Game theory needs to add “stupidity theory” to its playbook

soc_econ.pngI have big problems with objectivism, evolutionary psychology, and behavioral economics. At least, with the popular notions of them.

I’m talking about the ideas that people’s behavior can always (or even mostly) be explained in selfish motivations. It’s the idea that altruism can be explained as mate selection or get-laid schemes. It’s the idea that everything people do must have some kind of personal “payoff” or benefit.1

It has nothing to do with whether I find the models distasteful. There’s a practical real-world problem with using these ideas as explanatory models.

We can’t know anyone else’s motivation. Not really, anyway. We can know something about what areas of the brain are activated2, but that’s not the same thing as knowing motivation.

But we all know something about paranoia. We’ve experienced it. And that’s where these explanatory models go to hell really, really fast.

People do some really screwed-up things. If you believe that people are always (or even usually) acting in their own self-interest… then all of those screwed-up things must be done intentionally.

Quite simply, people aren’t that smart. Hanlon’s Razor (or Heinlein’s Razor) sums it up pretty damn well:

Never attribute to malice that which is adequately explained by stupidity.

But if you think that people are usually acting intentionally, and for their own gain, then sooner or later it all ends up looking like conspiracy. Specifically, a conspiracy against you. Any time someone’s an idiot, overlooks something, or is generally clueless, it gets interpreted as an intentional attack instead.

And given how often people do stupid things, that sounds like a pretty miserable way to live.

1I’m aware that I’m paraphrasing popular understanding of science, not the science itself. Short science-y answer to anyone about to pull citations on me: They explain one variable well, but they’ve not made the case that they are talking about sufficient cause in all cases.
2In laboratory cases, which has its own problems.