Prisoner’s Dilemma in Nature
Vervet monkeys and vampire bats have something in common—they risk their own safety to help others in their community. Vervet monkeys scream at their neighbors to warn them of nearby predators. Vampire bats give part of their blood meal to their group members. These actions seem to go against “survival of the fittest” because the ones that risk themselves are more likely to be noticed by predators and fail to pass their genes to their offsprings so this self-risk behavior should dwindle over time. However, the iterated Prisoner’s Dilemma proves that cooperative, rather than selfish strategies, are the best for survival. In the Prisoner’s Dilemma, each prisoner has a payoff associated with confessing or lying and each must decide what action to take to minimize their years in prison. Without knowing what the other prisoner will pick, the best strategy for each prisoner is to confess their crime. However, the scenario is different in nature where there is a large population with continuous interactions among the members. If the two players interact frequently, then they might end up cooperating for some number of rounds and receive a payoff that is higher than the payoff they have if either one of them defected. If the selfish player defects, then he will receive a higher payoff than the generous player. The selfish player will eventually encounter another selfish player that will also defect so they will both receive a very low payoff. The overall result of this is that the cooperative player will end up with a higher payoff than the selfish player.
It is interesting to see that the Prisoner’s Dilemma can be extended to a bigger population to help explain how cooperation came about in evolution. Although the Nash equilibrium is for both players to defect, it is not guaranteed that all players will defect in a world where there are constant interactions among the members. Rather, players will learn to cooperate over time, which a choice that the prisoners did not have. A recent study published by physicist Freeman Dyson and computer scientist William Press, however, found that playing cooperatively does not always yield the highest payoffs. If a few variables such as the tendency to defect is changed, then selfish strategies tend to have the highest payoffs. So while game theory is helpful in understanding the real world, it is still difficult to accurately model and predict it.
Source: https://www.quantamagazine.org/game-theory-explains-how-cooperation-evolved-20150212/