I must confess that I am running out of gas. But let me address crime as a strategy. The issue is not whether rape or murder is a successful strategy, but whether tolerance of rape and murder is a successful strategy. If a society condemns and punishes criminals. the society thrives. If it doesn't, it collapses into dystopia. Naturally, then, one would expect almost all of the people in a society that condemns rape and murder not to want to commit rape and murder. Outliers may or may not survive. Lots of criminals are severely punished, so it's not so clear that crime is a "successful strategy" so much as something some people get away with.

My point is not that we don't use our big brains to think about morality. My point is that when the smoke clears, the successful outcome will be consistent with what game theory would have suggested if the inputs were known. But, of course, the arguments are about the inputs (i.e., goals), so it's impossible to use game theory to decide how to reach them. It's a bit like the P vs. NP problem: some problems are very difficult to solve but their answers are very easy to verify.

One of my favorite quotations - which I may already have used in your presence - is from Groucho Marx. Asked to write a blurb for a book he had not read, he wrote "Outside of a dog, a book is man's best friend. Inside of a dog, it's too dark to read." As you read this, does it occur to you that you are NOT inside of a dog? I'm betting it does not. And you I am also betting that you are NOT inside of a dog. Likewise, we develop morality without any thought to natural selection or game theory. Yet, I will bet that any successful set of mores conduces to survival in a way describable as a Nash equilibrium.