When Rationality Leads to Ruin
Two criminals are arrested and held in separate cells. The police offer each a deal: betray your partner (Defect) or stay silent (Cooperate). Neither knows what the other will choose.
| Partner Cooperates | Partner Defects | |
|---|---|---|
| You Cooperate | -1, -1 Both serve 1 year |
-5, 0 You: 5 years, Partner: free |
| You Defect | 0, -5 You: free, Partner: 5 years |
-3, -3 Both serve 3 years |
Defecting is always the rational choice for each individual. If your partner cooperates, defecting gets you freedom instead of 1 year. If your partner defects, defecting gets you 3 years instead of 5.
But if both defect, both get 3 years—worse than if both had cooperated (1 year each). Individual rationality leads to collective disaster!
Play multiple rounds against different strategies. Points are inverted (higher = better):
Run a tournament where each strategy plays against every other strategy for 100 rounds. Click to see which strategy wins!
Start by cooperating, then copy whatever the opponent did last round. Simple, forgiving, and retaliatory. Won Axelrod's famous tournament!
Never cooperate. Maximizes individual gain against naive cooperators, but mutual defection is suboptimal.
Always cooperate regardless. Maximizes mutual benefit but is easily exploited by defectors.
Cooperate until the opponent defects once, then defect forever. Punishes betrayal harshly but can't forgive mistakes.
If the outcome was good (CC or DC), repeat your move. If bad (CD or DD), switch. Learns from outcomes.
Flip a coin each round. Unpredictable but doesn't exploit patterns or build cooperation.
Nations would be safer if none armed. But each has incentive to arm regardless of what others do. Result: everyone arms, everyone less safe.
All countries benefit from emissions cuts. But each has incentive to free-ride on others' efforts. Result: insufficient action.
Companies could profit more by keeping prices high. But each has incentive to undercut. Result: race to the bottom.
Teams work better when everyone shares credit. But individuals have incentive to claim credit. Result: dysfunction.
The Prisoner's Dilemma reveals a profound truth: individual rationality can lead to collective irrationality. What's best for each person isn't best for everyone.
Solutions require changing the game itself: repeated interaction (building trust), communication, enforceable agreements, or changing incentives. The invisible hand doesn't always work—sometimes it picks everyone's pocket.
Political scientist Robert Axelrod invited game theorists to submit strategies for an iterated Prisoner's Dilemma tournament. The surprise winner?
The simplest strategy submitted—just 4 lines of code—beat complex strategies from renowned experts. It succeeded by being:
This finding influenced fields from evolutionary biology to international relations. Cooperation can emerge even among self-interested actors—if the game repeats and reputation matters.