Game Theory: Why Cooperation Beats Competition in the Long Run

In a single interaction, cheating pays. Over repeated interactions with the same players, it doesn't. Axelrod's computer tournaments in the 1980s proved cooperation can evolve from pure self-interest.

The Prisoner's Dilemma, Briefly

Two suspects are interrogated separately. Each can cooperate (stay silent) or defect (betray the other). If both cooperate, they each serve one year. If both defect, three years each. If one defects while the other cooperates, the defector goes free and the cooperator serves five years. The rational individual choice — defect — produces a collectively worse outcome for both. This is the prisoner's dilemma, and it structures an enormous number of real situations: climate negotiations, antibiotic prescribing, arms races. In a one-shot game, defection is the dominant strategy. But what happens when the same players interact repeatedly?

Axelrod's Tournaments: A Natural Experiment in Cooperation

In 1980, political scientist Robert Axelrod ran an unusual experiment. He invited game theorists, economists, psychologists, and mathematicians to submit computer programs to play the iterated prisoner's dilemma. Programs played 200 rounds against every other program. Fourteen programs were submitted. The winner was the simplest: Tit-for-Tat, submitted by psychologist Anatol Rapoport. Two rules: cooperate on the first move; then do whatever your opponent did last round.

Axelrod ran a second tournament with 62 programs after publishing the results. Everyone knew Tit-for-Tat had won. It won again. In his 1984 book The Evolution of Cooperation, he analyzed why: Tit-for-Tat was nice (cooperated first, never defected unprovoked), retaliatory (immediately punished defection), forgiving (returned to cooperation as soon as the opponent did), and clear (its strategy was transparent).

Why Cooperation Evolves

Axelrod's simulations showed that cooperative strategies could invade populations of defectors if they could find each other. This gave evolutionary biology a framework for Robert Trivers' 1971 concept of "reciprocal altruism" — mutual aid between non-kin that evolves because long-term cooperation benefits exceed short-term costs. The conditions for cooperation to evolve are specific: repeated interactions with the same individuals, ability to identify cheaters, and sufficiently high probability of future interactions. Remove any of these and defection becomes attractive again.

Real-World Applications

Fisheries management is a classic tragedy of the commons. Elinor Ostrom (Nobel Prize 2009) showed empirically that communities often develop cooperative management regimes without privatization or state control — but only when conditions for repeated interaction and reputation effects are present. Climate agreements are iterated prisoner's dilemmas at the nation-state level. The Paris Agreement's voluntary commitments and transparency mechanisms attempt to create exactly the conditions Axelrod identified. Antibiotic resistance presents the same structure: each physician who prescribes unnecessarily is a defector in a game where everyone's future access depends on collective restraint.

FAQ

Q: Is Tit-for-Tat still considered the optimal strategy?

A: Follow-up tournaments found that "Generous Tit-for-Tat" — which occasionally forgives defection randomly — outperforms strict Tit-for-Tat by avoiding mutual recrimination spirals. In noisy environments where moves are sometimes miscommunicated, strict Tit-for-Tat can spiral into alternating defections. A 2012 study by Press and Dyson showed that "zero-determinant" strategies can extort cooperation from opponents — though in evolutionary simulations they tend to be outcompeted once the population learns to avoid them.

Q: Does game theory explain human altruism, or do people cooperate for reasons beyond self-interest?

A: Both. Experimental economics (Fehr and Gächter, 2000) demonstrates "strong reciprocity" — people punish cheaters even at personal cost in one-shot games, which pure self-interest doesn't explain. This appears to be a genuine moral emotion. But the evolutionary framework explains how this emotion could have evolved: populations with strong reciprocators cooperate more effectively and outcompete pure defectors. The proximate cause (genuine altruistic feeling) and the ultimate cause (evolutionary benefit) can both be true simultaneously.

Curious minds belong together. Find people who ask the same questions on Good People Map.