex plane.

Solution 6. For convenience, let α = π/4038. Denote by

0 ≤ θ < 2π = 8076α

the complex argument of z, selected uniformly at random from the interval [0, 2π). Then z itself lies in Quadrant
II if and only if
π
2019α = < θ < π = 4038α.
2
On the other hand, z 2019 has argument 2019θ, and hence it lies in Quadrant II if and only if there is some
integer k with
π
+ 2kπ < 2019θ < π + 2kπ
2
π π
⇐⇒ (4k + 1) · < 2019θ < (4k + 2) ·
2 2
⇐⇒ (4k + 1)α < θ < (4k + 2)α.

Because it is also true that 2019α < θ < 4038α, the set of θ that satisfies the conditions of the problem is the
union of intervals:
(2021α, 2022α) ∪ (2025α, 2026α) ∪ · · · ∪ (4037α, 4038α).
There are 505 such intervals, the j th interval consisting of (4j + 2017)α < θ < (4j + 2018)α. Each interval has
length α, so the sum of the intervals has length 505α. Thus the final answer is
505α 505 505
= = .
2π 2 · 4038 8076

Problem 7. Compute the least positive integer n such that the sum of the digits of n is five times the sum of the
digits of (n + 2019).

Solution 7. Let S(n) denote the sum of the digits of n, so that solving the problem is equivalent to solving
S(n) = 5S(n + 2019). Using the fact that S(n) ≡ n (mod 9) for all n, it follows that

n ≡ 5(n + 2019) ≡ 5(n + 3) (mod 9)

4n ≡ −15 (mod 9)
n ≡ 3 (mod 9).

Then S(n + 2019) ≡ 6 (mod 9). In particular, S(n + 2019) ≥ 6 and S(n) ≥ 5 · 6 = 30. The latter inequality
implies n ≥ 3999, which then gives n + 2019 ≥ 6018. Thus if n + 2019 were a four-digit number, then
S(n + 2019) ≥ 7. Moreover, S(n + 2019) can only be 7, because otherwise, S(n) = 5S(n + 2019) ≥ 40, which is
impossible (if n has four digits, then S(n) can be no greater than 36). So if n + 2019 were a four-digit number,
then S(n + 2019) = 7 and S(n) = 35. But this would imply that the digits of n are 8, 9, 9, 9 in some order,
contradicting the assumption that n + 2019 is a four-digit number. On the other hand, if n + 2019 were a
five-digit number such that S(n + 2019) ≥ 6, then the least such value of n + 2019 is 10005, and indeed, this
works because it corresponds to n = 7986, the least possible value of n.

Problem 8. Compute the greatest real number K for which the graphs of

(|x| − 5)2 + (|y| − 5)2 = K and (x − 1)2 + (y + 1)2 = 37

have exactly two intersection points.

6 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
√
Solution 8. The graph of the second equation is simply the circle of radius 37 centered at (1, −1). The first
graph is more interesting, and its behavior depends on K.
√
• For small values of K, the first equation determines a set of four circles of radius K with centers at
(5, 5), (5, −5), (−5, 5), and (−5, −5). Shown below are versions with K = 1, K = 4, and K = 16.
10 10 10

5 5 5

0 0 0

-5 -5 -5

-10 -10 -10

-10 -5 0 5 10 -10 -5 0 5 10 -10 -5 0 5 10

• However, when K > 25, the graph no longer consists of four circles! As an example, for K = 36, the value
x = 5 gives (|y| − 5)2 = 36; hence |y| = −1 or |y| = 6. The first option is impossible; the graph ends up
“losing” the portions of the upper-right circle that would cross the x- or y-axes compared to the graph
for (x − 5)2 + (y − 5)2 = 36. The graph for K = 36 is shown below.

-5

-10

-10 -5 0 5 10

• As K continues to increase, the “interior” part of the curve continues to shrink, until at K = 50, it simply
comprises the origin, and for K > 50, it does not exist. As examples, the graphs with K = 50 and K = 64
are shown below.

10 10

5 5

0 0

-5 -5

-10 -10

-10 -5 0 5 10 -10 -5 0 5 10

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 7
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
√
Overlay the graph of the circle of radius 37 centered at (1, −1) with the given graphs. When K = 25, this
looks like the following graph.

-5

-10

-10 -5 0 5 10

Note that the two graphs intersect at (0, 5) and (−5, 0), as well as four more points (two points near the
positive x-axis and two points near the negative y-axis). When K is slightly greater than 25, this drops to four
intersection points. The graph for K = 27 is shown below.

-5

-10

-10 -5 0 5 10

Thus for the greatest K for which there are exactly two intersection points, those two intersection points should
be along the positive x- and negative y-axes. If the intersection point on the positive x-axis is at (h, 0), then
(h − 1)2 + (0 + 1)2 = 37 and (h − 5)2 + (0 − 5)2 = K. Thus h = 7 and K = 29.

Solution 9. Call the original sequence of ten squares T = (12 , 22 , . . . , 102 ). A morphed sequence is one that can
be obtained by morphing T a finite number of times.

This solution is divided into three steps. In the first step, a characterization of the possible final morphed
sequences is given. In the second step, a lower bound on the number of steps is given, and in the third step, it
is shown that this bound can be achieved.
Step 1. Note the following.
• The sum of the elements of T is 12 + 22 + · · · + 102 = 385, and morphs are sum-preserving. So any
morphed sequence has sum 385 and a mean of 38.5.
• The sequence T has positive integer terms, and morphs preserve this property. Thus any morphed
sequence has positive integer terms.
• The sequence T is strictly increasing, and morphs preserve this property. Thus any morphed sequence
is strictly increasing.

8 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Now if the morphed sequence is an arithmetic progression, it follows from the above three observations
that it must have the form

(38.5 − 4.5d, 38.5 − 3.5d, . . . , 38.5 + 4.5d)

where d is an odd positive integer satisfying 38.5 − 4.5d > 0. Therefore the only possible values of d are
7, 5, 3, 1; thus there are at most four possibilities for the morphed sequence, shown in the table below.
Denote these four sequences by A, B, C, D.

T 1 4 9 16 25 36 49 64 81 100
d=7: A 7 14 21 28 35 42 49 56 63 70
d=5: B 16 21 26 31 36 41 46 51 56 61
d=3: C 25 28 31 34 37 40 43 46 49 52
d=1: D 34 35 36 37 38 39 40 41 42 43
P10 P10
Step 2. Given any two sequences X = (x1 , . . . , x10 ) and Y = (y1 , . . . , y10 ) with i=1 xi = i=1 yi = 385,
define the taxicab distance
10
X
ρ(X, Y ) = |xi − yi |.
i=1

Observe that if X 0 is a morph of X, then ρ(X 0 , Y ) ≥ ρ(X, Y ) − 2. Therefore the number of morphs
required to transform T into some sequence Z is at least 21 ρ(T, Z). Now
10
1 1X 2
ρ(T, A) = i − 7i = 56
2 2 i=1

and also ρ(T, A) < min (ρ(T, B), ρ(T, C), ρ(T, D)). Thus at least 56 morphs are needed to obtain sequence
A (and more morphs would be required to obtain any of sequences B, C, or D).

Step 3. To conclude, it remains to verify that one can make 56 morphs and arrive from T to A. One of many
possible constructions is given below.

T 1 4 9 16 25 36 49 64 81 100
6 morphs 1 4 9 16 25 42 49 58 81 100
2 morphs 1 4 9 16 27 42 49 56 81 100
8 morphs 1 4 9 16 35 42 49 56 73 100
10 morphs 1 4 9 26 35 42 49 56 63 100
2 morphs 1 4 9 28 35 42 49 56 63 98
12 morphs 1 4 21 28 35 42 49 56 63 86
10 morphs 1 14 21 28 35 42 49 56 63 76
6 morphs 7 14 21 28 35 42 49 56 63 70

Therefore the least number of morphs needed to transform T into an arithmetic progression is 56.

Remark: For step 3, one may prove more generally that any sequence of 56 morphs works as long as both of
the following conditions hold:
• each morph increases one of the first six elements and decreases one of the last three elements, and
• at all times, the ith term is at most 7i for i ≤ 6, and at least 7i for i ≥ 8.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 9
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Solution 10. To begin, denote by R the radius of ω. The semiperimeter of triangle ABC is 42, and then applying
Heron’s formula yields
14 · 30 · 40 √
[ABC] = = 42 · 28 · 12 · 2 = 168
4R
14 · 30 · 40
from which it follows that R = = 25.
4 · 168
Now consider the point circle with radius zero centered at T in tandem with the circle ω. Because P A = P T ,
it follows that P lies on the radical axis of these circles. Moreover, the midpoints of T B and T C lie on this
radical axis as well. Thus P lies on the midline of 4T BC that is parallel to BC.

B M C

To finish, let O denote the center of ω and M the midpoint of BC. √ right triangle T BO with
√ By considering
altitude BM , it follows that M T · M O = MB 2 , but also M O = OB 2 − MB 2 = 252 − 202 = 15, so

MB 2 400 80
MT = = = .
MO 15 3
Thus the distance from P to BC is 12 M T = 40
3 . Finally,

1 40 800
[P BC] = · · BC = .
2 3 3

10 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
4 Power Question 2019: Elizabeth’s Escape
Instructions: The power question is worth 50 points; each part’s point value is given in brackets next to the part.
To receive full credit, the presentation must be legible, orderly, clear, and concise. If a problem says “list” or “com-
pute,” you need not justify your answer. If a problem says “determine,” “find,” or “show,” then you must show
your work or explain your reasoning to receive full credit, although such explanations do not have to be lengthy. If a
problem says “justify” or “prove,” then you must prove your answer rigorously. Even if not proved, earlier numbered
items may be used in solutions to later numbered items, but not vice versa. Pages submitted for credit should be
NUMBERED IN CONSECUTIVE ORDER AT THE TOP OF EACH PAGE in what your team considers to be
proper sequential order. PLEASE WRITE ON ONLY ONE SIDE OF THE ANSWER PAPERS. Put the TEAM
NUMBER (not the team name) on the cover sheet used as the first page of the papers submitted. Do not identify
the team in any other way.

Elizabeth is in an “escape room” puzzle. She is in a room with one door which is locked at the start of the puzzle.
The room contains n light switches, each of which is initially off. Each minute, she must flip exactly k different light
switches (to “flip” a switch means to turn it on if it is currently off, and off if it is currently on). At the end of each
minute, if all of the switches are on, then the door unlocks and Elizabeth escapes from the room.

Let E(n, k) be the minimum number of minutes required for Elizabeth to escape, for positive integers n, k with
k ≤ n. For example, E(2, 1) = 2 because Elizabeth cannot escape in one minute (there are two switches and one
must be flipped every minute) but she can escape in two minutes (by flipping Switch 1 in the first minute and Switch
2 in the second minute). Define E(n, k) = ∞ if the puzzle is impossible to solve (that is, if it is impossible to have
all switches on at the end of any minute).

For convenience, assume the n light switches are numbered 1 through n.

1. Compute the following.

a. E(6, 1) [1 pt]
b. E(6, 2) [1 pt]
c. E(7, 3) [1 pt]
d. E(9, 5) [1 pt]
2. Find the following in terms of n.
a. E(n, 2) for positive even integers n [2 pts]
b. E(n, 3) for values of n of the form n = 3a + 2 where a is a positive integer [2 pts]
c. E(n, n − 2) for n ≥ 5 [2 pts]
3. Find an integer value of k with 1 < k < 2019 such that E(2019, k) = ∞. [3 pts]
n
4. a. Show that if n + k is even and 2 < k < n, then E(n, k) = 3. [3 pts]
b. Show that if n is even and k is odd, then E(n, k) = E(n, n − k). [3 pts]
5. Find the following.
a. E(2020, 1993) [3 pts]
b. E(2001, 501) [3 pts]
n
6. a. Show that if n and k are both even and k ≤ 2, then E(n, k) = d nk e. [3 pts]
n
b. Prove that if k is odd and k ≤ 2, then either E(n, k) = d nk e or E(n, k) = d nk e + 1. [4 pts]
7. Find all ordered pairs (n, k) for which E(n, k) = 3. [3 pts]

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 11
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
One might guess that in most cases, E(n, k) ≈ nk . In light of this guess, define the inefficiency of the ordered
pair (n, k), denoted I(n, k), as
n
I(n, k) = E(n, k) −
k
if E(n, k) 6= ∞. If E(n, k) = ∞, then by convention, I(n, k) is undefined.
8. a. Compute I(6, 3). [1 pt]
b. Compute I(5, 3). [1 pt]
15
c. Find positive integers n and k for which I(n, k) = 8 . [2 pts]
d. Prove that for any integer x > 2, there exists an ordered pair (n, k) for which I(n, k) > x. [3 pts]
n
9. Let S be the set of values of I(n, k) for all n, k for which k < 2 and I(n, k) is defined. Find the least upper
bound of S. Prove that your answer is correct. [4 pts]
10. Find two distinct non-integral positive rational numbers that are not the inefficiency of any ordered pair. That
is, find positive rational numbers q1 and q2 with q1 6= q2 such that neither q1 nor q2 is an integer and such that
neither q1 nor q2 is I(n, k) for any integers n and k. Prove that your answers are correct. [4 pts]

12 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
5 Solutions to Power Question
First, notice that a light switch is on if it has been flipped an odd number of times, and off if it has been flipped an
even number of times.

We use the notation {a1 , a2 , . . . , ak } to denote the set of k switches flipped in any given minute.

1. a. E(6, 1) = 6. Note that at least six minutes are required because exactly one switch is flipped each minute.
By flipping all six switches (in any order) in the first six minutes, the door will open in six minutes.
b. E(6, 2) = 3. The sequence {1, 2}, {3, 4}, {5, 6} will allow Elizabeth to escape the room in three minutes.
It is not possible to escape the room in fewer than three minutes because every switch must be flipped,
and that requires at least 62 = 3 minutes.
c. E(7, 3) = 3. First, note that E(7, 3) ≥ 3, because after only two minutes, it is impossible to flip each
switch at least once. It is possible to escape in three minutes with the sequence {1, 2, 3}, {1, 4, 5}, and
{1, 6, 7}.
d. E(9, 5) = 3. Notice that E(9, 5) 6= 1 because each switch must be flipped at least once, and only five
switches can be flipped in one minute. Notice also that E(9, 5) 6= 2 because after two minutes, there have
been 10 flips, but in order to escape the room, each switch must be flipped at least once, and this requires
9 of the 10 flips. However, the tenth flip of a switch returns one of the nine switches to the off position, so
it is not possible for Elizabeth to escape in two minutes. In three minutes, however, Elizabeth can escape
with the sequence {1, 2, 3, 4, 5}, {1, 2, 3, 6, 7}, {1, 2, 3, 8, 9}.

2. a. If n is even, then E(n, 2) = n2 . This is the minimum number of minutes required to flip each switch at
least once, and Elizabeth can clearly escape in n2 minutes by flipping each switch exactly once.
b. If n = 3a + 2 (a ≥ 1), then E(n, 3) = a + 2. The minimum number of minutes required to flip each switch
once is a + 1, but as in Problem 1d, this leaves exactly one “extra flip”, so some switch must be flipped
exactly twice. However, in a + 2 = n+4 3 minutes, Elizabeth can escape by starting with the sequence
{1, 2, 3}, {1, 2, 4}, {1, 2, 5}, and flipping each remaining switch exactly once.
c. If n ≥ 5, then E(n, n − 2) = 3. Note that Elizabeth cannot flip every switch in one minute, and after
two minutes, some switch (in fact, many switches) must be flipped exactly twice. However, Elizabeth can
escape in three minutes using the sequence {1, 4, 5, . . . , n}, {2, 4, 5, . . . , n}, {3, 4, 5, . . . , n}.

3. The answer is that k can be any even number between 2 and 2018 inclusive. If Elizabeth escapes, each switch
must have been flipped an odd number of times. Because there are 2019 switches, the total number of flips
must be odd. However, if an even number of flips are performed each minute, then the total number of flips
cannot be odd. Therefore the puzzle is impossible, and E(2019, k) = ∞.

Alternate Solution: As in Problem 2a, consider the number of lights that are on after each minute. Flipping
an even number of switches either leaves this number unchanged, or increases or decreases it by an even number.
Because the puzzle starts with 0 lights on, and 2019 is odd, it is impossible to have 2019 lights on at the end
of any minute. Thus E(2019, k) = ∞.

4. a. First, because k < n, not all switches can be flipped in one minute, and so E(n, k) 6= 1. Second, because
k > n2 , some switches must be flipped twice in the first two minutes, and so E(n, k) 6= 2.

Here is a strategy by which Elizabeth can escape in three minutes. Note that because n + k is even,
it follows that 3k − n = (k + n) + 2(k − n) is also even. Then let 3k − n = 2b. Note that b > 0 (because
k > n2 ), and b < k (because k < n).

Elizabeth’s strategy is to flip switches 1 through b three times each, and the remaining n − b switches once
each. This is possible because n−b = (3k −2b)−b = 3(k −b) is divisible by 3, and b+ n−b 3 = b+(k −b) = k.
Therefore, each minute, Elizabeth can flip switches 1 through b, and one-third of the 3(k − b) switches
from b + 1 through n. This allows her to escape in three minutes, as desired.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 13
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
b. Because n is even, and because each switch must be flipped an odd number of times in order to escape,
the total number of flips is even. Because k must be odd, E(n, k) must be even. To show this, consider
the case where E(n, k) is odd. If E(n, k) is odd, then an odd number of flips happen an odd number of
times, resulting in an odd number of total flips. This is a contradiction because n is even.

Call a switch “non-flipped” in any given minute if it is not among the switches flipped in that minute.
Because E(n, k) (i.e., the total number of minutes) is even, and each switch is flipped an odd number
of times, each switch must also be non-flipped an odd number of times. Therefore any sequence of flips
that solves the “(n, k) puzzle” can be made into a sequence of flips that solves the “(n, n − k)” puzzle
by interchanging flips and non-flips. These sequences last for the same number of minutes, and therefore
E(n, k) = E(n, n − k).
5. a. E(2020, 1993) = 76. By the result of Problem 4, conclude that E(2020, 1993) = E(2020, 27). Compute the
latter instead. Because 2020
27 > 74, it will require at least 75 minutes to flip each switch once. Furthermore,
E(2020, 27) ≥ 76 because the solution to Problem 4 implies that E(2020, 27) is even.

To solve the puzzle in exactly 76 minutes, use the following strategy. For the first 33 minutes, flip
switch 1, along with the first 26 switches that have not yet been flipped. The end result is that lights 1
through 26 · 33 + 1 = 859 are on, and the remaining 1161 lights are off. Note that 1161 = 27 · 43, so it
takes 43 minutes to flip each remaining switch exactly once, for a total of 76 minutes, as desired.
b. E(2001, 501) = 5. First, note that three minutes is not enough time to flip each switch once. In four
minutes, Elizabeth can flip each switch once, but has three flips left over. Because there are an odd
number of leftover flips to distribute among the 2001 switches, some switch must get an odd number of
leftover flips, and thus an even number of total flips. Thus E(2001, 501) > 4.

To solve the puzzle in five minutes, Elizabeth can flip the following sets of switches:
• in the first minute, {1, 2, 3, . . . , 501};
• in the second minute, {1, 2, 3, . . . , 102} and {502, 503, 504, . . . , 900};
• in the third minute, {1, 2, 3, . . . , 102} and {901, 902, 903, . . . , 1299};
• in the fourth minute, {1, 2, 3, . . . , 100} and {1300, 1301, 1302, . . . , 1700};
• in the fifth minute, {1, 2, 3, . . . , 100} and {1701, 1702, 1703, . . . , 2001}.
This results in switches 1, 2, 3, . . . , 100 being flipped five times, switches 101 and 102 being flipped three
times, and the remaining switches being flipped exactly once, so that all the lights are on at the end of
the fifth minute.
n
6. a. First, if k divides n, then E(n, k) = k = d nk e. Assume then that k does not divide n. Then let r = d nk e,
which implies (r − 1)k < n < rk.

Because (r − 1)k < n, it follows that r − 1 minutes are not enough to flip each switch once, so E(n, k) ≥ r.

Because n and k are even, it follows that rk and rk − n are also even. Then rk − n = 2b for some
integer b ≥ 1, and note that b < k because rk − k < n. Then the following strategy turns all of the lights
on in r minutes.
• In each of the first three minutes, flip switches 1 through b, along with the next k − b switches
that have not yet been flipped. This leaves b + 3(k − b) lights on, and the rest off. Note that
b + 3(k − b) = 3k − 2b = 3k − (rk − n) = n − k(r − 3).
• In each of the remaining r − 3 minutes, flip the first k switches that have never been flipped before.
This strategy turns on the remaining k(r − 3) lights, and Elizabeth escapes the room after r minutes.
n
b. First, if k divides n, then E(n, k) = k = d nk e. Assume then that k does not divide n. Then let r = d nk e,
which implies (r − 1)k < n < rk.

Because (r − 1)k < n, it follows that r − 1 minutes are not enough to flip each switch once, so E(n, k) ≥ r.

14 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Exactly one of rk − n and (r + 1)k − n is even. There are two cases.

Case 1: Suppose rk − n = 2b for some integer b ≥ 1. As in the solution to Problem 6a, b < k. Then the
following strategy turns all of the lights on in r minutes.
• In each of the first three minutes, flip switches 1 through b, along with the next k − b switches
that have not yet been flipped. This leaves b + 3(k − b) lights on, and the rest off. Note that
b + 3(k − b) = 3k − 2b = 3k − (rk − n) = n − k(r − 3).
• In each of the remaining r − 3 minutes, flip the first k switches that have never been flipped before.
This strategy turns on the remaining k(r − 3) lights, and Elizabeth escapes the room after r minutes.

Case 2: Suppose (r + 1)k − n = 2b for some integer b ≥ 1. As in the solution to Problem 6a, b < k. Then
the following strategy turns all of the lights on in r + 1 minutes.
• In each of the first three minutes, flip switches 1 through b, along with the next k − b switches
that have not yet been flipped. This leaves b + 3(k − b) lights on, and the rest off. Note that
b + 3(k − b) = 3k − 2b = 3k − (rk + k − n) = n − k(r − 2).
• In each of the remaining r − 2 minutes, flip the first k switches that have never been flipped before.
This strategy turns on the remaining k(r − 2) lights, and Elizabeth escapes the room after r + 1 minutes.
7. Consider the parity of n and of k. If n is odd and k is even, then as shown in Problem 3, E(n, k) = ∞ = 6 3.
If n is even and k is odd, then as noted in the solution to Problem 5, E(n, k) is even, so E(n, k) 6= 3. If n
and k are either both even or both odd and n3 ≤ k < n, then the argument in Problem 4 shows that E(n, k) = 3.

n
If k < 3, then three minutes is not enough time to flip each switch at least once, so E(n, k) > 3.

If k = n, then E(n, k) = 1.

n
Therefore E(n, k) = 3 if and only if n + k is even (that is, n and k have the same parity) and 3 ≤ k < n.
8. a. I(6, 3) = 0. By definition, I(6, 3) = E(6, 3) − 63 . Because 3 | 6, E(6, 3) = 6
3 = 2, and so I(6, 3) = 2 − 2 = 0.
4 5
b. I(5, 3) = By definition, I(5, 3) = E(5, 3) −
3. 3. By Problem 2b, E(5, 3) = E(3 · 1 + 2, 3) = 1 + 2 = 3,
and so I(5, 3) = 3 − 53 = 43 .
c. One such pair is (n, k) = (18, 16). If I(n, k) = 15 n 15
8 , then E(n, k) − k = 8 . Note that E(n, k) > 2 because
k < n.
Suppose E(n, k) = 3. Then nk = 89 . From Problem 7, E(n, k) = 3 if and only if n and k have the same
parity and n3 ≤ k < n, so let n = 18 and k = 16. Then, I(18, 16) = E(18, 16) − 18 9 15
16 = 3 − 8 = 8 , as
desired.
d. Let n = 2x and k = 2x − 1 for some positive integer x. Then n is even and k is odd, so Problem 4 applies,
2x
and E(n, k) = E(n, n − k) = E(n, 1) = 2x. Therefore I(n, k) = 2x − 2x−1 . Because x > 2, it follows that
2x
2x−1 < 2 < x, so I(n, k) > x, as desired.

9. The least upper bound of S is 2. First, note that Problems 6a and 6b together show that if k < n2 , then E(n, k)
is either d nk e or d nk e + 1 (or ∞ in the case where k is even and n is odd). Therefore I(n, k) ≤ 2 because E(n, k)
is at most d nk e + 1 and d nk e + 1 − nk ≤ 2.

Now it will be shown that I(n, k) can be arbitrarily close to 2. To do this, use values of n and k that
are both odd, such that d nk e is even. Specifically, let n = 3k + 2 for k odd; then d nk e = 4. Therefore E(n, k) = 5
by Problem 6b, and
3k + 2 2
I(n, k) = 5 − =2− .
k k
Letting k be a large odd integer, this gives an ordered pair (n, k) with I(n, k) arbitrarily close to 2, as desired.

Combining these two claims shows that the least upper bound of S is 2.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 15
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
10. It should be noted that integer answers are not possible because no positive integer can be the inefficiency of an
ordered pair (n, k). This is because if I(n, k) = E(n, k) − nk is an integer, then nk is an integer so E(n, k) = nk
and I(n, k) = 0.

To find non-integral rational numbers that cannot be the inefficiency of any ordered pair, start with a simple
lemma.

Lemma: Given positive integers n, k, and r, E(rn, rk) ≤ E(n, k).

Proof: If n is odd and k is even, then E(n, k) = ∞ and the lemma is trivially true. Otherwise there is
a sequence of E(n, k) sets of switches to flip which will result in escape. Partition the rn switches into r groups
of n switches, and apply the strategy within each of the r groups, which uses rk flips per minute. This will
achieve escape in E(n, k) minutes, so E(rn, rk) is at most E(n, k). 2

The values E(n, k) and E(rk, nk) can certainly equal one other, as when n is a multiple of k. The inequality
can also be strict. For example, E(4, 3) = E(4, 1) = 4 but E(8, 6) = 3 are obtained from values from earlier
work.

One implication of the above lemma is that I(n, k) ≥ I(rn, rk).

Now all rational numbers q with 0 < q < 1 are the inefficiency of some ordered pair (n, k). To show this,
let r = 3 − q = ab in lowest terms. Then r > 2 so a > 2b, and by Problem 6a, E(2a, 2b) = d ab e = 3 and
I(2a, 2b) = 3 − ab = q.

Next, all rational numbers q with 1 < q < 2 are the inefficiency of some ordered pair (n, k). Again, let
3 − q = ab in lowest terms. The fact that 1 < q < 2 implies that 1 < ab < 2. So by Problem 4a, E(2a, 2b) = 3
and I(2a, 2b) = 3 − ab = q.

If nk = 2, then I(n, k) = 0 as noted above, while if nk > 2, Problem 6b concludes that E(n, k) = d nk e or
E(n, k) = d nk e + 1. In either case, I(n, k) ≤ d nk e + 1 − nk < 2. As all nonnegative rational numbers less than 2
(except the excluded integer 1) have already been shown to be inefficiencies of ordered pairs, the cases nk ≥ 2
can now be excluded from consideration.

Consider ordered pairs (n, k) with 1 < nk < 2. If n + k is even, Problem 4a implies that E(n, k) = 3 so
I(n, k) = 3 − nk < 2 and once again, all such numbers are already known to be inefficiencies. Of course if n is
odd and k is even, then I(n, k) is undefined.

Combining these results shows that a rational number will not be the inefficiency of some ordered pair simply
whenever it is greater than 2 and avoids being the inefficiency of any ordered pair (n, k) where 1 < nk < 2, n is
even, and k is odd.

Let ab > 2 be a non-integral rational number in lowest terms. If ab is the inefficiency of some ordered pair (n, k),
then ab = E(n, k) − nk . Because E(n, k) is an integer, it follows that ab + nk is also an integer. Because 1 < nk < 2,
there is exactly one integer that it can be.

To find non-integral rational numbers that are not the inefficiency of any ordered pair (n, k), choose a ra-
tional number ab in such a way that the only possible choice for nk , in lowest terms, leads to I(n, k) < ab . Then
because I(rn, rk) ≤ I(n, k), no choice of n and k can possibly lead to I(n, k) = ab .

One such choice is ab = 11 3 . The only integer between 11 11 11

3 + 1 and 3 + 2 is 3 + 3 = 5.
4
Choosing
n
n = 4 and k = 3 satisfies the conditions that n is even, k is odd, and 1 < k < 2, and also leads to
I(4, 3) = E(4, 3) − 43 = E(4, 1) − 43 = 4 − 43 = 38 < 11 n
3 . Thus no ordered pair (n, k) in the ratio k = 3
4
11
can yield an inefficiency of 3 .

16 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Another choice that will work is ab = 17 n
5 . The only k that could yield such an inefficiency would be 5
8
n 8 8 12 17
(again, note that n is even, k is odd, and 1 < k < 2). Then I(8, 5) = E(8, 5) − 5 = 4 − 5 = 5 < 5 , and this
is a similar contradiction.

Note that there are many other non-integral rational numbers that are not the inefficiency of any ordered
pair. They just aren’t covered by the method outlined above and require different techniques. For instance, 52
is such a number. For such an ordered pair (n, k) would require nk to be an odd multiple of 12 . Now n cannot
be odd with k even; in that case, I(n, k) is not defined. So both n and k are even. But then previous work
shows that I(n, k) < 2, so 52 is not an inefficiency of any ordered pair (n, k).

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 17
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
6 Individual Problems

Problem 1. In rectangle P AU L, point D is the midpoint of U L and points E and F lie on P L and P A, respectively
PE 3 PF
such that = and = 2. Given that P A = 36 and P L = 25, compute the area of pentagon AU DEF .
EL 2 FA

Problem 2. Rectangle ARM L has length 125 and width 8. The rectangle is divided into 1000 squares of area 1
by drawing in gridlines parallel to the sides of ARM L. Diagonal AM passes through the interior of exactly n
of the 1000 unit squares. Compute n.

Problem 3. Compute the least integer n > 1 such that the product of all positive divisors of n equals n4 .

Problem 4. Each of the six faces of a cube is randomly colored red or blue with equal probability. Compute the
probability that no three faces of the same color share a common vertex.

Problem 5. Scalene triangle ABC has perimeter 2019 and integer side lengths. The angle bisector from C meets
AB at D such that AD = 229. Given that AC and AD are relatively prime, compute BC.

Problem 6. Given that a and b are positive and

20 − a = 19 − b = ab ,

compute the least upper bound of the set of possible values of a + b.

Problem 7. Compute the number of five-digit integers M A R T Y , with all digits distinct, such that M > A > R
and R < T < Y .

Problem 8. In parallelogram ARM L, points P and Q are the midpoints of sides RM and AL, respectively.
Point X lies on segment P Q, and P X = 3, RX = 4, and P R = 5. Point I lies on segment RX such that
[P QR]
IA = IL. Compute the maximum possible value of .
[LIP ]

Problem 9. Given that a, b, c, and d are positive integers such that

a! · b! · c! = d! and a + b + c + d = 37,

compute the product abcd.

Problem 10. Compute the value of

sin(6◦ ) · sin(12◦ ) · sin(24◦ ) · sin(42◦ ) + sin(12◦ ) · sin(24◦ ) · sin(42◦ ).

18 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
7 Answers to Individual Problems

Answer 1. 630

Answer 2. 132

Answer 3. 24
9
Answer 4. (or 0.28125)
32
Answer 5. 888
41
Answer 6. (or 8 51 or 8.2)
5
Answer 7. 1512
4
Answer 8. (or 1 13 or 1.3)
3
Answer 9. 2240
1
Answer 10. (or 0.0625)
16

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 19
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
8 Solutions to Individual Problems

Solution 1. For convenience, let P A = 3x and let P L = 5y. Then the given equations involving ratios of segment
lengths imply that P E = 3y, EL = 2y, P F = 2x, and F A = x. Then [P AU L] = (3x)(5y) = 15xy and

[AU DEF ] = [P AU L] − [P EF ] − [ELD]

Å ã
1 1 3x
= 15xy − (3y)(2x) − (2y)
2 2 2
3xy
= 15xy − 3xy −
2
21xy
= .
2
21xy
Because 15xy = 36 · 25, it follows that 3xy = 36 · 5 = 180 and that 2 = 72 (3xy) = 7
2 · 180 = 630.

Solution 2. Notice that 125 and 8 are relatively prime. Examining rectangles of size a × b where a and b are small
and relatively prime suggests an answer of a + b − 1. To see that this is the case, note that other than the
endpoints, the diagonal does not pass through any vertex of any unit square. After the first square, it must
enter each subsequent square via a vertical or horizontal side. By continuity, the total number of these sides
is the sum of the a − 1 interior vertical lines and b − 1 interior horizontal lines. The diagonal passes through
(a − 1) + (b − 1) = a + b − 2 additional squares, so the total is a + b − 1. Because 125 and 8 are relatively prime,
it follows that N = 125 + 8 − 1 = 132.

Remark: As an exercise, the reader is encouraged to show that the answer for general a and b is a+b−gcd(a, b).

Problem 3. Compute the least integer n > 1 such that the product of all positive divisors of n equals n4 .

Solution 3. Note that every factor pair d and nd have product n. For the product of all such divisor pairs to
equal n4 , there must be exactly 4 divisor pairs, or 8 positive integer divisors. A number has 8 positive integer
divisors if it is of the form a3 b1 or a7 where a and b are distinct primes. The prime factorization a3 b1 (a 6= b)
provides a set of divisors each of which has 4 options for using a (a0 , a1 , a2 , a3 ) and an independent 2 options
for using b (b0 , b1 ). Using the least values (a, b) = (2, 3), a3 b1 = 24. If instead the prime factorization is a7
(having divisors a0 , a1 , a2 , . . . , a7 ), the least answer would be 27 = 128. Thus the answer is 24.

Problem 4. Each of the six faces of a cube is randomly colored red or blue with equal probability. Compute the
probability that no three faces of the same color share a common vertex.

Solution 4. There are 26 = 64 colorings of the cube. Let r be the number of faces that are colored red. Define a
monochromatic vertex to be a vertex of the cube for which the three faces meeting there have the same color.
It is clear that a coloring without a monochromatic vertex is only possible in the cases 2 ≤ r ≤ 4. If r = 2 or
r = 4, the only colorings that do not have a monochromatic vertex occur when two opposing faces are colored
with the minority color (red in the r = 2 case, blue in the r = 4 case). Because there are 3 pairs of opposite

20 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
faces of a cube, there are 3 colorings without a monochromatic vertex if r = 2 and another 3 such colorings if
r = 4. For the r = 3 colorings, of which there are 20, the only cases in which there are monochromatic vertices
occur when opposing faces are monochromatic, but in different colors. There are 23 = 8 such colorings, leaving
20 − 8 = 12 colorings that do not have a monochromatic vertex. Therefore 3 + 3 + 12 = 18 of the 64 colorings
9
have no monochromatic vertex, and the answer is .
32

Problem 5. Scalene triangle ABC has perimeter 2019 and integer side lengths. The angle bisector from C meets
AB at D such that AD = 229. Given that AC and AD are relatively prime, compute BC.

Solution 5. Let BC = a, AC = b, AB = c. Also, let AD = e and BD = f . Then a + b + e + f = 2019, the

e b ae 229a
values a, b, and e + f are integers, and by the Angle Bisector Theorem, = . So b = = . Because
f a f f
229 is prime and gcd(b, e) = 1, conclude that f must be an integer multiple of 229. So let f = 229x for some
integer x. Then a = b · x and a + b + c = 2019 implies 2019 = bx + b + 229 + 229x = (b + 229)(1 + x). Because
2019 = 673 · 3, it follows that b = 444 and x = 2, from which BC = a = 888.

Problem 6. Given that a and b are positive and

20 − a = 19 − b = ab ,
compute the least upper bound of the set of possible values of a + b.

Solution 6. Let the common value of the three expressions in the given equation be N . Maximizing a + b
involves making at least one of a and b somewhat large, which makes the first two expressions for N small.
So, to maximize a + b, look for the least possible value of N . One can show that N = 14 is not possible
because that would require a > 5 and b > 4, which implies ab > 20. But N = 15 is possible by setting
a = 4 + x, b = 3 + y, where 0 < x, y ≤ 1. The goal is to find the least upper bound for x + y given
15 ≤ (4 + x)(3 + y) < 16 ⇒ 3 ≤ 3(x + y) + y + xy < 4. This is equivalent to seeking the maximum value of
x + y given 3(x + y) + y + xy ≤ 4. By inspection, if x = 1 and y = 15 , then 3(x + y) + y + xy = 4 ≤ 4. This
4 − 3x
is in fact optimal. To see this, consider that because 3x + 3y + y + xy ≤ 4, it follows that y ≤ , and so
x+4
4 − 3x x2 + x + 4
x+y ≤x+ ≤ , which is increasing on 0 ≤ x ≤ 1. Thus the maximum for x + y is attained
x+4 x+4
41
when x = 1. Hence the least upper bound for a + b is 5 + (3 + 51 ) = .
5

Problem 7. Compute the number of five-digit integers M A R T Y , with all digits distinct, such that M > A > R
and R < T < Y .

Solution 7. There are 10

5 = 252 ways to choose the values of the digits M, A, R, T, Y , without restrictions.
Because R is fixed as the least of the digits and because T < Y , it suffices to find the number of ways to choose
M and A. Once M and A are chosen, the other three digits are uniquely determined. There are 42 = 6 ways
to select M, A. Thus the number of five-digit integers of the type described is 252 · 6 = 1512.

Solution 8. Because AI = LI and AQ = LQ, line IQ is the perpendicular bisector of AL. Because ARM L is a
parallelogram, QI ⊥ RP . Note also that m∠RXP = 90◦ . Thus I is the orthocenter of triangle P QR, from

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 21
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
←
→
which it follows that P I ⊥ RQ and P I ⊥ P L (because P RQL is a parallelogram). Extend P I through I to
meet RQ at D. Then 2[P QR] = RQ · P D and 2[LIP ] = P I · P L = P I · RQ. Hence the problem is equivalent
to determining the maximum value of P D/P I.

Set m∠RP D = m∠RP I = α and m∠IP X = β, and note that P D = P R cos α = 5 cos α and P I = P X/cos β =
3/cos β. It follows that

PD 5 cos α cos β 5 cos(α + β) + cos(α − β) 5(3/5 + 1) 4
= = ≤ = ,
PI 3 6 6 3
with equality when α = β.

A R

D
I

P
Q X

L M

Problem 9. Given that a, b, c, and d are positive integers such that

a! · b! · c! = d! and a + b + c + d = 37,

compute the product abcd.

Solution 9. Without loss of generality, assume a ≤ b ≤ c < d. Note that d cannot be prime, as none of a!, b!, or
c! would have it as a factor. If d = p + 1 for some prime p, then c = p and a!b! = p + 1. The least possible
values of a!b! are 1, 2, 4, 6, 24, 36, 48, 120, 144, 240, so the case where d = p + 1 is impossible. If d ≥ 21, then
d!
a + b + c ≤ 16 and it is impossible to find values of a and b such that a! · b! = . If d = 16, either a!b! = 16
c!
or a!b! = 16 · 15 or a!b! = 16 · 15 · 14. Comparing to the list above, the only possible value a!b! on the list is
16 · 15 = 240 and so (a, b, c, d) = (2, 5, 14, 16) and abcd = 2240.

Problem 10. Compute the value of

sin(6◦ ) · sin(12◦ ) · sin(24◦ ) · sin(42◦ ) + sin(12◦ ) · sin(24◦ ) · sin(42◦ ).

Solution 10. Let S = (1+sin 6◦ )(sin 12◦ sin 24◦ sin 42◦ ). It follows from a sum-to-product identity that 1+sin 6◦ =
sin 90◦ + sin 6◦ = 2 sin 48◦ cos 42◦ . Because the sine of an angle is the cosine of its complement, it follows that

S = (2 sin 48◦ cos 42◦ )(sin 12◦ sin 24◦ sin 42◦ ) = 2(sin 48◦ )2 (sin 12◦ sin 24◦ cos 48◦ ).

By the double-angle formula, this means S = sin 12◦ sin 24◦ sin 48◦ sin 96◦ . By a product-to-sum identity,
√
◦ ◦ cos 36◦ − cos 60◦ 5−1
sin 12 sin 48 = = (1)
2 8

22 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
and √
◦ cos 72◦ − cos 120◦
◦ 5+1
sin 24 sin 96 = = . (2)
2 8
1
Multiply the expressions on the right-hand sides of (1) and (2) to obtain .
16

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 23
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
9 Relay Problems

Relay 1-1. Let a = 19, b = 20, and c = 21. Compute

a2 + b2 + c2 + 2ab + 2bc + 2ca

.
a+b+c

Relay 1-2. Let T = TNYWR. Lydia is a professional swimmer and can swim one-fifth of a lap of a pool in an
impressive 20.19 seconds, and she swims at a constant rate. Rounded to the nearest integer, compute the
number of minutes required for Lydia to swim T laps.

√
Relay 1-3. Let T = TNYWR. In 4ABC, m∠C = 90◦ and AC = BC = T − 3. Circles O and P each have
radius r and lie inside 4ABC. Circle O is tangent to AC and BC. Circle P is externally tangent to circle O
and to AB. Given that points C, O, and P are collinear, compute r.

√
Relay 2-1. Given that p = 6.6 × 10−27 , then p = a × 10b , where 1 ≤ a < 10 and b is an integer. Compute 10a + b
rounded to the nearest integer.

Relay 2-2. Let T = TNYWR. A group of children and adults go to a rodeo. A child’s admission ticket costs $5,
and an adult’s admission ticket costs more than $5. The total admission cost for the group is $10 · T . If the
number of adults in the group were to increase by 20%, then the total cost would increase by 10%. Compute
the number of children in the group.

Relay 2-3. Let T = TNYWR. Rectangles F AKE and F U N K lie in the same plane. Given that EF = T ,
AF = 4T 12
3 , and U F = 5 , compute the area of the intersection of the two rectangles.

24 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
10 Relay Answers

Answer 1-1. 60

Answer 1-2. 101

√
Answer 1-3. 3 − 2

Answer 2-1. 67

Answer 2-2. 67

Answer 2-3. 262

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 25
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
11 Relay Solutions

Relay 1-1. Let a = 19, b = 20, and c = 21. Compute

a2 + b2 + c2 + 2ab + 2bc + 2ca

.
a+b+c

Solution 1-1. Note that the numerator of the given expression factors as (a + b + c)2 , hence the expression to be
computed equals a + b + c = 19 + 20 + 21 = 60.

Solution 1-2. Lydia swims a lap in 5 · 20.19 = 100.95 seconds. The number of minutes required for Lydia to
swim T laps is therefore 100.95 · T /60. With T = 60, the desired number of minutes, rounded to the nearest
integer, is 101.

Solution 1-3. Let A0 and B 0 be the respective feet of the perpendiculars from O to AC and BC. Let H be
the foot of the altitude from C to AB. Because 4ABC is isosceles, it follows that A0 OB 0 C is a square,
m∠B 0 CO = 45◦ , and m∠BCH = 45 ◦
√ . Hence H lies on √ the same line as C, O, and√P . In terms of r, the
length CH
√
is CO + OP + P H√
= r 2 + 2r +√r = (3
√
+ 2)r. Because AC = BC = T − 3, it follows that
T −3 T −3 (3 2 − 2) T − 3 √ √ √
CH = √ . Thus r = √ √ = . With T = 101, T − 3 = 98 = 7 2, and it
2 2(3 + 2) 14
√
follows that r = 3 − 2.

√
Relay 2-1. Given that p = 6.6 × 10−27 , then p = a × 10b , where 1 ≤ a < 10 and b is an integer. Compute 10a + b
rounded to the nearest integer.

√ √ √
Solution 2-1. Note that p = 6.6 × 10−27 = 66 × 10−28 , so a = 66 and b√= −14. Note that 66 > 64 = 8.
Because 8.12 = 65.61 and 8.152 = 66.4225 > 66, conclude that 81 < 10 66 < 81.5, hence 10a rounded to the
nearest integer is 81, and the answer is 81 − 14 = 67.

Solution 2-2. Suppose there are x children and y adults in the group and each adult’s admission ticket costs $a.
The given information implies that 5x + ay = 10T and 5x + 1.2ay = 11T . Subtracting the first equation from
the second yields 0.2ay = T → ay = 5T , so from the first equation, 5x = 5T → x = T . With T = 67, the
answer is 67.

26 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Relay 2-3. Let T = TNYWR. Rectangles F AKE and F U N K lie in the same plane. Given that EF = T ,
AF = 4T 12
3 , and U F = 5 , compute the area of the intersection of the two rectangles.

Solution 2-3. Without loss of generality, let A, U , and N lie on the same side of F K. Applying the Pythagorean
Theorem to triangle AF K, conclude that F K = 5T 3 . Comparing the altitude to F K in triangle AF K to U F ,
2
note that the intersection of the two rectangles will be a triangle with area 2T3 if 4T 12
5 ≤ 5 , or T ≤ 3. Otherwise,
the intersection will be a trapezoid. In this case, using similarity, the triangular regions of F U N K that lie
outside of F AKE each have one leg of length 12 16 9
5 and the others of lengths 5 and 5 , respectively. Thus their
combined areas 12 · 12 16 9 5T 12
5 ( 5 + 5 ) = 6, hence the area of the intersection is 3 · 5 − 6 = 4T − 6. With T = 67, the
answer is therefore 262.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 27
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
12 Super Relay
1. Given that a, b, c, and d are integers such that a + bc = 20 and −a + cd = 19, compute the greatest possible
value of c.
2. Let T = TNYWR. Emile randomly chooses a set of T cards from a standard deck of 52 cards. Given that
Emile’s set contains no clubs, compute the probability that his set contains three aces.
AB
3. Let T = TNYWR. In parallelogram ABCD, = T . Given that M is the midpoint of AB and P and Q are
BC
[ABCD]
the trisection points of CD, compute .
[M P Q]
√
4. Let T = TNYWR. Compute the value of x such that logT x − 7 + logT 2 (x − 2) = 1.
5. Let T = TNYWR. Let p be an odd prime and let x, y, and z be positive integers less than p. When the
trinomial (px + y + z)T −1 is expanded and simplified, there are N terms, of which M are always multiples of p.
Compute M .
6. Let T = TNYWR. Compute the value of K such that 20, T − 5, K is an increasing geometric sequence
and 19, K, 4T + 11 is an increasing arithmetic sequence.
7. Let T = TNYWR. Cube C1 has volume T and sphere S1 is circumscribed about C1 . For n ≥ 1, the sphere Sn
is circumscribed about the cube Cn and is inscribed in the cube Cn+1 . Let k be the least integer such that the
volume of Ck is at least 2019. Compute the edge length of Ck .

15. Square KEN T has side length

√ 20. Point M lies in the interior of KEN T such that 4M EN is equilateral.
Given that KM 2 = a − b 3, where a and b are integers, compute b.
14. Let T = TNYWR. Let a, b, and c be the three solutions of the equation x3 − 20x2 + 19x + T = 0. Compute
a2 + b2 + c2 .
√
13. Let T = TNYWR and let K = T − 1. Compute (K − 20)(K + 1) + 19K − K 2 .
1 1
12. Let T = TNYWR. In 4LEO, sin ∠LEO = . If LE = for some positive real number n, then EO =
T n
n3 − 4n2 + 5n. As n ranges over the positive reals, compute the least possible value of [LEO].
1
11. Let T = TNYWR. Given that x, y, and z are real numbers such that x + y = 5, x2 − y 2 = , and x − z = −7,
T
compute x + z.
10. Let T = TNYWR. The product of all positive divisors of 2T can be written in the form 2K . Compute K.
9. Let T = TNYWR. At the Westward House of Supper (“WHS”), a dinner special consists of an appetizer, an
entrée, and dessert. There are 7 different appetizers and K different entrées that a guest could order. There
are 2 dessert choices, but ordering dessert is optional. Given that there are T possible different orders that
could be placed at the WHS, compute K.

8. Let S be the number you will receive from position 7 and let M be the number you will receive from position 9.
Sam and Marty each ride a bicycle at a constant speed. Sam’s speed is S km/hr and Marty’s speed is M km/hr.
Given that Sam and Marty are initially 100 km apart and they begin riding towards one another at the same
time, along a straight path, compute the number of kilometers that Sam will have traveled when Sam and
Marty meet.

28 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
13 Super Relay Answers

1. 39
2. 1
3. 6
4. 11
5. 55
6. 125
7. 15
15. 400
14. 362
13. 20
1
12.
40
11. 20
10. 210
9. 10
8. 60

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 29
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
14 Super Relay Solutions

Problem 1. Given that a, b, c, and d are integers such that a + bc = 20 and −a + cd = 19, compute the greatest
possible value of c.

Solution 1. Adding the two given equations yields bc + cd = c(b + d) = 39. The greatest possible value of c
therefore occurs when c = 39 and b + d = 1.

Problem 2. Let T = TNYWR. Emile randomly chooses a set of T cards from a standard deck of 52 cards. Given
that Emile’s set contains no clubs, compute the probability that his set contains three aces.

Solution 2. Knowing that 13 of the cards are not in Emile’s set, there are 39

T ways for him to have chosen a
set of T cards. Given that Emile’s set contains no clubs, the suits of the three aces are fixed (i.e., diamonds,
hearts, and spades). The number of possible sets of cards in which these three aces appear is therefore T36

−3 .
36
39
The desired probability is therefore T −3 / T . With T = 39, this probability is 1/1 = 1, which is consistent
with the fact that Emile’s set contains all cards in the deck that are not clubs, hence he is guaranteed to have
all three of the remaining aces.

AB
Problem 3. Let T = TNYWR. In parallelogram ABCD, = T . Given that M is the midpoint of AB and P
BC
[ABCD]
and Q are the trisection points of CD, compute .
[M P Q]

Solution 3. Let CD = 3x and let h be the length of the altitude between bases AB and CD. Then [ABCD] = 3xh
[ABCD] AB
and [M P Q] = 12 xh. Hence = 6. Both the position of M and the ratio = T are irrelevant.
[M P Q] BC

√
Problem 4. Let T = TNYWR. Compute the value of x such that logT x − 7 + logT 2 (x − 2) = 1.

√
Solution 4. It can readily be shown that loga b = loga2 b2 . Thus it follows that logT x − 7 = logT 2 (x − 7). Hence
the left-hand side of the given equation is logT 2 (x−7)(x−2) and the equation is equivalent to (x−7)(x−2) = T 2 ,
which is equivalent to x2 −9x+14−T 2 = 0. With T = 6, this equation is x2 −9x−22 = 0 =⇒ (x−11)(x+2) = 0.
Plugging x = −2 into the given equation leads to the first term of the left-hand side having a negative radicand
and the second term having an argument of 0. However, one can easily check that x = 11 indeed satisfies the
given equation.

Problem 5. Let T = TNYWR. Let p be an odd prime and let x, y, and z be positive integers less than p. When
the trinomial (px + y + z)T −1 is expanded and simplified, there are N terms, of which M are always multiples
of p. Compute M .

Solution 5. A general term in the expansion of (px + y + z)T −1 has the form K(px)a y b z c , where a, b, and c are
nonnegative integers such that a+b+c = T −1. Using the “stars and bars” approach, the number of nonnegative
integral solutions to a + b + c = T − 1 is the number of arrangements of T − 1 stars and 2 bars in a row (the
bars act has separators
and the “2” arises because it is one less than the number of variables in the equation).
Thus there are T +12 solutions. Each term will be a multiple of p unless a = 0. In this case, the number of
terms that are not multiples of p is the number of nonnegative integral solutions to the equation b + c = T − 1,
2
which is T (b can range from 0 to T − 1 inclusive, and then c is fixed). Hence M = T +1 − T = T 2−T . With

2
T = 11, the answer is 55.

30 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Problem 6. Let T = TNYWR. Compute the value of K such that 20, T − 5, K is an increasing geometric
sequence and 19, K, 4T + 11 is an increasing arithmetic sequence.

T −5 K
Solution 6. The condition that 20, T − 5, K is an increasing geometric sequence implies that 20 = T −5 ,
(T −5)2
hence K = 20 . The condition that 19, K, 4T + 11 is an increasing arithmetic sequence implies that
K − 19 = 4T + 11 − K, hence K = 2T + 15. With T = 55, each of these equations implies that K = 125.
Note that the two equations can be combined and solved without being passed a value of T . A quadratic
equation results, and its roots are T = 55 or T = −5. However, with T = −5, neither of the given sequences is
increasing.

Problem 7. Let T = TNYWR. Cube C1 has volume T and sphere S1 is circumscribed about C1 . For n ≥ 1, the
sphere Sn is circumscribed about the cube Cn and is inscribed in the cube Cn+1 . Let k be the least integer such
that the volume of Ck is at least 2019. Compute the edge length of Ck .

Solution 7. In general, √let cube Cn have edge length x. Then the diameter of sphere Sn is the space diagonal of Cn ,
which has length x 3. This in turn is the edge length √ of cube Cn+1 . Hence the edge lengths of C1 , C2 , . . . form
an increasing geometric sequence √ with common ratio 3 and volumes of C1 , C2 , . . . form an increasing geometric
sequence with common√ ratio 3 3. With T = 125, the edge length of C1 is 5, so the sequence of√edge lengths
of the cubes is 5, 5 3, 15, . . ., and the respective sequence of the volumes of the cubes is 125, 375 3, 3375, . . . .
Hence k = 3, and the edge length of C3 is 15.

Problem 15. Square KEN T has side length√ 20. Point M lies in the interior of KEN T such that 4M EN is
equilateral. Given that KM 2 = a − b 3, where a and b are integers, compute b.

Solution 15. Let s be the side length of square KEN T ; then M E = s. Let J √
be the foot of the altitude from M√to
KE. Then m∠JEM = 30◦ and m∠EM J = 60◦ . Hence M J = 2s , JE = s 2 3 , and KJ = KE − JE = s − s 2 3 .
Ä √ ä2 2 √
Applying the Pythagorean Theorem to 4KJM implies that KM 2 = s − s 2 3 + 2s = 2s2 − s2 3. With
s = 20, the value of b is therefore s2 = 400.

Problem 14. Let T = TNYWR. Let a, b, and c be the three solutions of the equation x3 − 20x2 + 19x + T = 0.
Compute a2 + b2 + c2 .

Solution 14. According to Vieta’s formulas, a + b + c = −(−20) = 20 and ab + bc + ca = 19. Noting that
a2 + b2 + c2 = (a + b + c)2 − 2(ab + bc + ca), it follows that a2 + b2 + c2 = 202 − 2 · 19 = 362. The value of T is
irrelevant.

√
Problem 13. Let T = TNYWR and let K = T − 1. Compute (K − 20)(K + 1) + 19K − K 2 .

2 2
√ value bars simplifies to K − 19K − 20 + 19K − K = −20. Hence
Solution 13. The expression inside the absolute
the answer is 20 and the value of K (= 361 = 19) is not needed.

1 1
Problem 12. Let T = TNYWR. In 4LEO, sin ∠LEO = . If LE = for some positive real number n, then
T n
EO = n3 − 4n2 + 5n. As n ranges over the positive reals, compute the least possible value of [LEO].

2
Solution 12. Note that [LEO] = 21 (sin ∠LEO) · LE · EO = 12 · T1 · n1 · (n3 − 4n2 + 5n) = n −4n+5
2T . Because T is
a constant, the least possible value of [LEO] is achieved when the function f (n) = n2 − 4n + 5 is minimized.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 31
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
This occurs when n = −(−4)/(2 · 1) = 2, and the minimum value is f (2) = 1. Hence the desired least possible
1 1
value of [LEO] is 2T , and with T = 20, this is 40 .

1
Problem 11. Let T = TNYWR. Given that x, y, and z are real numbers such that x + y = 5, x2 − y 2 = , and
T
x − z = −7, compute x + z.

1
Solution 11. Note that x2 −y 2 = (x+y)(x−y) = 5(x−y), hence x−y = 5T . Then x+z = (x+y)+(x−y)+(z−x) =
1 1 1
5 + 5T + 7 = 12 + 5T . With T = 40 , the answer is thus 12 + 8 = 20.

Problem 10. Let T = TNYWR. The product of all positive divisors of 2T can be written in the form 2K .
Compute K.

Solution 10. When n is a nonnegative integer, the product of the positive divisors of 2n is 20 · 21 · . . . · 2n−1 · 2n =
20+1+···+(n−1)+n = 2n(n+1)/2 . Because T = 20 is an integer, it follows that K = T (T2+1) = 210.

Problem 9. Let T = TNYWR. At the Westward House of Supper (“WHS”), a dinner special consists of an
appetizer, an entrée, and dessert. There are 7 different appetizers and K different entrées that a guest could
order. There are 2 dessert choices, but ordering dessert is optional. Given that there are T possible different
orders that could be placed at the WHS, compute K.

Solution 9. Because dessert is optional, there are effectively 2+1 = 3 dessert choices. Hence, by the Multiplication
T
Principle, it follows that T = 7 · K · 3, thus K = 21 . With T = 210, the answer is 10.

Problem 8. Let S be the number you will receive from position 7 and let M be the number you will receive from
position 9. Sam and Marty each ride a bicycle at a constant speed. Sam’s speed is S km/hr and Marty’s speed
is M km/hr. Given that Sam and Marty are initially 100 km apart and they begin riding towards one another
at the same time, along a straight path, compute the number of kilometers that Sam will have traveled when
Sam and Marty meet.

Solution 8. In km/hr, the combined speed of Sam and Marty is S + M . Thus one can determine the total time
they traveled and use this to determine the number of kilometers that Sam traveled. However, this is not
needed, and there is a simpler approach. Suppose that Marty traveled a distance of d. Then because Sam’s
S S S
speed is M of Marty’s speed, Sam will have traveled a distance of M · d. Thus, together, they traveled d + M · d.
100M S 100S
Setting this equal to 100 and solving yields d = M +S . Thus Sam traveled M · d = M +S . With S = 15 and
M = 10, this is equal to 60 km.

32 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
15 Tiebreaker Problems

Problem 1. Regular tetrahedra JAN E, JOHN , and JOAN have non-overlapping interiors. Compute tan ∠HAE.

Problem 2. Each positive integer less than or equal to 2019 is written on a blank sheet of paper, and each of
the digits 0 and 5 is erased. Compute the remainder when the product of the remaining digits on the sheet of
paper is divided by 1000.

Problem 3. Compute the third least positive integer n such that each of n, n + 1, and n + 2 is a product of exactly
two (not necessarily distinct) primes.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 33
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
16 Tiebreaker Answers
√
5 2
Answer 1.
2
Answer 2. 976

Answer 3. 93

34 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
17 Tiebreaker Solutions

Problem 1. Regular tetrahedra JAN E, JOHN , and JOAN have non-overlapping interiors. Compute tan ∠HAE.

Solution 1. First note that JN is a shared edge of all three pyramids, and that the viewpoint for the figure below
is from along the line that is the extension of edge JN .

Let h denote the height of each pyramid. Let X be the center of pyramid JOAN , and consider the plane
passing through H, A, and E. By symmetry, the altitude in pyramid JOHN through H and the altitude in
pyramid JAN E through E pass through X. Thus points H, X, and A are collinear, as are points E, X, and
O. Hence AH = OE = 2h. Using the result that the four medians in a tetrahedron are concurrent and divide
each other in a 3 : 1 ratio, it follows that AX = OX = 3h 5h
4 and XE = OE − OX = 4 . Applying the Law of
2
Cosines to triangle AXE yields cos ∠XAE = cos ∠HAE = 2−2h . Suppose, without loss of generality, that
» 3h √ √
the common side length of the pyramids is 1. Then h = 3 and cos ∠HAE = 96 . Hence sin ∠HAE = 975
2
√
5 2
and therefore tan ∠HAE = .
2

Solution 2. Count the digits separately by position, noting that 1 is irrelevant to the product. There are a total of
20 instances of the digit 2 in the thousands place. The digit 0 only occurs in the hundreds place if the thousands
digit is 2, so look at the numbers 1 through 1999. Each non-zero digit contributes an equal number of times, so
there are 200 each of 1, 2, 3, 4, 6, 7, 8, 9. The same applies to the tens digit, except there can be the stray digit
of 1 among the numbers 2010 through 2019, but again, these do not affect the product. In the units place, there
are 202 of each of the digits. Altogether, there are 602 each of 2, 3, 4, 6, 7, 8, 9, along with 20 extra instances of
the digit 2. Note that 9·8·7·6·4·3·2 = 3024·24 = 72,576 leaves a remainder of 576 when divided by 1000. Also
220 = 10242 ≡ 242 (mod 1000), so 220 contributes another factor of 576. The answer is therefore the remainder
when 576603 is divided by 1000. This computation can be simplified by using the Chinese Remainder Theorem
with moduli 8 and 125, whose product is 1000. Note 576603 ≡ 0 (mod 8) because 576 is divisible by 8. Also
576 ≡ 76 (mod 125). By Euler’s totient theorem, 576100 ≡ 1 (mod 125), so 576603 ≡ 763 (mod 125). This can
quickly be computed by noting that 763 = (75 + 1)3 = 753 + 3 · 752 + 3 · 75 + 1 ≡ 3 · 75 + 1 ≡ −24 (mod 125).
Observing that −24 ≡ 0 (mod 8), it follows that 576603 ≡ −24 (mod 1000), hence the desired remainder is 976.

ARML encourages the reproduction of our contest problems for non-commercial, educational purposes. 35
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.
Problem 3. Compute the third least positive integer n such that each of n, n + 1, and n + 2 is a product of exactly
two (not necessarily distinct) primes.

Solution 3. Define a positive integer n to be a semiprime if it is a product of exactly two (not necessarily distinct)
primes. Define a lucky trio to be a sequence of three consecutive integers, n, n + 1, n + 2, each of which is a
semiprime. Note that a lucky trio must contain exactly one multiple of 3. Also note that the middle number
in a lucky trio must be even. To see this, note that if the first and last numbers in a lucky trio were both even,
then exactly one of these numbers would be a multiple of 4. But neither 2, 3, 4 nor 4, 5, 6 is a lucky trio, and
if a list of three consecutive integers contains a multiple of 4 that is greater than 4, this number cannot be a
semiprime. Using this conclusion and because 3, 4, 5 is not a lucky trio, it follows that the middle number of a
lucky trio cannot be a multiple of 4. Hence it is necessary that a lucky trio has the form 4k + 1, 4k + 2, 4k + 3,
for some positive integer k, with 2k + 1 being a prime. Note that k 6≡ 1 (mod 3) because when k = 1, the
sequence 5, 6, 7 is not a lucky trio, and when k > 1, 4k + 2 would be a multiple of 6 greater than 6, hence
it cannot be a semiprime. Trying k = 2, 3, 5, 6, 8, 9, . . . allows one to eliminate sequences of three consecutive
integers that are not lucky trios, and if lucky trios are ordered by their least elements, one finds that the first
three lucky trios are 33, 34, 35; 85, 86, 87; and 93, 94, 95. Hence the answer is 93.

36 ARML encourages the reproduction of our contest problems for non-commercial, educational purposes.
Commercial usage of ARML problems without permission and posting entire contests or contest books are prohibited.

ARML Competition 2012
100% (5)
ARML Competition 2012
42 pages
2015 Arml
100% (1)
2015 Arml
40 pages
ARML 2014 Contest Final Version
No ratings yet
ARML 2014 Contest Final Version
36 pages
Week 22-ARML Team-Relay-Individual-Power-Answers-Solutions - 2014
No ratings yet
Week 22-ARML Team-Relay-Individual-Power-Answers-Solutions - 2014
20 pages
2019 CMC Arml PDF
No ratings yet
2019 CMC Arml PDF
14 pages
ARML Local 2016 Solutions
No ratings yet
ARML Local 2016 Solutions
14 pages
ARML-Local-2018-Solutions
No ratings yet
ARML-Local-2018-Solutions
11 pages
ARMLLocal2015Solutions
No ratings yet
ARMLLocal2015Solutions
10 pages
ARMLLocal2014Solutions
No ratings yet
ARMLLocal2014Solutions
13 pages
ARML Local 2017 Solutions
No ratings yet
ARML Local 2017 Solutions
11 pages
19th PMO National Stage Oral Phase
No ratings yet
19th PMO National Stage Oral Phase
9 pages
2020 Aoime2
No ratings yet
2020 Aoime2
9 pages
HMMT
No ratings yet
HMMT
5 pages
Sipnayan Gmatic M Wiz
No ratings yet
Sipnayan Gmatic M Wiz
106 pages
PMO21 National Orals Judges
No ratings yet
PMO21 National Orals Judges
9 pages
2019MMPC-part1_solutions
No ratings yet
2019MMPC-part1_solutions
19 pages
VTAMPS 18 SS Set 3
No ratings yet
VTAMPS 18 SS Set 3
13 pages
deExamSolutions
No ratings yet
deExamSolutions
5 pages
ARML Local 2024 Solutions
No ratings yet
ARML Local 2024 Solutions
18 pages
AMC 10 MOCK TEST 2
No ratings yet
AMC 10 MOCK TEST 2
4 pages
2010contestentiredraftv1 4
No ratings yet
2010contestentiredraftv1 4
45 pages
2010 Arml
No ratings yet
2010 Arml
45 pages
ARML 2015 Contest
No ratings yet
ARML 2015 Contest
40 pages
VTAMPS 7.0 Secondary 2 Set 4
No ratings yet
VTAMPS 7.0 Secondary 2 Set 4
12 pages
2023 PWMTC Advanced Individual 2 - Solutions
No ratings yet
2023 PWMTC Advanced Individual 2 - Solutions
4 pages
PMO Area Stage Answers Only
No ratings yet
PMO Area Stage Answers Only
7 pages
American Regions Math League (ARML) - Local Contests - Problems (2018)
No ratings yet
American Regions Math League (ARML) - Local Contests - Problems (2018)
5 pages
The Online Math Open Spring Contest Official Solutions March 22 - April 2, 2019
No ratings yet
The Online Math Open Spring Contest Official Solutions March 22 - April 2, 2019
18 pages
ARML Local 2020 Solutions
No ratings yet
ARML Local 2020 Solutions
24 pages
Week 24-ARML Team-Relay-Individual-Power Questions-2009
No ratings yet
Week 24-ARML Team-Relay-Individual-Power Questions-2009
5 pages
NYCIML - Junior - Contest - 1 - s23 - COMPLETE
No ratings yet
NYCIML - Junior - Contest - 1 - s23 - COMPLETE
10 pages
cdExamSolutions
No ratings yet
cdExamSolutions
6 pages
ARML2015
No ratings yet
ARML2015
40 pages
2024 MSSolutions
No ratings yet
2024 MSSolutions
9 pages
CMIMC_2025_Contest_AlgNT_Sols
No ratings yet
CMIMC_2025_Contest_AlgNT_Sols
8 pages
2015 cdExamSolutions
No ratings yet
2015 cdExamSolutions
5 pages
Singapore Mathematical Olympiads 2010 (By Tay Seng Et Al)
100% (1)
Singapore Mathematical Olympiads 2010 (By Tay Seng Et Al)
64 pages
Singapore Mathematical Olympiads (2010)
83% (6)
Singapore Mathematical Olympiads (2010)
64 pages
BBB 2021 Finals Primary 6 Set 2
100% (2)
BBB 2021 Finals Primary 6 Set 2
12 pages
solutions (1)
No ratings yet
solutions (1)
6 pages
d404b932-2184-486b-927b-7a5a08becec6
No ratings yet
d404b932-2184-486b-927b-7a5a08becec6
4 pages
2020 Eleventh - Hour - Mock - Solutions
No ratings yet
2020 Eleventh - Hour - Mock - Solutions
6 pages
HMM T February 2019 Guts Round Test
No ratings yet
HMM T February 2019 Guts Round Test
7 pages
FSJM QF 2020 Solutions
No ratings yet
FSJM QF 2020 Solutions
6 pages
AMC-10-Mock-2016-Final
No ratings yet
AMC-10-Mock-2016-Final
4 pages
Amc 10a 2005
No ratings yet
Amc 10a 2005
6 pages
8280 S_Alpha Bowl Solutions
No ratings yet
8280 S_Alpha Bowl Solutions
5 pages
P1 Jr IOQM Mock 25 Sol
No ratings yet
P1 Jr IOQM Mock 25 Sol
10 pages
Vtamps Worksheet Solution
No ratings yet
Vtamps Worksheet Solution
10 pages
HMMTNovember 2018 General Round Solutions
No ratings yet
HMMTNovember 2018 General Round Solutions
4 pages
High School Math Contest: University of South Carolina January 31st, 2015
No ratings yet
High School Math Contest: University of South Carolina January 31st, 2015
16 pages
Solutions
No ratings yet
Solutions
17 pages
Vtamps 18.0 SS Session 2
No ratings yet
Vtamps 18.0 SS Session 2
9 pages
VTAMPS 18 SS Set 2
No ratings yet
VTAMPS 18 SS Set 2
13 pages
Practice Makes Perfect in Geometry: Angles, Triangles and other Polygons
From Everand
Practice Makes Perfect in Geometry: Angles, Triangles and other Polygons
John Parnell
5/5 (1)
Practice Makes Perfect in Geometry: Angles, Triangles and other Polygons with Answers
From Everand
Practice Makes Perfect in Geometry: Angles, Triangles and other Polygons with Answers
John Parnell
No ratings yet
Practice Makes Perfect in Geometry: Three-Dimensional Figures
From Everand
Practice Makes Perfect in Geometry: Three-Dimensional Figures
John Parnell
No ratings yet
Worked Examples in Mathematics for Scientists and Engineers
From Everand
Worked Examples in Mathematics for Scientists and Engineers
G. Stephenson
No ratings yet
Algebraic Equations
From Everand
Algebraic Equations
Demetrios P. Kanoussis
No ratings yet
Origami Dots: Folding paper to explore geometry
From Everand
Origami Dots: Folding paper to explore geometry
Andy Parkinson
5/5 (1)
Key Of The Week - Fb Major
No ratings yet
Key Of The Week - Fb Major
1 page
LitProf 23
No ratings yet
LitProf 23
2 pages
College Research - Midland
No ratings yet
College Research - Midland
3 pages
Saturn Poster
No ratings yet
Saturn Poster
1 page
Current Events March 19
No ratings yet
Current Events March 19
5 pages
Daily Lesson Log of M9Al-Ia-B-1 Grade Level Learning Area Quarter I. Objectives
No ratings yet
Daily Lesson Log of M9Al-Ia-B-1 Grade Level Learning Area Quarter I. Objectives
3 pages
Enhanced Mathematics 7 - Q1 Q4 - SDO PASAY
No ratings yet
Enhanced Mathematics 7 - Q1 Q4 - SDO PASAY
6 pages
0208231623255664436
No ratings yet
0208231623255664436
2 pages
2025 Old Mutual Olympiad Test 1-1
No ratings yet
2025 Old Mutual Olympiad Test 1-1
5 pages
Advanced Mathematics - 1
No ratings yet
Advanced Mathematics - 1
4 pages
Ch.9 Practice Tests
No ratings yet
Ch.9 Practice Tests
8 pages
Excel Review Center ECE Refresher/Coaching Course Math TH 1: (Ans.) (Ans.)
No ratings yet
Excel Review Center ECE Refresher/Coaching Course Math TH 1: (Ans.) (Ans.)
3 pages
Simplification - Square, Cube - Theory
No ratings yet
Simplification - Square, Cube - Theory
3 pages
Pub Factorization Unique and Otherwise
No ratings yet
Pub Factorization Unique and Otherwise
262 pages
Gr8_Checkpoint_Rev_Wrsht_2
No ratings yet
Gr8_Checkpoint_Rev_Wrsht_2
9 pages
Grade 8 Ch7 Parent
No ratings yet
Grade 8 Ch7 Parent
78 pages
RRB Group D Syllabus - Docx 9f92dafc
No ratings yet
RRB Group D Syllabus - Docx 9f92dafc
25 pages
MATH2089 NM Lectures Topic1
No ratings yet
MATH2089 NM Lectures Topic1
14 pages
C 03 Quadratic Relationships
No ratings yet
C 03 Quadratic Relationships
68 pages
PBS Maths Form 2 Chapter 2
25% (4)
PBS Maths Form 2 Chapter 2
3 pages
WK7 Largado G7 DLL
No ratings yet
WK7 Largado G7 DLL
7 pages
AHsd - Important Legal
No ratings yet
AHsd - Important Legal
6 pages
Lesson 22: Multiplying and Dividing Expressions With Radicals
No ratings yet
Lesson 22: Multiplying and Dividing Expressions With Radicals
16 pages
Chapter 21 I Variations ENHANCE
No ratings yet
Chapter 21 I Variations ENHANCE
21 pages
Heron
No ratings yet
Heron
2 pages
Notes Algebraic Expressions
No ratings yet
Notes Algebraic Expressions
2 pages
Math-7 Q1 SLM WK6
No ratings yet
Math-7 Q1 SLM WK6
14 pages
Maths Methods 2023 Unit 1 Trial Exam 1 Solutions
No ratings yet
Maths Methods 2023 Unit 1 Trial Exam 1 Solutions
5 pages
L1 - SR Oly GIF
No ratings yet
L1 - SR Oly GIF
10 pages
Wk1 Day 1 4 Prototype Lesson Plan in Math 9
No ratings yet
Wk1 Day 1 4 Prototype Lesson Plan in Math 9
13 pages
CSEET - Quantitative Aptitude Edition II - CS Vaibhav Chitlangia - Yes Academy, Pune
No ratings yet
CSEET - Quantitative Aptitude Edition II - CS Vaibhav Chitlangia - Yes Academy, Pune
44 pages
Simple Practice Problems On Numbers-1
No ratings yet
Simple Practice Problems On Numbers-1
3 pages
enoch2TheoremItaly1
No ratings yet
enoch2TheoremItaly1
11 pages
Multiplication & Division of Radicals
No ratings yet
Multiplication & Division of Radicals
7 pages
4-4 Study Guide and Intervention: Complex Numbers
No ratings yet
4-4 Study Guide and Intervention: Complex Numbers
2 pages