Game Theory

Game theory is the study of decisions where the result depends not only on what you choose, but on what other people choose too. Whenever you have to guess how someone else will act, and they are guessing about you, you are playing a game in this sense.

It began as mathematics and economics, but it now explains everything from price wars and arms races to why people tip, queue, vote and trust strangers. Its most famous puzzles show that individually sensible choices can add up to outcomes that are worse for everyone.

How real people decide

Where game theory came from

John von Neumann and Oskar Morgenstern founded the modern field with their 1944 book “Theory of Games and Economic Behavior”. They showed that strategic situations could be described precisely, with players, choices and payoffs.

In 1950 John Nash introduced the idea now called the Nash equilibrium, extending the theory to games that are not pure contests. The same year Merrill Flood and Melvin Dresher at the RAND Corporation devised the game that Albert Tucker framed as the prisoner’s dilemma.

The prisoner’s dilemma

Two suspects are questioned separately. If both stay silent, both get a short sentence. If one betrays the other, the betrayer goes free and the silent one gets a long sentence. If both betray, both get a fairly long sentence.

Whatever the other person does, betraying gives you a better result, so a purely self-interested player always betrays. Yet if both reason that way, they end up worse off than if both had cooperated. That gap between individual logic and shared benefit is the heart of the dilemma.

The same structure appears in climate agreements, doping in sport, advertising battles between rival firms and overfishing. Each side gains by defecting, and everyone loses when all of them do.

Nash equilibrium explained

A Nash equilibrium is a set of choices where no player can do better by changing their own move while everyone else keeps theirs. It is a stable point, not necessarily a good one.

In the prisoner’s dilemma, mutual betrayal is the equilibrium even though mutual cooperation pays both players more. Many games have several equilibria, and then the real question becomes which one people coordinate on.

Other classic games

The stag hunt, drawn from Jean-Jacques Rousseau: two hunters can catch a stag together or each settle for a hare alone. Hunting the stag pays more, but only if you trust the other person to stay. It is a game about trust rather than temptation.

Chicken: two drivers race towards each other. Swerving looks weak, but if neither swerves both crash. It models standoffs where backing down is costly and not backing down is catastrophic.

The ultimatum game: one person proposes how to split a sum of money, and the other can accept or reject. If they reject, nobody gets anything. Theory says any offer should be accepted, yet many people turn down offers they see as unfair.

Public goods games: everyone can contribute to a shared pot that is multiplied and split equally. The group does best when all contribute, but each person does best by keeping their money while others pay in.

First-mover and coordination games: some situations reward whoever commits first, while others reward simply choosing the same option as everyone else, such as which side of the road to drive on.

Why real people do not play like the theory

Classical game theory assumes players are rational and care only about their own payoff. Experiments since the 1950s have shown that real people cooperate more, punish unfairness even at a cost to themselves, and care about intentions and reputation.

In Robert Axelrod’s computer tournaments around 1980, a simple strategy called tit for tat, submitted by Anatol Rapoport, did remarkably well in the repeated prisoner’s dilemma. It cooperated first, then copied the other player’s last move. Being nice, retaliatory, forgiving and clear turned out to be a winning combination.

Behavioural game theory studies these gaps. It treats fairness, trust and reciprocity as part of how humans actually decide, rather than as mistakes.

Why it matters for AI

AI agents increasingly negotiate, trade, bid and coordinate with people and with each other. An agent that plays strictly by textbook equilibrium may behave in ways humans see as cold, exploitative or untrustworthy.

Resonant records how real people choose in these strategic situations, anonymously, so the gap between theoretical and human play can be measured and used to check whether AI systems act the way people expect.

Common questions

What is game theory in simple terms?

It is the study of decisions where your best choice depends on what others choose. It models players, their options and the payoffs of each combination.

What is the prisoner’s dilemma?

A game where each player gains by betraying the other, but both are worse off if both betray than if both cooperate.

What is a Nash equilibrium?

A set of choices where no player can improve their result by changing only their own decision. It is stable, but not always the best outcome for the group.

What is the best strategy in the prisoner’s dilemma?

In a single round, betrayal is the rational choice. In repeated games, cooperative strategies that reciprocate, such as tit for tat, often do better.

Do people actually behave rationally in games?

Not in the narrow sense. People often cooperate, reject unfair offers and reward trust, which is why real results differ from textbook predictions.

Can I play these games myself?

Yes. Resonant lets you answer classic game theory scenarios anonymously and compare your choice with how other people decided.

All game theory results · The Trolley Problem · Game Theory · Human–AI Alignment · Articles · Resonant