I don’t think that PD (or any of its variants) is a good proxy for cheating, because cheating involves deception or rule breaking, while “defect” is just a legal move.
A better proxy might be something like nuptial gifts in some spider species. So in some species, the male will present a female with whom he wishes to mate a nuptial gift - an insect wrapped in webbing. But the “cheat” move is when either the insect has already been sucked dry or when it’s snatched back too quickly for the female to feed.
We can estimate the degree to which cheaters prosper by looking at how common these and similar behaviors are in their respective populations - let evolution do the calculations. Animal behavior is replete with deceptive and manipulative communication, and because so much of it is genetically determined we can be reasonably confident that we have an objective metric.
https://en.wikipedia.org/wiki/Prisoner's_dilemma#Zero-determinant_strategies
Actually, mathematically speaking, in the long run they tend to eventually fail.
I don’t think that PD (or any of its variants) is a good proxy for cheating, because cheating involves deception or rule breaking, while “defect” is just a legal move.
A better proxy might be something like nuptial gifts in some spider species. So in some species, the male will present a female with whom he wishes to mate a nuptial gift - an insect wrapped in webbing. But the “cheat” move is when either the insect has already been sucked dry or when it’s snatched back too quickly for the female to feed.
We can estimate the degree to which cheaters prosper by looking at how common these and similar behaviors are in their respective populations - let evolution do the calculations. Animal behavior is replete with deceptive and manipulative communication, and because so much of it is genetically determined we can be reasonably confident that we have an objective metric.