2. Game theory’s most studied scenario is the two-player, no-communication Prisoners Dilemma (developed for Cold War nuclear strategy, assuming all are untrustworthy). Very briefly: Players have two options, cooperate or defect, and four payoffs, typically ranked Temptation > Reward > Punishment > Sucker. If both cooperate, each payoff = Reward.  If only one defects, his payoff = Temptation, the other’s payoff = Sucker. Both defect, and both payoffs = Punishment. 3. Conventional “logic” says the other player is “rational” and untrustworthy, so he’ll defect. So it’s “rational” for you to defect, guaranteeing poor results. But evolution can be smarter. 4. Computer contests of iterated Prisoner’s Dilemmas show “that evolution probably selected for two stable orientations… tit-for-tat… and permanent defection.” Permanent defection = low payoffs of the “red in tooth and claw” variety. 5. Tit-for-tat involves cooperating initially, then mimicking the other player’s last move. Thus tit-for-tat-players retaliate if defected against, but forgive a defector that starts cooperating. Once established, tit-for-tat improves productivity for all, and can become an "evolutionarily stable strategy," meaning it can’t be beaten by other approaches. Even with short-term incentives for defection, cooperation can thrive. - http://bigthink.com/errors-we-live-by/game-theory-behavioural-telescope