68 Chapter Highlights
Chapter 29
Game Theory
This is a fun chapter. Students like it a lot, and faculty usually enjoy teaching
it. Game theory is hot stuff in economics these days, and this chapter tries to
convey some of the reasons why.
The first two equilibrium concepts, that of a dominant strategy equilibrium
and that of a Nash equilibrium, are reasonably easy to convey. The idea of a
Nash equilibrium in mixed strategies is a little harder. Here’s an example that
will motivate the idea.
Consider the game of baseball. The pitcher has two strategies: pitch high or
pitch low. Likewise, the batter has two strategies, swing high or swing low. If
the batter connects, he gets a payoff of 1 and the pitcher gets zero. If the batter
misses, the pitcher gets a payoff of 1.
What are the Nash equilibria in this game? If the pitcher always pitches
high, the batter will always swing high, and if the pitcher always pitches low,
then the batter will always swing low. It is clear from this observation—and from
observing baseball games—that the equilibrium strategy must involve a mixed
strategy. The pitcher will flip a coin and decide whether to pitch high or low,
and the batter flips a coin to decide whether to swing high or low. The batter
will connect 50% of the time. Here students are very willing to accept that the
optimal strategy must involve randomization.
If you really want to get them buzzing, you can talk about the following
paradox. If the batter really believes that the pitcher will really randomize 50–
50, then he might as well swing high all the time. But of course, once the pitcher
detects this departure from randomizing, he will modify his own behavior to
exploit the batter’s sloppiness. This example drives home the important point
that what keeps the players at the Nash equilibrium is the desire to avoid being
psyched out by their opponents.
Most students have heard of the prisoners’ dilemma by now, but they haven’t
seen the analysis of the repeated game. The reason why the repeated game
is different from the one-shot game is that in the repeated game, the strategy
choice at time tcan depend on the entire history of the game up until t.Thus
choices at time t−1 may have some influence on choices at time t. This opens
the possibility of tit-for-tat and other strategies that can allow for cooperative
solutions.