A short story about "game theory"

Jan 22, 2006 10:16 GMT  ·  By

Let's play a game, a more exotic one. Imagine you have committed a crime together with a friend. You are caught by the police, separated from your friend and you are both being interrogated. You have two options: either to betray your friend and admit to everything or to keep your mouth shut hoping your friend will do the same. What should you do?

Well, let's see what are the advantages and disadvantages of each one of these choices. We have to take into account what both persons choose to do, because they both influence the general outcome of the situation. The police have already established that: if one person confesses to the crime and the other one keeps quiet, the first one goes free and the other spends 10 years in jail. If they both confess, they both get 2 years in prison. If both keep quiet about what happened, then due to lack of evidence, they will both get a minor-6 month in prison-conviction.

So if you are one of the two, you have to always take into account what the other person might do. You might decide, on a first assessment of the situation, that it's better to take the safe way and confess. In this case, no matter what the other one does, you won't get more than 2 years in jail, or you might even go free if you betray the other person and he/she doesn't betray you. But on second thought you might realize that if you both cooperate and don't betray each other the general outcome is better-you will both get less time in prison. Of course if you choose to keep quiet the risk is greater because you can be betrayed and then you'll rot in jail. The thing is-will you trust the other and take that risk?

This is called the Prisoner's Dilemma. It's a game, but of a more serious type; people theorize on these kinds of games and apply the results to other domains. The general research field is called Game theory. It's pretty important stuff; you can tell that even from the definition we find in the popular online encyclopedia Wikipedia:

Game theory is a branch of applied mathematics that studies strategic situations where players choose different actions in an attempt to maximize their returns. First developed as a tool for understanding economic behavior, game theory is now used in many diverse academic fields, ranging from biology to philosophy. Game theory saw substantial growth and its first formalization by John von Neumann before and during the Cold War, mainly due to its application to military strategy, most notably to the concept of mutual assured destruction. Beginning in the 1970s, game theory has been applied to animal behavior, including species' development by natural selection. Because of interesting games like the prisoner's dilemma, in which mutual self-interest hurts everyone, game theory has been used in political science, ethics and philosophy. Finally, game theory has recently drawn attention from computer scientists because of its use in artificial intelligence and cybernetics.

Returning to our Prisoner's Dilemma? the dilemma is actually: should one act to secure only his/her individual well-being, even if that means betraying, or should he/she take into account the well-being of the other person, risk to keep quiet and in the end, if he/she is lucky, obtain a better general outcome? A liberal would say individual well-being is more important- if one strives to progress as an independent individual then society will progress. Liberals encourage a free society where there are no constraints imposed on people, constrains that would prevent them from pursuing their individual goals. A socialist would tend to defend a more cooperative point of view: we can see that a society where people cooperate and think about the well-being of the other person is better even from the results of this game. It's not a question of being nice, moral, but rather a practical option- cooperating implies greater common profit.

I will contrast this with another interesting game: The Rendezvous Problem. Imagine two people getting lost from one another in a maze (it can be a park, a supermarket, whatever). They have to find each other and have no extra clues. Each of them has, again, two options: either to stand still and wait for the other to look for him/her, or to start looking. If they both stand still they will never meet. But if they both look it will take longer to find each other than if one stands still and the other one looks. So the best possible result can be achieved if they do opposite things.

This is clearly different from the Prisoner's Dilemma, the two have to coordinate by doing different things, not the same, and there is less room here for rational calculus. This game seams to be more about guessing what the other might do. It relies more on intuition, and might be useful for modeling relationships on a smaller scale, like relationships between two people. This game is less relevant for whole societies; liberals or socialist might not really be able to argue in favor of their points of view using this as an example (like they do using the Prisoner's Dilemma). Nevertheless, this might be just the right example of how a relationship between two people works: one has to search and one to wait, you just have to guess when it's time to search and when to wait. And all the applied math and all the probabilities in the world might not help you with that. You just have to rely on intuition sometimes, I guess.

Photo Gallery (4 Images)

+1more