Adversarial Testing, was Thou shalt have no 1626

"Trevor L. Jackson, III"

No, it is that both will apply the same process of reasoning. And since that is decidedly *not* explicitly part of the rules of the game (as presented for example in the article to which I referred), it is an incorrect buttumption to use.

That is what I was referring to by "wishful thinking": there happens to be one element of the array that is larger than the others, so that's supposed to be the inevitable outcome of optimal strategies? Anyone who knows clbuttical game theory knows that that is often not the case.

For the actual one-shot PD game as standardly described, each player has free choice and no way to collude with the other (which also would need some enforcement mechanism). In particular, there is no way to make the other person use your own process of reasoning, however wonderful you might believe it to be. Under these circumstances, with pure strategy A you will do best against the other player's choice of pure strategy a, and with pure strategy A you will do best against the other player's choice of pure strategy b. Since the other player's grand strategy *has* to be some mix of pure strategies a and b, you will do best by choosing pure strategy A, regardless of the other player's strategy. Note that you will do better even in the case that the other player uses the "superrational" reasoning and chooses pure strategy b. *He* will do much worse, but by the announced rules for the game you do not buttign any utility to what happens to the other guy.