In 1980, the political scientist Robert Axelrod invited game theorists to a tournament. Each entrant would submit a computer program to play a simple game, the prisoner’s dilemma, two hundred rounds at a time, against every other program in the field. Fourteen arrived, written by economists, mathematicians, psychologists, and computer scientists. Some were elaborate, built to probe for weakness and exploit it.¹

The winner was four lines long. The mathematician and psychologist Anatol Rapoport sub...