One thing that often confounds game theorist, economists and (while I can’t speak for other autistic people,) myself, is that people often don’t behave in a rational manner. Before jumping to any conclusions, you should first verify that you what it means to “behave in a rational manner.” In particular, you should avoid using any straw vulcans:

A [straw vulcan is a] straw man used to show that emotion is better than logic.

It starts by having characters who think “logically” try to solve a problem. And they can’t. Either they can’t find any answer, or they’re caught in some kind of standoff, or they’re even stuck in a Logic Bomb-type loop. Once this is established, someone who uses good old human emotion comes up with a solution that the logical thinker can’t. This proves An Aesop that emotion is superior and that the logical thinker shouldn’t trust logic so much.

This is, of course, completely broken. Fiction often gets the concept of logic wrong in a number of ways.

A true rationalist — by which I include competent game theorists and economists… and myself! — is aware that most other people are not rational. That’s not the confounding part. The confounding part is just how damn irrational people can be. Perhaps even more confounding is the fact an irrational (but otherwise reasonable) person is able to persist in a specific irrational behaviour, even after having been shown that such behaviour is irrational.

Dan Gilbert has an excellent TED Talk where he gives an excellent example of irrational behaviour. Paraphrased from memory, he says:

Imagine you want to see a play, and in your wallet, you have a twenty dollar bill, and a ticket for a play which you’ve paid twenty dollars for. Upon arriving at the theatre, you take out your wallet, and realize you’ve lost the movie ticket. Will you buy a new ticket with your remaining twenty dollar bill? Most people would say no. They say to themselves “I’ve already paid for this ticket once! I don’t want to pay $40 instead of $20 to see this play!”

Now imagine you want to see the play, and in your wallet, you have two twenty dollar bills. You arrive at the theatre, and you realize you’ve lost one of your bills. Will you buy a new ticket with your remaining twenty dollar bill? Most people would say yes. They say to themselves “Well, I cam here specifically to see the play, right? What does the lost of the $20 have to do with anything?”

In case it isn’t exceedingly clear, notice that in both cases, you had two pieces of paper in your wallet, each of which you could have exchanged to watch a play which you did indeed wanted to see. In both cases, you lost one of those two pieces of paper, and you had to opportunity to exchange the remaining piece of paper to see the play. So why is it in one case you decided to see the play, and in the other you decided no to see the play?

(Note that of the three paragraphs “quoted” from memory, I took the most liberties with the third one, where I’m pretty sure I belaboured the point much more strongly than Gilbert did, but I wasn’t comfortable taking it out of the quote block, and claiming it as my own words, as all of the ideas are from Gilbert.)

So what astonishes us rationalist is not so much that people behave in the way that Gilbert describes. Probably most rationalist behaved the same way before hearing Gilbert speak. What is astonishing is after hearing this talk from Gilbert, irrational people will laugh, say “Yes, it’s true…” and then continue to behave in that way.

Traditionally, researchers would try to “rationalize” (pun intended) this behaviour, the prototypical example being the Ultimatum game. In the Ultimatum game, one player (usually referred to as “A”) is given a certain amount of money (e.g. $100) and is given the opportunity to share any amount of it to a second player, player “B”. The catch is that player “B” has the option of rejection the money, in which case both players receive nothing. In a world involving purely rational agents, B should accept any and all offers above $0, because receiving any amount of money at all is better than receiving nothing. A knows this, and thus will always offer B the minimum, e.g. giving himself $99, and B $1. When this game is played practice, B will instead usually reject any offer which seems “unfair”. For example, B will often reject an offer of $20. Traditionally, a rationalist would wonder why B would do this, since they are throwing away an offer for a free twenty bucks. The rationalization is that perhaps humans, being highly social creatures, are rejecting the offer as a form of punishment. That is, B is willing to forego the short term benefits of $20 in exchange of the longterm benefits of punishing those who would treat others unfairly, and thus promote fairness in the society that they both live in. I’ll refer to this as the “punishment” theory.

A recent experiment has shown this not to be case. In this modified version of the ultimatum game (in the study, they called the modified version the “private impunity game”), A is again given some amount of money, and free to split it between himself and player B in any way he sees fit. The difference now is that whatever split player A decides on, he immediately gets to keep and leaves the game. All that remains is to take whatever sum that A had left over, and offer it to B, who is then free to accept it or reject it. In other words, in this variant of the game, B has no opportunity to actually punish A, not even via guilt, seeing as how A will have left from the game before B ever has the chance to make the decision of whether to accept or reject the offer. And yet it turns out even in this case, B will occasionally reject the offer, harming no one but themselves! The rejection rate is not quite as high as in the Ultimatum game, thus giving some credence to the “punishment” theory, but simultaneously, the fact that there is any rejection at all strongly argues against the “punishment” theory.

Perhaps equally frightening was a secondary result of the study, which revealed that if the rules of the “Impunity” game were explained more formally and succinctly as a series of if-then statements, rather than as natural English (i.e. in the form “if A chooses X and B chooses Y, then A receives $i and B receives $j.”) the rejection rates reverted back to those of the Ultimatum game, suggesting “that people can’t even be bothered to perform a rational analysis when money is on the line, much less engage in rational actions.” This conclusion really drained the blood from my head, and I had to sit down to regain my composure.

As a human, I’m genetically encoded to have empathy for fellow human beings. But… well, I don’t know how to express this next part adequately, but in when someone differs fundamentally from your way thinking, you lose empathy and see them as an alien (or more commonly, a “monster”). Consider any sociopathic serial killer, for example. Or less dramatically, a presumably insane homeless person screaming in the streets. Most people don’t feel sympathy nor empathy for these homeless, but rather an extreme discomfort and a desire to ignore and get away from the person, to pretend that they do no exist in your world.