Q. So it seems that the SAC insider trading case is a pretty good real world example of the classic “prisoner’s dilemma” experiment from game theory. So before we get to implications for the SAC case, can you give us a refresher on the basic idea?
A. Sure. Imagine a really smart prosecutor who arrests two suspects but needs at least one to cooperate to get a conviction. So he separates them into different rooms so they can’t talk to each other, and says: Here’s the deal. If you cooperate, and the other guy stays quiet, I’ll nail him for 10 years. But the opposite is also true: if he talks and you stay quiet, you’ll be doing the 10. If you both cooperate, it’ll be five years each. But if neither of you talks, I admit I’ll have to settle for a minor charge, and you’ll both just do 6 months.
Q. So, collectively, they’re much better off by staying quiet, of course. But neither one knows what the other will do…
A. And that’s the problem. For each of them, it’s better to talk: if the other guy stays quiet, you walk; and the worst you can do is five years. But if you stay quiet while he talks, you’re doing 10. Since you don’t know what the other guy is doing, the only rational individual answer is: talk, even thought the best joint answer is: stay quiet.
Q. And in a nutshell, that’s why most people expect Martoma to cooperate?
A. Right — even though we just described an abstract experiment, it pretty well describes this situation, and judges do tend to be quite lenient with cooperating defendants. The only difference is, here, the “players” can talk to each other, and maybe Martoma gets offered a sweet deal by the other possible defendant– changing the total payoff for not cooperating.
Q. So outside these real world examples, why was the prisoner’s dilemma so important to game theory, such a big part of A Beautiful Mind?
A. Because it raises the question of why in the real world humans don’t act purely selfishly more often than they do in, say, the business world. And the answer to that puzzle lies in something called the “iterated prisoner’s dilemma.” Real life is a series of prisoner’s dilemma games, and people who act selfishly in one round are identified and punished in later rounds by the other players.
Q. So, happily, game theory actually teaches us to be nice?
A. Yep, the Nobel prize winners and your mother agree.