The Selfish Gene

 “Nice guys finish last” is a common saying. But Dawkins thinks there’s also a sense in which nice guys finish first. He thinks about birds who are “grudgers” (those that pick parasites off other birds, but remember the ones that don’t return the favor and ignore them the next time around). That strategy actually beats out the “cheat” strategy (accepting help with parasites but not reciprocating). Dawkins thinks the “grudger” is the kind of “nice guy” who “finishes first.” This is the individual who engages in “reciprocal altruism.”   Dawkins agrees with Robert Axelrod and Hamilton that many wild animals are “engaged in ceaseless games of the Prisoner’s Dilemma, played out in evolutionary time,” which explains why nice guys finish first.

The Prisoner’s Dilemma is a strategy game in which two people each have two cards. One card says “cooperate” and the other says “defect.” Both players have the same cards. Each chooses one card and puts it face-down on the table, and the cards are revealed at the same time. If both people play “cooperate,” they each win $300. If both people play “defect,” they each pay a $10 fine. If one person plays “cooperate” and the other plays “defect,” the person who plays “cooperate” has to pay $100, and the person who plays “defect” wins $500.   

Even though mutual cooperation is the best strategy (both people win money), most people will play “defect,” because the risk of being the only one to cooperate is too high. This is why the game is called a “dilemma.” The most logical thing to do is defect—it won’t yield the best potential outcome, but it’s the least risky strategy.  

There’s another version of the game called the “Iterated (Repeated) Prisoner’s Dilemma.” In this version, the same players play the game over and over again, meaning they know what cards the other player chose in the previous round. Dawkins thinks the birds who were picking parasites off each other’s backs were playing the iterated Prisoner’s Dilemma, since they picked parasites multiple times and remembered the other birds’ strategies. 

Axelrod ran a competition where he programmed a computer to play all the possible strategies in the iterated Prisoner’s Dilemma (including one strategy that was simply “random”) 200 times. There are 15 possible strategies. It turns out that the winning strategy was “Tit for Tat,” meaning be nice on the first round, and in the next round play whatever card one’s opponent played in the previous round.

Axelrod ran the “Tit for Tat” strategy against a lot of other strategies, including some very similar ones. Two similar strategies are “Naïve Prober” (which is basically “Tit for Tat” with a random extra “Defect” thrown in now and again), and “Remorseful Prober” (which will randomly play “Defect,” but then play “Cooperate” on the next round). Dawkins calls these “nasty” strategies because they involve defecting first. Tit for Tat, however, is a “nice” strategy, because it only defects in retaliation. It turns out, that when ranked, the eight best strategies were “nice” strategies (in which the player never defects unprovoked), and the strategies that trailed behind were “nasty” strategies (in which the player does defect unprovoked, in some way or another).

 “Tit for Tat” is also a forgiving strategy: one only play the card one’s opponent played on the previous round. The “grudger” strategy that birds play when picking parasites off each other, in contrast, is unforgiving. The bird remembers that another bird didn’t help them for the whole game and never helps them again, even if the cheating bird changes its mind and helps the grudger. It turns out that the “grudger” strategy ranked next to last in Axelrod’s simulation. Dawkins thinks this means that the best strategies are both “nice” and “forgiving.”

Axelrod called “Tit for Tat” a “robust” strategy. Dawkins calls it an evolutionarily stable strategy. Interestingly, “Tit for Tat” doesn’t win in an environment where most other strategies are nasty. Dawkins thinks this means that evolutionarily stable strategies work when there are a lot of others who act in similar ways. This is exactly what happens in nature, because winning strategies produce offspring that inherit similar behavior tendencies.   

Axelrod ran his Prisoner’s Dilemma game multiple times. Dawkins thinks each round of games is analogous to a generation. When multiple rounds are played it turns out that the winning strategy is whichever more or less “stable” one is dominant first.  If “always defect” is played often, and early in the game, it will dominate and become the evolutionarily stable strategy.

Dawkins wonders which strategy (between “Tit for Tat” and “always defect”) wins out in real life, if both are stable. “Tit for Tat” tends to do well when other people play the same strategy. He decides that since individuals tend to live near their kin, it’s likely that if one person tends to play “nice,” their genetic relatives will too, and “Tit for Tat” prevails. An “always defect” family would stay that way, but fair badly in the long run.

Games are either “zero sum” games (meaning one side wins and the other loses), or “nonzero sum” games (meaning both sides can win). Football, chess, rugby, and tennis are all zero sum games—usually there’s one winner and one loser. Dawkins thinks that many situations in nature are nonzero sum games. If both sides cooperate, they can both win, which explains how “cooperation and mutual assistance can flourish even in a basically selfish world.”

Dawkins talks about times during World War I when German and British soldiers were nice to each other between battles, even though they were enemies in the war. One Christmas, for example, they put down their guns, drank wine, and celebrated together on the battleground. Dawkins wonders if niceness can evolve in nature in the ways that Axelrod’s simulation games showed (and how the World War I soldiers behaved). Dawkins concludes that it can, if the game is nonzero sum, and if it’s iterated (repeated).

For example, fig trees and fig wasps cooperate. Fig wasps pollinate fig trees, and fig trees provide nourishment for fig wasp eggs. Both sides could be exploitative (or “defect”)—if the wasps lay too many eggs, or if the trees shed flowers with eggs in them—but they “cooperate” because over time (for generations) the benefit to each is higher than if they exploit each other. 

 Wilkinson’s research on vampire bats indicates that the “cooperate” strategy comes up elsewhere in nature, as well. Wilkinson discovered that vampire bats who are able to feed often donate their food (which is blood) to other bats. He examined 110 donations, and found that 33 were to non-relatives. He then discovered that bats can recognize each other. Dawkins thinks these 33 examples were the “Tit for Tat” policy in action. The bats were sharing their meals in case they needed help next time. Dawkins thinks that he has shown that even in a world driven by selfish genes, “nice guys can finish first.”