Solutions Pamphlet

MAA American Mathematics Competitions

69th Annual

AMC 12A
American Mathematics Competition 12A
Wednesday, February 7, 2018

This Pamphlet gives at least one solution for each problem on this year’s competition
and shows that all problems can be solved without the use of a calculator. When more
than one solution is provided, this is done to illustrate a significant contrast in methods,
e.g., algebraic versus geometric, computational versus conceptual, elementary versus
advanced. These solutions are by no means the only ones possible, nor are they superior
to others the reader may devise.
We hope that teachers will inform their students about these solutions, both as
illustrations of the kinds of ingenuity needed to solve nonroutine problems and as
examples of good mathematical exposition. However, the publication, reproduction, or
communication of the problems or solutions for this contest during the period when students
are eligible to participate seriously jeopardizes the integrity of the results. Dissemination
at any time via copier, telephone, email, internet, or media of any type is a violation of
the competition rules.
Correspondence about the problems/solutions for this AMC 12 and orders for any
publications should be addressed to:
MAA American Mathematics Competitions
Attn: Publications, PO Box 471, Annapolis Junction, MD 20701
Phone 800.527.3690 | Fax 240.396.5647 | [email protected]
The problems and solutions for this AMC 12 were prepared by
MAA’s Subcommittee on the AMC10/AMC12 Exams, under the direction of the
co-chairs Jerrold W. Grossman and Carl Yerger.
© 2018 Mathematical Association of America
2018 AMC 12A Solutions 2

1. Answer (D): There are currently 36 red balls in the urn. In order
for the 36 red balls to represent 72% of the balls in the urn after some
blue balls are removed, there must be 36 ÷ 0.72 = 50 balls left in the
urn. This requires that 100 − 50 = 50 blue balls be removed.

2. Answer (C): The 5-pound rocks have a value of $14 ÷ 5 = $2.80

per pound; the 4-pound rocks have a value of $11 ÷ 4 = $2.75 per
pound; the 1-pound rocks have a value of $2 per pound. It is not to
Carl’s advantage to take 1-pound rocks when he can take the larger
rocks. Therefore the only issue is how many of the more valuable
5-pound rocks to take, including as many 4-pound rocks as possible
in each case. The viable choices are displayed in the following table.

5-pound rocks 4-pound rocks 1-pound rocks value

($14 each) ($11 each) ($2 each)
3 0 3 $48
2 2 0 $50
1 3 1 $49
0 4 2 $48

The maximum possible value is $50.

Note: Although the 5-pound rocks are the most valuable per pound,
it was not to Carl’s advantage to take as many of them as possible.
This situation is an example of the classic knapsack problem for which
the so-called “greedy algorithm” is not optimal.

3. Answer (E): There are 4 choices for the periods in which the
mathematics courses can be taken: periods 1, 3, 5; periods 1, 3, 6;
periods 1, 4, 6; and periods 2, 4, 6. Each choice of periods allows
3! = 6 ways to order the 3 mathematics courses. Therefore there are
4 · 6 = 24 ways of arranging a schedule.

4. Answer (D): Because the statements of Alice, Bob, and Charlie

are all incorrect, the actual distance d satisfies d < 6, d > 5, and
d > 4. Hence the actual distance lies in the interval (5, 6).

5. Answer (E): Factoring x2 − 3x + 2 as (x − 1)(x − 2) shows that its

roots are 1 and 2. If 1 is a root of x2 − 5x + k, then 12 − 5 · 1 + k = 0
and k = 4. If 2 is a root of x2 − 5x + k, then 22 − 5 · 2 + k = 0 and
k = 6. The sum of all possible values of k is 4 + 6 = 10.
2018 AMC 12A Solutions 3

6. Answer (B): Note that the given conditions imply that the 6 values
are listed in increasing order. Because the median of the these 6 values
is n, the mean of the middle two values must be n, so

(m + 10) + (n + 1)
= n,
2
which implies m = n − 11. Because the mean of the set is also n,

(n − 11) + (n − 7) + (n − 1) + (n + 1) + (n + 2) + 2n
= n,
6
so 7n−16 = 6n and n = 16. Then m = 16−11 = 5, and the requested
sum is 5 + 16 = 21.

7. Answer (E): Because 4000 = 25 · 53 ,

 n
2
4000 · = 25+n · 53−n .
5

This product will be an integer if and only if both of the factors 25+n
and 53−n are integers, which happens if and only if both exponents
are nonnegative. Therefore the given expression is an integer if and
only if 5 + n ≥ 0 and 3 − n ≥ 0. The solutions are exactly the integers
satisfying −5 ≤ n ≤ 3. There are 3 − (−5) + 1 = 9 such values.

8. Answer (E): The length of the base DE of 4ADE is 4 times

the length of the base of a small triangle, so the area of 4ADE is
42 · 1 = 16. Therefore the area of DBCE is the area of 4ABC minus
the area of 4ADE, which is 40 − 16 = 24.

9. Answer (E): If 0 ≤ x ≤ π and 0 ≤ y ≤ π, then sin(x) ≥ 0,

sin(y) ≥ 0, cos(x) ≤ 1, and cos(y) ≤ 1. Therefore

sin(x + y) = sin(x) · cos(y) + cos(x) · sin(y) ≤ sin(x) + sin(y).

The given inequality holds for all y such that 0 ≤ y ≤ π.

10. Answer (C): The graph of the system is shown below.

2018 AMC 12A Solutions 4

–4 –2 2 4

–2

–4

The graph of the first equation is a line with x-intercept (3, 0) and
y-intercept (0, 1). To draw the graph of the second equation, consider
the equation quadrant by quadrant. In the first quadrant x > 0 and
y > 0, and thus the second equation is equivalent to |x−y| = 1, which
in turn is equivalent to y = x ± 1. Its graph consists of the rays with
endpoints (0, 1) and (1, 0), as shown. In the second quadrant x < 0
and y > 0. The corresponding graph is the reflection of the first quad-
rant graph across the y-axis. The rest of the graph can be sketched by
further reflections of the first-quadrant graph across the coordinate
axes, resulting in the figure shown. There are 3 intersection points:
(−3, 2), (0, 1), and 32 , 12 , as shown.

OR

The second equation implies that x = y ± 1 or x = −y ± 1. There are

four cases:
• If x = y + 1, then (y + 1) + 3y = 3, so (x, y) = 23 , 12 .

• If x = y − 1, then (y − 1) + 3y = 3, so (x, y) = (0, 1).

• If x = −y + 1, then (−y + 1) + 3y = 3, so again (x, y) = (0, 1).
• If x = −y − 1, then (−y − 1) + 3y = 3, so (x, y) = (−3, 2).
It may be checked that each of these ordered pairs actually satisfies
the given equations, so the total number of solutions is 3.

11. Answer (D): The paper’s long edge AB is the hypotenuse of right
triangle ACB, and the crease lies along the perpendicular bisector
of AB. Because AC > BC, the crease hits AC rather than BC. Let
D be the midpoint of AB, and let E be the intersection of AC and the
line through D perpendicular to AB. Then the crease in the paper
2018 AMC 12A Solutions 5
DE CB 3
is DE. Because 4ADE ∼ 4ACB, it follows that AD = AC = 4.
Thus
CB 5 3 15
DE = AD · = · = .
AC 2 4 8

5
D
3

A C
4 E

12. Answer (C): If 1 ∈ S, then S can have only 1 element, not 6

elements. If 2 is the least element of S, then 2, 3, 5, 7, 9, and 11 are
available to be in S, but 3 and 9 cannot both be in S, so the largest
possible size of S is 5. If 3 is the least element, then 3, 4, 5, 7, 8, 10,
and 11 are available, but at most one of 4 and 8 can be in S and at
most one of 5 and 10 can be in S, so again S has size at most 5. The
set S = {4, 6, 7, 9, 10, 11} has the required property, so 4 is the least
possible element of S.

OR

At most one integer can be selected for S from each of the following
6 groups: {1, 2, 4, 8}, {3, 6, 12}, {5, 10}, {7}, {9}, and {11}. Because
S consists of 6 distinct integers, exactly one integer must be selected
from each of these 6 groups. Therefore 7, 9, and 11 must be in S.
Because 9 is in S, 3 must not be in S. This implies that either 6 or
12 must be selected from the second group, so neither 1 nor 2 can be
selected from the first group. If 4 is selected from the first group, the
collection of integers {4, 5, 6, 7, 9, 11} is one possibility for the set S.
Therefore 4 is the least possible element of S.
Note: The two collections given in the solutions are the only ones
with least element 4 that have the given property. This problem is
a special case of the following result of Paul Erdős from the 1930s:
Given n integers a1 , a2 , . . . , an , no one of them dividing any other,
with a1 < a2 < · · · < an ≤ 2n, let the integer k be determined by the
inequalities 3k < 2n < 3k+1 . Then a1 ≥ 2k , and this bound is sharp.
2018 AMC 12A Solutions 6

13. Answer (D): Let S be the set of integers, both negative and non-
negative, having the given form. Increasing the value of ai by 1 for
0 ≤ i ≤ 7 creates a one-to-one correspondence between S and the
ternary (base 3) representation of the integers from 0 through 38 − 1,
so S contains 38 = 6561 elements. One of those is 0, and by symmetry,
half of the others are positive, so S contains 1 + 12 · (6561 − 1) = 3281
elements.

OR

First note that if an integer N can be written in this form, then

N − 1 can also be written in this form as long as not all the ai in
the representation of N are equal to −1. A procedure to alter the
representation of N so that it will represent N − 1 instead is to find
the least value of i such that ai 6= −1, reduce the value of that ai by 1,
and set ai = 1 for all lower values of i. By the formula for the sum of
a finite geometric series, the greatest integer that can be written in
the given form is
38 − 1
= 3280.
3−1
Therefore, 3281 nonnegative integers can be written in this form,
namely all the integers from 0 through 3280, inclusive. (The negative
integers from −3280 through −1 can also be written in this way.)

OR

Think of the indicated sum as an expansion in base 3 using “digits”

−1, 0, and 1. Note that the leftmost digit ak of any positive integer
that can be written in this form cannot be negative and therefore
must be 1. Then there are 3 choices for each of the remaining k digits
to the right of ak , resulting in 3k positive integers that can be written
in the indicated form. Thus there are
7
X 38 − 1
3k = = 3280
3−1
k=0

positive numbers of the indicated form. Because 0 can also be written

in this form, the number of nonnegative integers that can be written
in the indicated form is 3281.
2018 AMC 12A Solutions 7

14. Answer (D): By the change-of-base formula, the given equation is

equivalent to
log 4 log 8
=
log 3x log 2x
2 log 2 3 log 2
=
log 3 + log x log 2 + log x
2 log 2 + 2 log x = 3 log 3 + 3 log x
log x = 2 log 2 − 3 log 3
4
log x = log .
27
4
Therefore x = 27 , and the requested sum is 4 + 27 = 31.

OR

Changing to base-2 logarithms transforms the given equation into

2 3
=
log2 3x log2 2x
2 log2 2x = 3 log2 3x
log2 (2x)2 = log2 (3x)3
(2x)2 = (3x)3 ,
4
so x = 27 , and the requested sum is 4 + 27 = 31.

15. Answer (B): None of the squares that are marked with dots in
the sample scanning code shown below can be mapped to any other
marked square by reflections or non-identity rotations. Therefore
these 10 squares can be arbitrarily colored black or white in a symmet-
ric scanning code, with the exception of “all black” and “all white”.
On the other hand, reflections or rotations will map these squares to
all the other squares in the scanning code, so once these 10 colors
are specified, the symmetric scanning code is completely determined.
Thus there are 210 − 2 = 1022 symmetric scanning codes.
2018 AMC 12A Solutions 8

OR

The diagram below shows the orbits of each square under rotations
and reflections. Because the scanning code must look the same under
these transformations, all squares in the same orbit must get the same
color, but one is free to choose the color for each orbit, except for the
choice of “all black” and “all white”. Because there are 10 orbits,
there are 210 − 2 = 1022 symmetric scanning codes.

A B C D C B A
B E F G F E B
C F H I H F C
D G I J I G D
C F H I H F C
B E F G F E B
A B C D C B A

16. Answer (E): Solving the second equation for x2 gives x2 = y + a,

and substituting into the first equation gives y 2 + y + (a − a2 ) = 0.
The polynomial in y can be factored as (y + (1 − a))(y + a), so the
solutions are y = a − 1 and y = −a. (Alternatively, the solutions
can be obtained using the quadratic formula.) The corresponding
equations for x are x2 = 2a − 1 and x2 = 0. The second equation
always has the solution x = 0, corresponding to the point of tangency
at the vertex of the parabola y = x2 − a. The first equation has
2 solutions if and only if a > 21 , corresponding to the 2 symmetric
intersection points of the parabola with the circle. Thus the two
curves intersect at 3 points if and only if a > 12 .
2018 AMC 12A Solutions 9

OR

Substituting the value for y from the second equation into the first
equation yields

2
x2 + x2 − a = a2 ,
which is equivalent to
x2 x2 − (2a − 1) = 0.

The first factor gives the solution x = 0, and the second factor gives
2 other solutions if a > 12 and no other solutions if a ≤ 12 . Thus there
are 3 solutions if and only if a > 12 .

17. Answer (D): Let the triangle’s vertices in the coordinate plane be
(4, 0), (0, 3), and (0, 0), with [0, s] × [0, s] representing the unplanted
portion of the field. The equation of the hypotenuse is 3x+4y−12 = 0,
so the distance from (s, s), the corner of S closest to the hypotenuse,
to this line is given by
|3s + 4s − 12|
√ .
32 + 42
Setting this equal to 2 and solving for s gives s = 22 2
7 and s = 7 , and
the former is rejected because the square must lie within the triangle.

2
The unplanted area is thus 72 = 49 4
, and the requested fraction is
4
49 145
1− 1 = .
2 ·4·3 147

OR

Let the given triangle be described as 4ABC with the right angle
at B and AB = 3. Let D be the vertex of the square that is in the
interior of the triangle, and let s be the edge length of the square.
Then two sides of the square along with line segments AD and CD
decompose 4ABC into four regions. These regions are a triangle
with base 5 and height 2, the unplanted square with side s, a right
triangle with legs s and 3 − s, and a right triangle with legs s and
4 − s. The sum of the areas of these four regions is
1 1 1 7
· 5 · 2 + s2 + s(3 − s) + s(4 − s) = 5 + s,
2 2 2 2
and the area of 4ABC is 6. Solving 5 + 72 s = 6 for s gives s = 2
7,
and the solution concludes as above.
2018 AMC 12A Solutions 10

18. Answer (D): Because AB is 56 of AB + AC, it follows from the

Angle Bisector Theorem that DF is 56 of DE, and BG is 65 of BC.
Because trapezoids F DBG and EDBC have the same height, the
area of F DBG is 65 of the area of EDBC. Furthermore, the area of
4ADE is 14 of the area of 4ABC, so its area is 30, and the area of
trapezoid EDBC is 120−30 = 90. Therefore the area of quadrilateral
F DBG is 56 · 90 = 75.

D F E

G
B C

Note: The figure (not drawn to scale) shows the situation in which
∠ACB is acute. In this case BC ≈ 59.0 and ∠BAC ≈ 151◦ . It is also
possible for ∠ACB to be obtuse, with BC ≈ 41.5 and ∠BAC ≈ 29◦ .
These values can be calculated using the Law of Cosines and the sine
formula for area.

19. Answer (C): Elements of set A are of the form 2i · 3j · 5k for

nonnegative integers i, j, and k. Note that the product
   
1 1 1 1 1 1
1 + + 2 + ··· 1 + + 2 + ··· 1 + + 2 + ···
2 2 3 3 5 5
will produce the desired sum. By the formula for infinite geometric
series, this product evaluates to
1 1 1 3 5 15
1 · 1 · 1 =2· · = .
1− 2 1− 3 1− 5
2 4 4
The requested sum is 15 + 4 = 19.

20. Answer √ (D): It follows from the Pythagorean Theorem that CM =

M B = 32 2. Because quadrilateral AIM E is cyclic, opposite angles
are supplementary and thus ∠IM E is a right angle. Let x = CI and
y = BE; then AI = 3 − x and AE = 3 − y. By the Law of Cosines in
4M CI,
2
3√ 3√

9
IM 2 = x2 + 2 −2·x· 2 · cos 45◦ = x2 − 3x + .
2 2 2
2018 AMC 12A Solutions 11

Similarly, M E 2 = y 2 − 3y + 92 . By the Pythagorean Theorem in right

triangles EM I and IAE,
   
2 9 2 9
x − 3x + + y − 3y + = (3 − x)2 + (3 − y)2 ,
2 2
which simplifies to x + y = 3. Because the area of 4EM I is 2, it
follows that IM 2 · M E 2 = 16. Therefore
  
9 9
x2 − 3x + (3 − x)2 − 3(3 − x) + = 16,
2 2

2
which simplifies to x2 − 3x + 92 = 16. Because y > x, the only real
√
solution is x = 3−2 7 . The requested sum is 3 + 7 + 2 = 12.

C
x
I

3–x

3–y y
A E B

OR

Place the figure in the coordinate plane with A at (0, 0), B at (3, 0),
and C at (0, 3). Then M is at 32 , 32 . Let s = AE and t = CI. Then

the coordinates of E are (s, 0), and the coordinates of I are (0, 3 − t).
Because AIM E is a cyclic quadrilateral and ∠EAI is a right angle,
∠IM E is a right angle. Therefore M I and M E are perpendicular,
so the product of their slopes is
3 3
2 t− 2
3 · 3 = −1;
2 −s 2

this equation simplifies to s = t. Then, with brackets indicating area,

[ABC] = [CIM ] + [BM E] + [AEI] + [IM E]

9 1 3 1 3 1
= · · t + · · (3 − t) + · t · (3 − t) + 2,
2 2 2 2 2 2
2018 AMC 12A Solutions 12
√
3± 7
which simplifiles to 2t2 − 6t + 1 = 0. Therefore t = 2 ,
and
√
because AI > AE, the length of CI is 3−2 7 and the requested sum
is 3 + 7 + 2 = 12.

21. Answer (B): By Descartes’ Rule of Signs, none of these polynomi-

als has a positive root, and each one has exactly one negative root.
Because each polynomial is positive at x = 0 and negative at x = −1,
it follows that each has exactly one root between −1 and 0. Note also
that each polynomial is increasing throughout the interval (−1, 0).
Because x19 > x17 for all x in the interval (−1, 0), it follows that the
polynomial in choice A is greater than the polynomial in choice B
on that interval, which implies that the root of the polynomial in
choice A is less than the root of the polynomial in choice B. Because
x13 > x11 for all x in the interval (−1, 0), it follows that the polyno-
mial in choice C is greater than the polynomial in choice A on that
interval, which implies that the root of the polynomial in choice C
is less than the root of the polynomial in choice A and therefore less
than the root of the polynomial in choice B. The same reasoning
shows that the root of the polynomial in choice D is less than the
root of the polynomial in choice B.
Furthermore, 2018 > 2018x6 on the interval (−1, 0), so x6 + 2018 >
2019x6 , from which it follows that x11 (x6 + 2018) < 2019x17 . There-
fore the polynomial in choice B is less than 2019x17 + 1 on the in-
1


terval (−1, 0). The polynomial in choice E has root − 1 − 2019 .
Bernoulli’s Inequality shows that (1 + x)17 > 1 + 17x for all x > −1,
which implies that
 17  
1 17
−2019 1 − + 1 < −2019 1 − + 1 = −2001 < 0,
2019 2019
so the polynomial in choice B is negative at the root of the polynomial
in choice E. This shows that the root of the polynomial in choice B
is greater than the root in choice E.
Because the unique real root of the polynomial in choice B is greater
than the unique root of the polynomial in each of the other choices,
that polynomial has the greatest real root.

22. Answer (A): Let z = a+bi be a solution of the first

√ equation, where
a and b are real numbers. Then (a + bi)2 = 4 + 4 15i. Expanding
the left-hand side and equating real and imaginary parts yields
√
a2 − b2 = 4 and 2ab = 4 15.
2018 AMC 12A Solutions 13
√
2 15
From the second equation, b = a , and substituting this into the
first equation and simplifying gives (a ) −4a2 −60 = 0, which factors
2 2
√
as (a2 − 10)(a2 + 6) = 0. Because a is real, √ it follows that a = ± 10,
from which it then√follows √ that b =√± 6.√Thus two vertices of the
parallelogram are 10 + 6i and − 10 − 6i. A similar calculation
with the other given √ equation shows
√ that the other two vertices of the
parallelogram are 3 + i and − 3 − i. The area of this parallogram
can be computed using the shoelace formula, which gives the area of
a polygon in terms of the coordinates of its vertices (x1 , y1 ), (x2 , y2 ),
. . . , (xn , yn ) in clockwise or counter-clockwise order:

1
· (x1 y2 + x2 y3 + · · · + xn−1 yn + xn y1 )
2

− (y1 x2 + y2 x3 + · · · + yn−1 xn + yn x1 ) .

√ √ √ √
In this √
case x1 = √10, y1 = 6, x2 = 3, y2 = √1, x3 = √ − 10,
y3 = − 6, x4 = − 3, and y4 = −1. The area is 6 2 − 2 10, and
the requested sum of the four positive integers in this expression is
20.

OR
√
The solutions of z 2 = 4 + 4 15i = 16 cis√ 2θ1 are z1 = 4 cis θ1 and its
opposite, with 0 < θ1 < π4 and tan 2θ1 = 15. Then cos 2θ1 = 41 , and
√ √
by the half-angle identities, cos θ1 = 410 and sin θ1 = 46 . Similarly,
√
the solutions of z 2 = 2 + 2 3i = 4 cis θ√ 2 are z2 = 2 cis θ2 √ and its
opposite, with 0 < θ2 < π4 and tan 2θ2 = 3. Then cos θ2 = 23 and
sin θ2 = 12 .
The area of the parallelogram in the complex plane with vertices z1 ,
z2 , and their opposites is 4 times the area of the triangle with vertices
0, z1 , and z2 , and because the area of a triangle is one-half the product
of the lengths of two of its sides and the sine of their included angle,
it follows that the area of the parallelogram is
 
1
4 · 4 · 2 · sin (θ1 − θ2 ) = 16 (sin θ1 cos θ2 − cos θ1 sin θ2 )
2
√ √ √ !
6 3 10 1
= 16 · − ·
4 2 4 2
√ √
= 6 2 − 2 10.

Therefore, p + q + r + s = 6 + 2 + 2 + 10 = 20.
2018 AMC 12A Solutions 14

23. Answer (E): Extend P N through N to Q so that P N = N Q.

Segments U G and P Q bisect each other, implying that U P GQ is
a parallelogram. Therefore GQ k P T , so ∠QGA = 180◦ − ∠T =
∠T P A + ∠T AP = 36◦ + 56◦ = 92◦ . Furthermore GQ = P U = AG,
so 4QGA is isosceles, and ∠QAG = 21 (180◦ − 92◦ ) = 44◦ . Because
M N is a midline of 4QP A, it follows that M N k AQ and
∠N M P = ∠QAP = ∠QAG + ∠GAP = 44◦ + 56◦ = 100◦ ,
so acute ∠N M A = 80◦ . (Note that the value of the common length
P U = AG is immaterial.)
T

Q
N
U G
P A
M

OR

Place the figure in the coordinate plane with P = (−5, 0), M = (0, 0),
A = (5, 0), and T in the first quadrant. Then
U = (−5 + cos 36◦ , sin 36◦ ) and G = (5 − cos 56◦ , sin 56◦ ),
and the midpoint N of U G is
 
1 1
(cos 36◦ − cos 56◦ ), (sin 36◦ + sin 56◦ ) .
2 2
The tangent of ∠N M A is the slope of line M N , which is calculated
as follows using the sum-to-product trigonometric identites:
sin 36◦ + sin 56◦
tan(∠N M A) =
cos 36◦ − cos 56◦
◦ ◦ ◦ ◦
2 sin 36 +56
2 cos 36 −56
2
= ◦ ◦ ◦ ◦
−2 sin 36 +562 sin 36 −56
2
cos 10◦
= = cot 10◦ = tan 80◦ ,
sin 10◦
and it follows that ∠N M A = 80◦ .
2018 AMC 12A Solutions 15

24. Answer (B): Because Alice and Bob are choosing their numbers
uniformly at random, the cases in which two or three of the chosen
numbers are equal have probability 0 and can be ignored. Suppose
Carol chooses the number c. She will win if her number is greater than
Alice’s number and less than Bob’s, and she will win if her number
is less than Alice’s number and greater than Bob’s. There are three
cases.
• If c ≤ 21 , then Carol’s number is automatically less than Bob’s, so
her chance of winning is the probability that Alice’s number is less
than c, which is just c. The best that Carol can do in this case is to
choose c = 21 , in which case her chance of winning is 21 .
• If c ≥ 23 , then Carol’s number is automatically greater than Bob’s,
so her chance of winning is the probability that Alice’s number is
greater than c, which is just 1 − c. The best that Carol can do in this
case is to choose c = 32 , in which case her chance of winning is 13 .
• Finally suppose that 21 < c < 2
3. The probability that Carol’s
number is less than Bob’s is
2
3 −c
2 = 4 − 6c,
3 − 12
so the probability that her number is greater than Alice’s and less
than Bob’s is c(4 − 6c). Similarly, the probability that her number
is less than Alice’s and greater than Bob’s is (1 − c)(6c − 3). Carol’s
probability of winning in this case is therefore

c(4 − 6c) + (1 − c)(6c − 3) = −12c2 + 13c − 3.

The value of a quadratic polynomial with a negative coefficient on its

quadratic term is maximized at −b 2a , where a is the coefficient on its
quadratic term and b is the coefficient on its linear term; here that is
when c = 13 1 2
24 , which is indeed between 2 and 3 . Her probability of
winning is then
 2
13 13 25 24 1
−12 · + 13 · −3= > = .
24 24 48 48 2

Because the probability of winning in the third case exceeds the prob-
abilities obtained in the first two cases, Carol should choose 13
24 .

25. Answer (D): The equation Cn − Bn = A2n is equivalent to

2
102n − 1 10n − 1
 n
10 − 1
c· −b· = a2 .
9 9 9
2018 AMC 12A Solutions 16

Dividing by 10n − 1 and clearing fractions yields

(9c − a2 ) · 10n = 9b − 9c − a2 .

As this must hold for two different values n1 and n2 , there are two
such equations, and subtracting them gives

9c − a2 (10n1 − 10n2 ) = 0.

The second factor is non-zero, so 9c−a2 = 0 and thus 9b−9c−a2 = 0.

2
From this it follows that c = a3 and b = 2c. Hence digit a must
be 3, 6, or 9, with corresponding values 1, 4, or 9 for c, and 2, 8, or
18 for b. The case b = 18 is invalid, so there are just two triples of
possible values for a, b, and c, namely (3, 2, 1) and (6, 8, 4). In fact,
in these cases, Cn − Bn = A2n for all positive integers n; for example,
4444 − 88 = 4356 = 662 . The second triple has the greater coordinate
sum, 6 + 8 + 4 = 18.

Problems and solutions were contributed by Risto Atanasov, George

Bauer, Chris Bolognese, Ivan Borsenco, Thomas Butts, Barbara Cur-
rier, Steven Davis, Steve Dunbar, Marta Eso, Zuming Feng, Zachary
Franco, Peter Gilchrist, Jerrold Grossman, Chris Jeuell, Jonathan
Kane, Joe Kennedy, Michael Khoury, Hugh Montgomery, Mohamed
Omar, Albert Otto, Joachim Rebholz, David Wells, Barry Weng, and
Carl Yerger.
 The
MAA American Mathematics Competitions
are supported by
Innovator’s Circle
The D. E. Shaw Group
Susquehanna International Group
Tudor Investment Corporation
Two Sigma
Winner’s Circle
MathWorks
Achiever’s Circle
Art of Problem Solving
Jane Street Capital
Sustainer’s Circle
American Mathematical Society
Ansatz Capital
Army Educational Outreach Program
Collaborator’s Circle
American Statistical Association
Casualty Actuarial Society
Conference Board of the Mathematical Sciences
Mu Alpha Theta
Society for Industrial and Applied Mathematics