What Is the Prisoner's Dilemma?
In short
The prisoner's dilemma is a two-player game in which each player does better by defecting, whatever the other does, yet two defectors end up worse off than two cooperators.[1] It is usually credited to Merrill Flood and Melvin Dresher, who tested it at RAND in early 1950; Albert Tucker added the story of two prisoners deciding whether to confess.[1]

Key facts
- Usually credited to
- Merrill Flood and Melvin Dresher, RAND Corporation (experiment in early 1950)[1]
- Prisoner story
- Albert Tucker, in a 1950 handout for Stanford psychologists[1]
- Name first in print
- Luce and Raiffa, 1957 (per Guerra-Pujol and Martinez-Garcia, cited by the SEP)[1]
- The two moves
- Cooperate (stay silent) or defect (confess)[1]
- Only strict Nash equilibrium
- Both players defect[1]
- Axelrod's example payoffs
- Defect alone $5; both cooperate $3 each; both defect $1 each; cooperate alone $0[3]
- Winner of Axelrod's tournaments
- Tit for tat, submitted by Anatol Rapoport (won both rounds)[4]
The story of the two prisoners
Two partners in crime are arrested and held in separate cells, and a prosecutor offers each of them the same choice: confess or stay silent.[1] In the version told by the Stanford Encyclopedia of Philosophy (SEP), a prisoner who confesses while the partner stays silent walks free, and the silent partner gets a heavy sentence.[1] If both confess, both are convicted but get early parole; if both stay silent, each gets only a token sentence on a minor firearms charge.[1]
Whatever the partner does, each prisoner is better off confessing, yet when both confess, each ends up worse off than if both had kept quiet.[1] A common reading is that the puzzle shows a conflict between what is rational for each individual and what is rational for the group.[1]
A payoff table for the prisoner's dilemma
Game theorists reduce the story to two moves: "cooperate" (stay silent) and "defect" (confess).[1] The political scientist Robert Axelrod gave one version with payoffs in dollars, shown below.[3]
| Other player cooperates | Other player defects | |
|---|---|---|
| You cooperate | $3 each | You $0, other player $5 |
| You defect | You $5, other player $0 | $1 each |
In general, the four outcomes are called the temptation (defecting alone), the reward (both cooperate), the punishment (both defect) and the "sucker" payoff (cooperating alone), and a prisoner's dilemma requires them to be ranked in that order.[1][4]
Why defecting is the dominant strategy
If the other player cooperates, defecting pays $5 rather than $3; if the other player defects, defecting still pays $1 rather than nothing.[3] Because defecting is better whatever the other side does, game theorists say it strictly dominates cooperating, so two rational players both defect and collect the punishment instead of the larger reward.[1]
Mutual defection is also the game's only strict Nash equilibrium: the only outcome from which each player could only do worse by changing their own move alone.[1] John Nash developed this equilibrium concept for non-cooperative games, where binding agreements are not possible, in work reported in his 1950 doctoral dissertation, and shared the 1994 economics prize in memory of Alfred Nobel with John Harsanyi and Reinhard Selten.[5]
Who invented the prisoner's dilemma?
The game is usually credited to Merrill Flood and Melvin Dresher, two of the academics the RAND Corporation recruited after the Second World War to think scientifically about military strategy.[1] In early 1950 they persuaded a couple of colleagues to play a game with this payoff structure a hundred times in a row.[1] The work appeared in a 1952 RAND research memorandum by Flood, "Some Experimental Games".[1][2]
The prison story came from Albert Tucker, who wanted to make Flood and Dresher's game easier to grasp for an audience of psychologists at Stanford; the SEP dates his handout to 1950.[1] The name is often credited to Tucker, but the SEP cites Guerra-Pujol and Martinez-Garcia, who note that its earliest use in print is in Luce and Raiffa (1957).[1]
What changes when the game is repeated?
When the same players meet again and again, a defection can be punished, and cooperation rewarded, in later rounds.[1] If both players know exactly how many rounds there will be, an argument called backward induction says they should defect in the last round, therefore in the one before, and so on back to the first.[1][4] In practice, the SEP notes, people in long games of fixed length behave much as in games of unknown length, except in the final few rounds.[1]
Axelrod argued that cooperation can emerge only when the number of rounds is indefinite or unknown, and stays stable only if the next meeting matters enough: in his phrase, "the future must have a sufficiently large shadow".[3] The 2005 economics prize in memory of Alfred Nobel went partly to Robert Aumann, whom the prize announcement calls the first to carry out a full formal analysis of infinitely repeated games.[6]
Robert Axelrod's tournaments and tit for tat
Axelrod invited game theory experts to submit computer programs, and paired each strategy with every other in repeated play.[3] In the first round, described by Axelrod and William Hamilton in Science in 1981, 14 entries and a random strategy met in a round robin of 200-move games.[4] The highest average score went to the simplest entry, tit for tat, which cooperates on the first move and then copies the other player's previous move.[4]
Entrants to a second round knew the first round's results; Axelrod and Hamilton report 62 entries from six countries, while the SEP gives the number of entrants as sixty-three.[1][4] Anatol Rapoport, the first round's winner, submitted tit for tat again, and it won again.[4]
Axelrod linked its success to four traits: it is "nice", never the first to defect; it retaliates when provoked; it forgives after retaliating; and its behaviour is clear enough for others to recognise.[1][3] Enthusiasm has since cooled: in tournaments held in 2004 and 2005 for the twentieth anniversary of Axelrod's book, tit for tat finished only fourteenth of fifty strategies in the event closest to his original design.[1]
Real-world examples of the prisoner's dilemma
Axelrod pointed to two industrial nations that keep trade barriers against each other's exports: both would gain from free trade, but whatever one country does, the other is better off keeping its barriers.[3] He also described the "live and let live" system in the trenches of the western front in the First World War, where front-line soldiers often held back from shooting to kill as long as the other side did the same.[3]
The SEP lists defence spending by military rivals and price setting by two competing firms among situations said to have this structure, though it adds that they are better modelled by the repeated game.[1] It also points to "public goods dilemmas" such as clean air, where each person benefits from others' contributions whether or not they contribute.[1] The 2005 prize announcement says repeated-game theory helps explain price wars and trade wars, and why some communities manage shared resources better than others.[6]
Frequently asked questions
Why is it called the prisoner's dilemma?
The name comes from the story Albert Tucker used to explain Flood and Dresher's game: two arrested partners in crime must each decide, separately, whether to confess or stay silent.[1] Each does better by confessing, whatever the other does, yet both confessing leaves each worse off than both staying silent.[1] The label is often credited to Tucker, but the SEP cites researchers who found its earliest use in print in Luce and Raiffa (1957).[1]
What is the best strategy in the prisoner's dilemma?
In a single round, defecting is the dominant strategy, because it pays more whatever the other player does.[1] In repeated play, the SEP says no strategy scores highest against every possible group of opponents, though tit for tat won both of Robert Axelrod's computer tournaments.[1][4] In later tournaments held in 2004 and 2005, it finished only fourteenth of fifty in the event closest to Axelrod's design.[1]
Is mutual defection a Nash equilibrium?
Yes. Both players defecting is the game's only strict Nash equilibrium: from there, each player could only do worse by changing their own move alone.[1] The dilemma is that this outcome leaves both players worse off than mutual cooperation, from which each player would gain by switching to defection.[1][3]
What are real-life examples of the prisoner's dilemma?
Robert Axelrod cited two countries that keep trade barriers against each other even though both would gain from free trade, and listed arms races and military escalation among international problems that can take the form of a repeated dilemma.[3] The Stanford Encyclopedia of Philosophy mentions defence spending by military rivals and price setting by two competing firms as situations often said to have this structure, adding that they are better modelled as repeated games, as well as contributions to public goods such as clean air.[1]
Sources
- [1]Prisoner's Dilemma (Stanford Encyclopedia of Philosophy, substantive revision 24 October 2025) · Stanford Encyclopedia of Philosophy (Stanford University) · accessed October 8, 2026
- [2]Some Experimental Games (Research Memorandum RM-789-1) · RAND Corporation · accessed October 8, 2026
- [3]The Evolution of Cooperation (essay adapted from Axelrod's 1984 book, in Breakthrough: Emerging New Thinking) · Robert Axelrod, via Stanford University (Breakthrough online edition) · accessed October 8, 2026
- [4]The Evolution of Cooperation (Science, vol. 211, 1981) · Robert Axelrod and William D. Hamilton, Science (copy hosted by the University of Michigan) · accessed October 8, 2026
- [5]Press release: The Prize in Economic Sciences 1994 · The Royal Swedish Academy of Sciences (NobelPrize.org) · accessed October 8, 2026
- [6]Press release: The Prize in Economic Sciences 2005 · The Royal Swedish Academy of Sciences (NobelPrize.org) · accessed October 8, 2026
Cite this article
- APA
- BodaWiki Editorial. (2026, October 12). What Is the Prisoner's Dilemma? BodaWiki. https://bodawiki.com/wiki/what-is-the-prisoners-dilemma
- MLA
- “What Is the Prisoner's Dilemma?” BodaWiki, 12 Oct. 2026, bodawiki.com/wiki/what-is-the-prisoners-dilemma.
- Chicago
- BodaWiki Editorial. “What Is the Prisoner's Dilemma?” BodaWiki. October 12, 2026. https://bodawiki.com/wiki/what-is-the-prisoners-dilemma.