Game theory studies situations in which the outcome for each participant depends on the choices of others. Economists, political scientists and biologists group these situations into types, because the structure of a game, such as who moves when, how often it is played and whether players' interests clash or align, largely decides how it is likely to be played.
The types below are not exclusive: a game can be both sequential and repeated, and a coordination game can be played once or many times. Simultaneous and sequential games differ in timing; repeated games add the shadow of the future; coordination games and hawk-dove games differ in whether players want to match or to differ. Each is usually analysed with the tools of strategies, payoffs and best responses and the Nash equilibrium, a set of strategies from which no player can gain by changing alone.
Simultaneous Game
A simultaneous game is one in which players choose their actions without seeing what the others have chosen. They need not literally move at the same moment; what matters is that no one knows the others' choices when deciding. Such games are usually shown in normal form, as a payoff matrix listing each player's strategies along the rows and columns and the payoffs for every combination.
The best-known example is the prisoner's dilemma. Two suspects, questioned separately, each do better by confessing whatever the other does, so both confess, even though both would be better off staying silent. Sealed-bid auctions, rival firms setting prices or output at the same time, and rock-paper-scissors are other examples. Players must form beliefs about what others will do, and a Nash equilibrium is a situation where everyone's beliefs turn out to be correct. Some simultaneous games, such as rock-paper-scissors, have equilibria only in mixed strategies, where players randomise.
Sequential Game
A sequential game is one in which players move in turn, and later players observe at least some of the earlier moves before choosing. Such games are usually drawn as a game tree, with branches for each possible move at each stage and payoffs at the ends.
Sequential games are solved by backward induction: start at the last decision, work out what that player would choose, then step back to the previous decision knowing how the game will end, and so on to the start. The result is a subgame perfect equilibrium, a refinement associated with Reinhard Selten that rules out threats a player would not actually carry out. Timing can confer a first-mover advantage, as when a firm builds capacity before a rival decides whether to enter, or a second-mover advantage, when learning from others' moves is valuable. Chess, wage bargaining with offers and counter-offers, and market entry decisions are standard examples.
Repeated Game
A repeated game is a game played more than once by the same players, with each round's payoffs added together. Repetition changes behaviour because today's actions can be rewarded or punished tomorrow. Two players who would both defect in a one-off prisoner's dilemma may cooperate when they expect to meet again.
Strategies such as tit-for-tat, which cooperates in the first round and then copies the opponent's last move, performed well in computer tournaments run by Robert Axelrod around 1980. The folk theorem shows that in infinitely repeated games, or games with no known last round, many outcomes including full cooperation can be sustained if players value the future enough. If the number of rounds is finite and known, backward induction can unravel cooperation from the last round back to the first. Repeated games help explain the stability of cartels and tacit collusion in an oligopoly, and the value of reputation in banking and trade.
Coordination Game
A coordination game is one with two or more Nash equilibria in which players do best by choosing matching actions. The difficulty is not conflict but agreeing on which equilibrium to reach. Driving on the same side of the road is the simplest example: both left and right work, provided everyone does the same.
In the stag hunt, or assurance game, described by Jean-Jacques Rousseau, two hunters can together catch a stag, which is worth more, or each can safely catch a hare alone. Hunting the stag is the better equilibrium for both, but it is risky if a player doubts the other will cooperate. Coordination problems arise in technical standards, choice of currency and payment networks, where network effects make a method more useful as more people adopt it. Bank panics can be read in the same way: if depositors expect others to stay calm, staying is best; if they expect a run, withdrawing is best. Conventions, leadership and institutions help players settle on one outcome.
Hawk-Dove Game
The hawk-dove game, also called the game of chicken, models a conflict in which each player can be aggressive (hawk) or yield (dove). Two doves share a prize; a hawk facing a dove takes it all; and two hawks fight, which costs both more than the prize is worth. Each player wants to do the opposite of the other, so the two pure equilibria are one hawk and one dove.
The game was developed in evolutionary biology by John Maynard Smith and George Price in 1973 to explain why animal contests are often settled without serious injury. With a prize worth 4 and a fighting cost of 6, two hawks each get , a hawk against a dove gets 4 and the dove 0, and two doves get 2 each. The mixed equilibrium is to play hawk with probability : then hawk earns and dove earns , so neither does better. Price wars and brinkmanship in negotiations follow the same logic.
Comparing the types
| Type | Key feature | Typical tool | Example |
|---|---|---|---|
| Simultaneous | Choices made without seeing others' | Payoff matrix | Prisoner's dilemma |
| Sequential | Players move in turn | Game tree, backward induction | Market entry |
| Repeated | Same game played many times | Trigger strategies, folk theorem | Tacit collusion |
| Coordination | Gains from matching actions | Multiple equilibria | Side of the road |
| Hawk-dove | Gains from doing the opposite | Mixed strategies | Price war |
Unlike a zero-sum game, all five types can leave both players better or worse off together, which is why fairness and trust, discussed under social preferences and fairness, often shape how real people play them.
Ref 1100001750