Back to Marilyn

So now I want to bring things back from Batman and the Prisoner’s Dilemma to Marilyn, through a somewhat roundabout way. Marilyn Monroe said that we had to learn that we could only trust ourselves, and there’s sets of circumstances where that seems true, at first glance. But in the long run, it seems like that isn’t a good idea. 

The quintessential example of when it seems like you should only trust yourself is the Prisoner’s Dilemma, a coordination game where the rational thing to do seems to be betraying the other player, even though cooperating with them would net both of you a better result. It exploits even a healthy kind of self-interest, resulting in both participants being worse off. But life isn’t just a single prisoner’s dilemma. We find ourselves in these situations with the same people over and over again, and what they did last time can influence what we do, and what we think they’ll do. We can use strategies to try to get the best results, but there are a lot of different strategies available for us to use.

I outlined five of them, each one exemplified by a character from Batman. Batman never forgives, and always defects once he’s been betrayed. Leslie always cooperates. Two-Face acts randomly, the Joker always defects, and Harvey Bullock plays tit for tat, doing whatever the other player did last. We can see the results of these strategies here. This chart totals ten iterations based on the rules in the matrix above. Mouse over the picture for Two-Face’s coin flips (D for defect and C for cooperate). Remember, each number is a year spent in prison, so lower numbers are better. Note that Batman beats out Harvey, but only does so because he can better exploit the randomness of Two-Face. If we exclude the Two-Face strategy, and thus any random elements, they tie at 175.

What this means is we can see that the Marilyn solution of trusting only yourself (the Joker strategy here) leaves us strictly worse off than other strategies. Robert Axelrod, one of the original scholars working on the Prisoner’s Dilemma, found that Tit for Tat proved to be better than Batman against a wider array of strategies, because it was able to forgive. He examined this at some length in The Evolution of Strategies in the Iterated Prisoner’s Dilemma.

But you already know all of this. You act on it all the time because you trust people around you in all kinds of ways. Babysitters, crossing guards, doctors, contractors, and engineers. You rely on them without a second thought, provided you see them cooperating. The ones who don’t, who betray that trust, you develop a strategy to deal with them. Legal sanctions, for instance. Incentives not to defect. You’re participating in these prisoner’s dilemmas all the time, and hopefully now you can spot them. Being aware of them can help you strategize better, and as we can see, makes everyone better off.