THE PRISONER’S DILEMMA
There is a Game Theory concept known as the Prisoners’ Dilemma. This is what it entails: suppose you and I are arrested and placed in separate rooms where we can’t communicate. But, fortunately (?) neither of us can be indicted on the primary charge (say, assault), which carries 3 years in prison. But the cops can put us away for a lesser charge (say, breaking and entering), which carries a 1-year sentence. But they want to strike big. So the good-cop-bad-cop routine starts. They want to turn us against each other. “Testify, and you’ll go free”, they promise.
Now, here’s how your options pan out:
- You can sing (testify against me) and go free
- You can keep quiet (and accept the 1 year in prison)
But since you can’t communicate with me – you don’t know what I’m up to. Surely, the cops have offered me the same deal. The real dilemma is that if we both testify, we both also get 2 years in prison. What do you do?
It seems easy, doesn’t it? Just sell me out and get the hell outta there! But, wait! What if I am doing just the same? If we both sing – both get 2 years remember? That’s worse than the 1 year each if no one talks. So why not just keep quiet? Well …because if you keep mum, I’ll probably sell you out either way, shack up with your wife and you end up with three years.
Its roughly similar to the problem presented to passengers of the two ferries in the Dark Knight. If none detonated the bomb, everyone lived. One could kill the other, while the worst outcome was mutual destruction. And neither knows what the other is thinking. Take a look at the scenarios and their corresponding prison sentences below:
THE RATIONAL CHOICE
I’d appreciate it if you didn’t rat me out. Because I’d do that for you. But emotions / sentiments seldom play a part in rational decisions. In Game Theory, the dominant choice is to betray. Think about it, if you betray me, this is what happens. if I stay silent, you go free. If I talk, you get 2 years instead of three. The underlying assumption is that I, too, will fess up. This is what the purely rational being would do.
Two researchers Menusch Khadjavi and Andreas Lange, of the University of Hamburg, put the theory up for test. They employed female inmates from a Saxony Prison and students. Instead of sentences, they used Euros (for students) and coffee / cigarettes (for prisoners). They conducted both simultaneous and sequential (i.e. one deciding before the other, then the other decides, then first decides again and so on) versions of the experiment.
In the simultaneous (i.e. real time) version, 37% of students cooperated between themselves (Game Theory says 100% would act selfishly and never cooperate). Inmates cooperated a staggering 56% of the time. 13% got mutual cooperation – the best possible combination of outcomes.
In the sequential game, one prisoner had to act first – based on either pure-rationality or blind trust / good faith. The other could then decide to act rationally, return the favor or show goodness of heart. A whopping 63% students cooperated, so the mutual cooperation rate skyrockets to 39%. Inmates cooperated about 50%+ times.
WHAT DOES IT ALL MEAN?
What the experiment – like the ferry-bomb outcome in the Dark Knight – suggests, is that human beings are not purely rational creatures. They don’t always act to promote self interest as suggested by the profit-maximizing maxims of our times. Instead, we make complex decisions that are fueled by kindness, consideration, social-consciousness and vengeance. To me, that represents a better understanding of our world than the one where my friends always sell me out.