Why Would Two Prisoners Both Confess When Silence Is a Better Deal?

The Interrogation: A Story That Started It All

In 1950, two mathematicians, Merrill Flood (1908–1991) and Melvin Dresher (1911–1992), invented a simple game to test how people make decisions. The game puts you in the shoes of a prisoner. You and your partner in crime have been arrested and locked in separate rooms. The police don’t have enough evidence to convict you on the big charge, but they can definitely convict you on a smaller one. They give each of you the same offer:

• If you both stay silent, you get one year in jail for the small crime.
• If you both confess to the big crime, you each get three years.
• But if one of you confesses and the other stays silent, the confessor walks free right away, and the silent one gets five years.

You can’t talk to your partner. What should you do?

Reason it out. If your partner stays silent, you can either stay silent too and get one year, or confess and go free. Free beats one year, so confessing looks better. If your partner confesses, you can stay silent and get five years, or confess and get three. Three beats five, so confessing is smarter again. No matter what the other prisoner does, confessing gives you a lighter sentence. The same logic runs in your partner’s head. Both of you confess, and both end up with three years — even though you could have both stayed silent and walked out after one year.

Flood and Dresher had uncovered a maddening puzzle. Two people, each acting perfectly rationally, can produce an outcome that is worse for both of them. They called this the Prisoner’s Dilemma. The move to stay silent is called cooperation, because it would benefit both if you both did it. The move to confess is called defection.

Why Would Smart People Do That? The Logic of the Trap

The Prisoner’s Dilemma feels like a trick. There’s a term for the reasoning that leads you to confess: dominant strategy. A move is strictly dominant if it always gives you a better result, no matter what the other player does. In this game, defection is dominant for both players. When both follow their dominant strategy, they land in a spot where neither can do better by changing only their own move — this is called a Nash equilibrium (named after John Nash (1928–2015)). Both confessing is the only Nash equilibrium in the simple game.

Notice the problem: the outcome where both cooperate is Pareto superior — it makes both better off than the equilibrium. So each person’s private reasoning steers them toward a result nobody wanted. The exact numbers don’t drive the trap; the ranking does. As long as the temptation to squeal beats the reward of mutual silence, which beats the punishment of mutual betrayal, which beats the sucker’s payoff, the same logic forces two players into the worse corner.

Think of it like two neighbours who would both benefit from a shared fence. Each would rather the other pay for it, so if both argue that way, no fence gets built. The trap snaps shut whenever everyone optimises for themselves without considering that others are doing the same.

When the Dilemma Escapes the Prison: Real‑World Traps

The PD isn’t just a puzzle for prisoners. It pops up wherever people could cooperate but have strong reasons not to. Economists see it in price wars: two companies that would earn more by keeping prices high may slash prices, each hoping to steal customers, until both lose profits. Political scientists see it in arms races: two countries that build more weapons to feel safer can end up with less security overall because each provokes the other. Environmental problems often have the same shape — a clean lake benefits everyone, but each lakeside resident might secretly dump waste, thinking “my little bit won’t matter,” and soon the lake is ruined.

Even everyday life offers examples. In a school group project, each member hopes someone else will do the hard work. If everyone adopts that free‑rider logic, the project fails and everyone gets a bad grade. The PD isn’t a rare glitch; it’s a pattern that appears whenever individual gain tempts us to break from the common good.

Playing Again and Again: How Tit‑for‑Tat Tamed the Trap

Is there any escape? In the 1970s and 1980s, political scientist Robert Axelrod (b. 1943) wondered what would happen if the same two players faced the dilemma over and over. He invited computer programmers to submit strategies for this iterated Prisoner’s Dilemma (IPD) and staged a tournament. The winner was a shockingly simple program called Tit‑for‑Tat (TFT). Its rule: start by cooperating, and on every round after that, simply copy whatever the other player did on the previous round.

TFT succeeded because it had four everyday virtues. It was nice — it never defected first. It was retaliatory — if the other defected, it immediately struck back. It was forgiving — after punishing once, it returned to cooperation as soon as the opponent did. And it was clear — opponents could quickly figure out what it would do. By combining these traits, TFT encouraged mutual cooperation: players learned that defecting brought immediate punishment, while cooperating kept the rewards flowing.

The iterated game doesn’t guarantee cooperation; if everyone knows the exact last round, clever backward reasoning can unravel everything. But in real life, we rarely know when our last meeting will be. As long as there is a decent chance of meeting again — a strong shadow of the future — strategies like TFT, and even more generous cousins that forgive occasional mistakes, can keep cooperation alive. Even purely self‑interested players can learn to work together.

So Why Does This Matter Today?

The Prisoner’s Dilemma matters because it gives a name to a feeling we’ve all had: “If I do the right thing but nobody else does, I’ll be the sucker.” It shows that purely selfish reasoning, when everyone follows it, can steer us toward outcomes nobody wants — from climate inaction to gridlocked friendships. And it helps us see that building trust, repeating interactions, and being both firm and forgiving can unlock cooperation that pure logic seems to forbid.

Philosophers, economists, and computer scientists still study the PD to design better institutions: systems of rewards and punishments that make cooperation the smart move. Understanding the trap is the first step to avoiding it. The next time you face a situation where everyone is waiting for someone else to give first, you might recognise the pattern — and choose to break the cycle.

Think about it

1. If you knew you would never see someone again after this one interaction, would it be wise to cooperate? Why or why not?
2. Can you think of a rule, a tool, or a habit that could help a group escape a real‑life Prisoner’s Dilemma without a boss forcing them?
3. Tit‑for‑Tat works best when both players remember past rounds. In a situation where people forget or make mistakes, what could go wrong? How might you fix it?