you are viewing a single comment's thread.

view the rest of the comments →

[–]wizzwizz4 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (1 child)

The best outcome is mutual peace. However, the first defecting party – the first to use violence – gets a better position, and the only action that the other party can take is to also defect, until they're both violent. This is worse for both parties than both being peaceful, but it's strictly dominant.

Is this model inaccurate?

[–]sodasplash 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (0 children)

Your model is not untrue.

However, the usefulness of the prisoner’s dilemma exists in a larger scope than you’re considering, that scope being game theory. The prisoner’s dilemma was created to explain a certain very basic aspect of game theory. You posit that violence necessarily begets violence. In the context of game theory, it’s not that simple.

Certainly in a 1 v 1, winner take all scenario, what you describe is correct. However, I have two responses and I’ll get back to the better one in a second.

The Prisoner’s Dilemma tends to be used most effectively in the context of a larger “game.” In larger games, it tends to be presented as trade fair or screw over trade partner. I.E. sell bread for one dollar or steal bread. But there’s communication between parties so the thief eventually gets frozen out.

Say you have 50 traders and they all come to market and they all get to trade with each other. Eventually word will get and the ones screwing over other people (by selling defective products, running ponzi schemes, etc) begin to get frozen out and they fail.

This is essentially how modern international trading began. People realized trading was better than invading and so trading proliferated and larger and larger trading spheres (i.e. cities then nations then trading blocks) were created. It’s a model often thought of in terms of zerosum (stealing) or nonzerosum (trading) as described brilliantly in Robert Wright’s Nonzero.

Where it gets more interesting is in the more minute description of the 1 v 1 scenario. There are two prisoners. And they each have two choices. So there are actually FOUR outcomes, best described in a square.

The key is that in a prisoner’s dilemma both parties make their decision in the dark and know that the other is also doing the same. (Somewhat unrealistic). Imagine two people are arrested and they can each either snitch or keep quiet.

  • Both snitch: both get 5 years.

  • A snitches, B keeps quiet: B gets 10 years, A walks free

  • B snitches, A keeps quiet: A gets 10 years, B walks free

  • Both keep quiet: both get 1 year.

Obviously the best outcome for both is to keep quiet (peace) but the second best outcome for both is to snitch (attack). This is more of a purely philosophical question about the nature of man and trust.

Your proposed outlook has more realistic application in a sense but it doesn’t really have an application within game theory which was the point of the prisoner’s dilemma being made up in the first place. Not to come off as insulting because your model does have more real world theory in many cases, all you are saying ultimately is that violence leads to violence.

All real world cases of the prisoner’s dilemma being played out with multiple parties shows that violence leads to ostracization. I feel like that can be seen as reflected in the world today. The regimes which are seen as most rogue are the ones most likely to be gone to war with and face international sanctions.

You can argue over the morality of the world community’s decisions obviously (as I would) but there’s little doubt that those regimes are not playing by the laws the world has set forth (whether you and I agree with those laws or not).