0
0
Two people are caught and put in separate rooms. They can’t talk to each other. Each one gets the same offer: tell on your partner and you walk free, while your partner takes the full punishment.
If both stay quiet, both get a short sentence. If both tell, both get a long one. If only one tells, that one goes home and the quiet one carries everything.
That little story is the prisoner’s dilemma, and it is one of the most studied puzzles in the world. It is studied because it is not really about prisoners. It is about roommates, business partners, neighbours, brothers and sisters, and two countries sharing one river. It is about every time you have to decide whether to trust someone who might not trust you back.
And the most surprising answer to it came from a contest between computer programs, where the winner turned out to be the one that kept giving people another chance.
What is the prisoner’s dilemma, in plain words?
Look at it from inside one of those rooms. You don’t know what your partner will do. So you think it through.
- If my partner stays quiet, I do better by telling. I walk free.
- If my partner tells, I also do better by telling. Otherwise I take the whole hit.
Either way, telling looks smarter. But your partner is thinking the exact same thing. So both of you tell, and both of you get the long sentence. If you had both just trusted each other, you would both be better off.
That is the trap. Each person makes the “smart” choice, and together they land in the worst shared result. The puzzle was first worked out by researchers at the RAND Corporation around 1950, and the mathematician Albert Tucker gave it the prison story and the name.
Here is the part that changes everything: in real life, we almost never play this game just once.
When the game repeats, everything changes
If you will only meet someone once, betraying them can look like it pays. But your coworker will be there tomorrow. Your neighbour will be there next year. Your sister will be there for the rest of your life.
When the same two people meet again and again, what you did last time follows you into the next round. People remember. This is called the repeated prisoner’s dilemma, and it is much closer to how life actually works.
So what is the best way to play a game like that? Around 1980, a political scientist at the University of Michigan named Robert Axelrod decided to find out with a contest.
The tournament: experts built programs to find the best strategy
Axelrod invited experts in game theory, economics, psychology, maths and politics to send him a computer program. Each program would play the repeated prisoner’s dilemma against every other program, many rounds in a row. The program with the best total score would win.
Fourteen programs came in. Some were long and clever. Some tried to trick their opponents. Some tried to guess what the other side would do next.
The winner was the shortest program in the whole contest. It was sent in by Anatol Rapoport, a psychologist at the University of Toronto, and it was called Tit for Tat. It had only two rules:
- On the first move, cooperate. Be kind first.
- After that, do whatever the other player did last time.
That’s it. If you were kind, it was kind back. If you betrayed it, it betrayed you once in return, and only once. In that sense it is the old rule of “an eye for an eye,” which, as it happens, started out as a limit on revenge, not a license for it. And the moment you went back to being kind, it went back to being kind too. It did not hold a grudge.
Axelrod then ran a second, bigger tournament. This time 62 programs entered, and everyone knew Tit for Tat had won the first time. Many tried to beat it. Tit for Tat won again.
The strange detail: Tit for Tat never beats anyone
Here is the part most summaries skip, and it might be the most important thing in this whole story.
Tit for Tat cannot score more than the player it is facing in any single match. Think about why. It never betrays first. It only ever hits back after it has been hit. So the best it can do in any one game is tie. Against a sneaky opponent, it loses a little.
And yet it won the whole tournament. Twice.
How? Because it was not trying to beat the person in front of it. It was trying to build something with them. When it met kind programs, the two of them cooperated for the whole game and both scored high. The tricky programs won small fights and kept ending up in long, ugly feuds with each other. Their total scores sank.
Axelrod looked at what the best programs had in common and found four traits. They were nice (never the first to betray). They were provocable (they did push back when someone betrayed them). They were forgiving (they went back to kindness fast). And they were clear (easy for others to read and trust). He wrote it all up in his 1984 book, The Evolution of Cooperation.
The players who were busy “winning” every exchange lost the game. The one that only cared about the relationship came out on top.
Then came mistakes, and forgiveness got even more important
Real life has a problem that the first tournaments mostly left out: mistakes.
Messages get lost. Someone is tired and snaps. A text is read in the wrong tone. You meant to help, and it looked like a betrayal.
In a world like that, plain Tit for Tat has a weakness. Picture two Tit for Tat players getting along fine. Then one of them makes a single mistake. The other hits back, as its rule says. The first one hits back for that. Now they are stuck, trading blows forever, each one “just responding” to the last thing the other did. One accident echoes down every round that follows.
You have probably seen this in a family. Nobody remembers how it started. Everyone is just answering the last hurt.
In 1992, two researchers, Martin Nowak and Karl Sigmund, published a study in the journal Nature that tested what happens when strategies have to live in a world with errors. The strategy that came out on top was a gentler version called Generous Tit for Tat. It still answers kindness with kindness. But when it is betrayed, it lets roughly one in three betrayals go, and simply cooperates anyway.
That small amount of extra grace is what breaks the echo. One forgiven move, and the feud stops. The two players are back to building something together.
The honest catch: forgiveness is not the same as being a doormat
This is the part the short version of this story usually leaves out, and it matters.
In his first report, Axelrod pointed out that a program called Tit for Two Tats would have won the first tournament if anyone had sent it in. It was more patient. It only hit back after being betrayed twice in a row. So in the second tournament, someone did enter it.
It did not win. The second time around, some players had learned to take advantage of extra patience. They would betray once, wait, and betray again, knowing they would get away with it. Tit for Tat beat it.
And the strategy that always cooperates no matter what? It gets eaten alive by anyone willing to use it. Even Generous Tit for Tat still pushes back on most betrayals. It just doesn’t keep pushing.
So the lesson is not “let people walk all over you.” It is something more careful and more useful: be kind first, be honest when you’ve been wronged, and then don’t keep score. The winners were never the ones who let anything slide. They were the ones who refused to let one bad round decide every round after it.
How to play the repeated game in real life
- Go first with kindness. Don’t wait for the other person to prove they are safe. Every winning program started by cooperating.
- Say it when something hurts. Pushing back isn’t the opposite of forgiveness. The winners were provocable. Name the problem clearly and calmly.
- Then drop it. Once the other person turns back toward you, turn back toward them. Don’t bring last month into this week.
- Assume some “betrayals” are mistakes. In a noisy world, a lot of what feels like an attack was a tired person or a lost message. Let some of them go without a reply.
- Be easy to read. Clear, steady people are easier to trust. Mixed signals start feuds.
If you are carrying an old hurt and you’re not sure you can let it go yet, our free Am I Ready to Forgive? quiz is a gentle place to start. And if what you’re stuck in feels more like a running tally you replay at night, how to let go of resentment by putting down the ledger walks through that exact problem.
The oldest answer to the prisoner’s dilemma
There is something quietly funny about all this. It took experts, computers and decades of maths to prove that the strategy that keeps forgiving outlasts the strategy that keeps score.
Very old wisdom got there first. There is an ancient conversation where someone asked how many times they should forgive a person who keeps hurting them. Seven times? That already sounded generous. The answer that came back was so much bigger that no one could keep count. Which may have been the point. You can’t hold a ledger that long.
Maybe the reason forgiveness wins, again and again, is that the world was made by a God who forgives first, and it works best when we play by the same rule.
Most of us are in a few long games right now. A marriage. A friendship. A job. A family group chat. The question is not whether someone will hurt us in the next round. They will, and we will hurt them too, mostly by accident.
The question is what we do with the round after that.
A question for you
Do you think forgiveness is a strength or a weakness in the real world, and does it depend on who you’re dealing with? Tell us what you think in the comments.
A good place to start:
The Beginner’s Guide to Feeling God’s Presence Every Day
A short video guide and companion PDF for noticing God in the ordinary moments of a real, messy life, including the relationships that test you most.
Get the free guide
Free.
Share this
“The winning strategy in the prisoner’s dilemma never beat a single opponent. It won anyway, because it kept forgiving.” https://bgodinspired.com/?p=118347
“Be kind first. Push back when you’re wronged. Then don’t keep score. Turns out that isn’t just nice advice, it’s the strategy that won the most famous game-theory contest ever run.” https://bgodinspired.com/?p=118347
“One mistake can echo forever between two people who both keep score. The fix that worked in the research was surprisingly simple: let some of it go.” https://bgodinspired.com/?p=118347
Common Questions About the Prisoner’s Dilemma
What is the prisoner’s dilemma in simple terms?
The prisoner’s dilemma is a puzzle about two people who each choose to cooperate or betray without knowing what the other will do. Betraying looks like the smart choice for each person alone, but if both betray, both end up worse off than if both had cooperated. It shows how people acting in their own interest can land in a result that is bad for everyone.
Who won Robert Axelrod’s prisoner’s dilemma tournament?
A strategy called Tit for Tat, submitted by Anatol Rapoport of the University of Toronto, won both of political scientist Robert Axelrod’s computer tournaments around 1980. Tit for Tat cooperates on the first move and then copies whatever the other player did on the previous move. It was the simplest program entered in the first tournament, which had 14 entries, and it won again against a field of 62 entries in a second, larger round.
Why does Tit for Tat win if it never beats its opponent?
Tit for Tat never betrays first, so in any single match it can only tie or slightly lose against the player it faces. It wins a tournament on total score because it builds long runs of cooperation with cooperative players, which earn high points for both sides, while aggressive strategies get stuck in costly feuds with each other.
What is Generous Tit for Tat?
Generous Tit for Tat is a forgiving version of Tit for Tat. It answers cooperation with cooperation, but after being betrayed it sometimes cooperates anyway, forgiving roughly one in three betrayals. In a 1992 study in Nature, Martin Nowak and Karl Sigmund found it outperformed plain Tit for Tat in a world where mistakes happen, because the occasional forgiveness breaks endless cycles of revenge.
Does the prisoner’s dilemma mean you should always forgive?
No. Strategies that always cooperate are easily exploited, and in Axelrod’s second tournament, the more patient Tit for Two Tats lost to Tit for Tat. The most successful strategies are kind first, push back when betrayed, and then return to cooperation quickly instead of holding a grudge.
