The Prisoner’s Dilemma

Two suspects are arrested and held separately. Neither can communicate with the other. The prosecutor makes each the same offer: testify against your partner, and if they stay silent, you go free while they take the full sentence. If you both testify against each other, you both get a moderate sentence. If you both stay silent, you both get a light sentence, the best joint outcome available. But neither of you knows what the other will choose.

That’s the Prisoner’s Dilemma, one of the most studied problems in game theory, and it turns out to describe an enormous number of everyday human interactions far better than most of us realise.

Why It Matters Outside the Interrogation Room

Strip away the police-procedural framing and the underlying structure shows up everywhere. Two colleagues can either share credit generously or each try to claim it for themselves. Two neighbouring businesses can compete fairly or undercut each other into the ground. Two people in a relationship can each show up generously or each hold back, waiting to see what the other does first. In every case, both parties cooperating produces the best joint outcome, but each individual is tempted to defect, to take the selfish option, because it looks better for them personally, regardless of what the other person does.

The tragedy of the dilemma is that if both people reason this way, both defect, and both end up worse off than if they’d simply trusted each other. Rational self-interest, applied individually, produces a collectively worse outcome. That’s the whole puzzle.

The Tournament That Changed the Field

In the early 1980s, the political scientist Robert Axelrod ran a famous tournament, inviting game theorists to submit computer strategies for playing the Prisoner’s Dilemma repeatedly against each other. Dozens of complex, cunning strategies were submitted. The winner was embarrassingly simple: Tit for Tat. Cooperate on the first move. After that, do whatever your opponent did last time.

Tit for Tat won not because it was clever, but because it was nice (it never defected first), retaliatory (it punished defection immediately), forgiving (it went straight back to cooperating the moment the other side did), and clear (its pattern was easy for an opponent to read and therefore easy to cooperate with). Axelrod’s conclusion, which reshaped how game theorists think about cooperation, was that when an interaction is repeated rather than one-off, starting from cooperation and responding in kind consistently outperforms starting from suspicion or aggression.

That matters because most of your real relationships, with colleagues, family, long-term business partners, are repeated interactions, not one-off encounters. The strategy that wins in a single, anonymous, never-to-be-repeated transaction is often very different from the strategy that wins with someone you’ll be dealing with for years.

Where the Genuinely Dangerous People Fit In

The people we covered a few lessons ago, the small minority who operate with little regard for anyone else, tend to play this game differently. Rather than reciprocating cooperation, they’re more inclined toward exploiting a cooperative opponent for as long as it’s profitable to do so, and moving on once the relationship stops paying off for them. That’s not universally true of every individual, but it’s a reasonable pattern to watch for: someone who consistently takes the cooperative option you offer and repays it with a selfish one, again and again, without adjusting, is telling you something important about how they intend to treat you.

What This Means Practically

Start most relationships from cooperation. It’s the strategy that wins over time, and it’s also, simply, the more pleasant way to move through the world. But don’t be a pushover about it. If someone defects, respond in kind, clearly and promptly, so they understand cooperation is conditional, not automatic. And if they come back to cooperating, meet them there. Grudges outlast their usefulness quickly; Tit for Tat’s forgiveness is as important to its success as its retaliation.

There’s one exception worth naming plainly, because it comes up in practice more than people expect. When you’re dealing with someone you have real reason to believe will defect regardless of what you do, get in the first move. Don’t extend the trust that a Tit for Tat strategy assumes will eventually be reciprocated, because with a genuinely exploitative person, it won’t be. Knowing which situation you’re in, an ordinary relationship worth building through cooperation, or a genuinely one-sided one, is most of the skill.

Further Reading

• Axelrod, R. (1984). The Evolution of Cooperation. Basic Books.
• Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489), 1390-1396.
• Poundstone, W. (1992). Prisoner’s Dilemma. Doubleday.
• Nowak, M. A. (2006). Five rules for the evolution of cooperation. Science, 314(5805), 1560-1563.
• Rand, D. G., & Nowak, M. A. (2013). Human cooperation. Trends in Cognitive Sciences, 17(8), 413-425.