Lecture I-II: Motivation and Decision Theory

Markus M. Möbius
February 15, 2005

1 Two Motivating Experiments

Experiment 1 Each of you (the students in this course) have to declare an
integer between 0 and 100 to guess ”2/3 of the average of all the responses”.
More precisely, each student who guesses the highest integer which is not
higher than 2/3 of the average of all responses, will receive a prize of 10
Dollars.

How should you play this game? A naive guess would be that other players
choose randomly a strategy. In that case the mean in the game would be
around 50 and you should choose 33. But you realize that other players make
the same calculation - so nobody should say a number higher than 33. That
means that you should not name a number greater than 22 - and so on. The
winning number was 13. That means that people did this iteration about 3
times. But in fact, the stated numbers were all over the place - ranging from
0 to 40. That means that different students had different estimates of what
their fellow students would do.

• Being aware of your fellow players’ existence and trying to anticipate

their moves is called strategic behavior. Game theory is mainly about
designing models of strategic behavior.

• In this game, the winner has to correctly guess how often his fellow
players iterate. Assuming infinite iterations would be consistent but
those who bid 0 typically lose badly. Guessing higher numbers can
mean two things: (a) the player does not understand strategic behavior
or (b) the player understands strategic behavior but has low confidence

1
 in the ability of other players to understand that this is a strategic
game. Interestingly, most people knew at least one other person in the
class (hence there was at least some degree of what a game theorist
would call common knowledge of rationality).1

Experiment 2 I am going to auction off $20. Each of your can bid secretly
on the book and the highest bidder wins the auction. However, all of you have
to pay your bid regardless of whether you win or lose.

In this game there is no optimal single bid for all players. You can check
that for all cases where each player i bids some fixed bid bi at least one of
the players will regret her decision and try to reverse it - we say that there
is no pure strategy Nash equilibrium in this game. Consider for example the
case where all player bid 55 Dollars. Then some player should bid 55 and 5
cents. No equilibrium!
There is an equilibrium if we allow players to randomize. You can check
that with two players who pick random numbers between 0 and 60 with equal
probability no player would want to change her pick - all picks will give her
zero profit in expectation.
The more players there are, the more the bid distribution is skewed to-
wards 0 (check)! We will formally discuss mixed strategy Nash equilibria in
a few lectures time.

2 What is game theory?

Definition 1 Game theory is a formal way to analyze interaction among a
group of rational agents who behave strategically.

This definition contains a number of important concepts which are dis-

cussed in order:
Group: In any game there is more than one decision maker who is
referred to as player. If there is a single player the game becomes a decision
problem.
1
There is common knowledge of rationality between players A and B if A knows that
B is rational (and vice versa), if A knows that B knows that A is rational (and vice versa)
etc.

2
 Interaction: What one individual player does directly affects at least
one other player in the group. Otherwise the game is simple a series of
independent decision problems.
Strategic: Individual players account for this interdependence.
Rational: While accounting for this interdependence each player chooses
her best action. This condition can be weakened and we can assume that
agents are boundedly rational. Behavioral economics analyzes decision prob-
lems in which agents behave boundedly rational. Evolutionary game theory
is game theory with boundedly rational agents.

Example 1 Assume that 10 people go into a restaurant. Every person pays

for her own meal. This is a decision problem. Now assume that everyone
agrees before the meal to split the bill evenly amongst all 10 participants. Now
we have a game.

Game theory has found numerous applications in all fields of economics:

1. Trade: Levels of imports, exports, prices depend not only on your own
tariffs but also on tariffs of other countries.

2. Labor: Internal labor market promotions like tournaments: your chances

depend not only on effort but also on efforts of others.

3. IO: Price depends not only on your output but also on the output of
your competitor (market structure ...).

4. PF: My benefits from contributing to a public good depend on what

everyone else contributes.

5. Political Economy: Who/what I vote for depends on what everyone

else is voting for.

3 Decision Theory under Certainty

It makes sense to start by discussing trivial games - those we play against
ourselves, e.g. decision problems. Agents face situations in which they have
to make a choice. The actions of other agents do not influence my preference
ordering over those choices - therefore there is no strategic interaction going
on. Proper games will be discussed in the next lectures.

3
 A decision problem (A, ¹) consists of a finite set of outcomes A = {a1 , a2 , .., an }
and a preference relation ¹. The expression a ¹ b should be interpreted as
”b is at least as good as a”. We expect the preference relation to fulfill two
simple axioms:

Axiom 1 Completeness. Any two outcomes can be ranked, e.g. a ¹ b or

b ¹ a.

Axiom 2 Transitivity implies that if a ≥ b and b ≥ c then a ≥ c.

Both axioms ensure that all choices can be ordered in a single chain without
gaps (axiom 1) and without cycles (axiom 2).
Although the preference relation is the basic primitive of any decision
problem (and generally observable) it is much easier to work with a consistent
utility function u : A → < because we only have to remember n real numbers
{u1 , u2 , .., un }.

Definition 2 A utility function u : A → < is consist with the preference

relationship of a decision problem (A, ¹) if for all a, b ∈ A:

a¹b if and only if u (a) ≤ u (b)

Theorem 1 Assume the set of outcomes is finite. Then there exists a utility
function u which is consistent.

Proof: The proof is very simple. Simple collect all equivalent outcomes in
equivalence classes. There are finitely many of those equivalence classes
since there are only finitely many outcomes. Then we can order these
equivalence classes in a strictly increasing chain due to completeness
and transitivity.

Note that the utility function is not unique. In fact, any monotonic
transformation of a consistent utility function gives another utility function
which is also consistent.
We can now define what a rational decision maker is.

Definition 3 A rational decision maker who faces a decision problem (A, ¹)

chooses an outcome a∗ ∈ A which maximizes his utility (or, equivalently, for
each a ∈ A we have a ¹ a∗ ).

4
Remark 1 When there are infinitely many choices we want to make sure
that there is a continuous utility function. This requires one more axiom
which makes sure that preferences are continuous. For that purpose, one has
to define topology on the set of outcomes. We won’t deal with that since we
won’t gain much insight from it.

4 Decision Theory under Uncertainty

Lotteries are defined over the of outcomes A (which is again assumed to be
finite to keep things simple).

DefinitionP4n A simple lottery is defined as the set {(a1 , p1 ) , (a2 , p2 ) , .. (an , pn )}

such that i=1 pi = 1 and 0 ≤ pi ≤ 1. In a simple lottery the outcome ai
occurs with probability pi .

When there are up to three outcomes we can conveniently describe the

set of lotteries in a graphical way (see triangle).
Under certainty the preference relationship can still be written down ex-
plicitly for finite A (simply write down all of the n(n+1) 2
rankings). Under
uncertainty there are suddenly infinitely many lotteries. This poses two
problems. First of all, it’s impractical to write a large number of lottery
comparisons down. A second (and deeper) point is the observation that the
preference relationship is in principle unobservable because of the infinite
number of necessary comparisons.
John von Neumann and Oscar Morgenstern showed that under some ad-
ditional restrictions on preferences over lotteries there exists a utility function
over outcomes such that the expected utility of a lottery provides a consistent
ranking of all lotteries.

Definition 5 Assume a utility function u over the outcomes A. The expected

utility of the lottery L = {(a1 , p1 ) , (a2 , p2 ) , .. (an , pn )} is defined as
n
X
u (L) = u (ai ) pi
i=1

Before we introduce the additional axioms we discuss the notion of com-

pound (two stage) lotteries.

5
Definition 6 The compound lottery L̃ is expressed as L̃ = {(L1 , q1 ) , (L2 , 1 − q1 )}.
With probability q1 the simple lottery L1 is chosen and with probability 1 − q1
the simple lottery L2 is chosen.

Note, that we implicitly distinguish between simple and compound lotteries.

Therefore, we allow that a simple lottery L might have the same outcome
distribution as the compound lottery L̃ but L ≺ L̃.
The first axiom assumes that only outcomes matter - the process which
generates those outcomes is irrelevant.

Axiom 3 Each compound lottery is equivalent to a simple lottery with the

same distribution over final outcomes.

In some books the equivalence of simple and compound lotteries is assumed

in the definition of a lottery. However, it is useful to keep those types of
lotteries separate because we know that the framing of a decision problem
influences how people make choices (i.e. both the process and the final out-
come distribution matter).
The next axiom is fairly uncontroversial.

Axiom 4 Monotonicity. Assume that the lottery L1 is preferred to lot-

tery L2 . Then the compound lottery {(L1 , α) , (L2 , 1 − α)} is preferred to
{(L1 , β) , (L2 , 1 − β)} if α > β.

Axiom 5 Archimedian. For any outcomes a < b < c there is some lottery
L = {(a, α) , (c, 1 − α)} such that the agent is indifferent between L and b.

The substitution axiom (also know as independence of irrelevant alterna-

tives) is the most critical axiom.

Axiom 6 Substitution. If lottery L1 is preferred to lottery L2 then any mix-

ture of these lotteries with any other lottery L3 preserves this ordering:

{(L1 , α) , (L3 , 1 − α)} ≥ {(L2 , α) , (L3 , 1 − α)}

This axiom is also known as independence of irrelevant alternatives.

Under these axioms we obtain the celebrated result due to John von
Neumann and Oskar Morgenstern.

Theorem 2 Under the above axioms an expected utility function exists.

6
Proof: First of all we find the best and worst outcome b and w (possible be-
cause there are only finitely many outcomes). Because of the Archime-
dian axiom we can find a number αa for each outcome a such that L =
{(b, α) , (w, 1 − α)}. We can define a utility function over each outcome
a such that u (a) = α. Using the monotonicity axiom it can be shown
that this number is unique. For each lottery we can now calculate its ex-
pected utility. It remains to be shown that this expected utility function
is consistent with the original preferences. So take two lotteries L1 and
L2 such that L1 ¹ L2 . We can write L1 = {(a1 , p1 ) , (a2 , p2 ) , .. (an , pn )}
and L2 = {(a1 , q1 ) , (a2 , q2 ) , .. (an , qn )}. Now replace each outcome
ai by the above Pn lotteries. The compound Pn lottery can be rewritten
as L1 = {(b,P i=1 pi u (ai )) , (w, 1
P−n i=1 pi u (ai ))}. Similarly, we get
L2 = {(b, Pni=1 qi u (ai )) , (w,
Pn 1 − i=1 qi u (ai ))}. From this we can de-
n
duce that i=1 pi u (ai ) ≤ i=1 qi u (ai ), e.g. EU (L1 ) ≤ EU (L2 ). Oth-
erwise, by the monotonicity axiom we could deduce that L1 Â L2 which
is a contradiction. QED

From now on all payoffs in our course will be assumed to represent vNM
utility values. The expected payoff will be the expected utility.

4.1 Puzzles
EUT forms the basis of modern micro economics. Despite its success there
are important behavioral inconsistencies related to it. Some of those we are
going to discuss briefly before we turn our attention to proper games.

4.2 Allais Paradox

Consider the following choice situation (A) among two lotteries:

• Lottery A1 promises a sure win of 3000,

• Lottery A2 is a 80 percent chance to win 4000 (and zero in 20 percent

of the cases).

Typically, A1 is strictly preferred to A2. Now, consider two further choice

pairs (B) and (C):

• Lottery B1 promises a 90 percent chance of winning 3000,

7
 • Lottery B2 is a 72 percent chance to win 4000.

This choice is included to see if there is a certainty effect.

• Lottery C1 promises a 25 percent chance of winning 3000,

• Lottery C2 is a 20 percent chance to win 4000.

Most people in our class now preferred C2 over C1.

It can be checked that the lotteries Bi and Ci are derived from Ai just
by mixing the original lotteries with an irrelevant alternative - in the case of
(B) there is a 10 percent chance of getting nothing and a 90 percent chance
of getting (A), and in case of (C), there is a 75 percent chance of getting
nothing.
The Allais paradox is the most prominent example for behavioral incon-
sistencies related to the von Neumann Morgenstern axiomatic model of choice
under uncertainty. The Allais paradox shows that the significant majority of
real decision makers orders uncertain prospects in a way that is inconsistent
with the postulate that choices are independent of irrelevant alternatives.
There is an alternative explanation for the failure of EUT in this case.
Assume, that agents face the compound lottery instead of the simple lotteries
(B) and (C). Now the relationship to (A) is much more transparent - in fact,
one could tell a story such as: ”with 75 percent probability you are not invited
to choose between these two outcomes, and with 25 percent probability you
can choose either A1 or A2”. It’s likely that choices would be much more
consistent now.
The standard explanation for the failure of EUT is peoples’ inability to
keep small probability differences apart. 80 percent and 100 percent ’looks’
quite different and people focus on the probabilities. 20 percent and 25 per-
cent ’looks’ the same - so people focus on the values instead. Prospect theory
(Kahnemann and Tversky) can deal with the Allais paradox by weighting
probabilities accordingly.

4.3 Framing effects

Framing effects are preference reversals induced by changes in reference
points.
Consider the following choice situation (A):

8
 Table 1: McNeil, Pauker and Tversky (1988)
Survival Mortality Both
Radiation Surgery Radiation Surgery Radiation Surgery
immediate 100 90 0 10
1 year 77 68 23 32
5 year 22 34 78 66
US 16 84 50 50 44 56
Israeli 20 80 45 56 34 66

Pair 1: 600 people are struck with a disease that could kill. Vaccine 1
will save 400 lives for sure while the second one will either save no one (1/3)
or will save everyone (with probability 2/3).
Pair 2: 600 people are struck with a disease that could kill. Vaccine 1
will kill 200 people for sure while the second one implies a 2/3 chance that
no one will die and a 1/3 chance that everyone will die.
Note that both situations are identical because save is equal to not kill.
However, people tend to be risk averse in saving lives and risk loving if it is
phrased in terms of losses (kills).
Preference reversals have real effects and do not just appear in cute ex-
amples. McNeil, Pauker and Tversky (1988) asked American doctors and
Israeli medical students about how they would choose between two cancer
treatments (surgery and radiation) - they presented one group with survival
statistics, a second group with mortality statistics and a third group with
both. Table 1 sums up their choices.