Coordination game
A game where players win by matching each other, not by outsmarting them.
What it means
A strategic situation with multiple equilibria in which players are rewarded for choosing compatible actions, so their interests are largely aligned rather than opposed. The difficulty is not conflict but selection: with several ways to coordinate, players must somehow converge on the same one, often guided by focal points, norms, or communication. Coordination games model driving on the same side of the road, technology standards, language, and meeting conventions. They matter because much of social order is solved coordination, and because they reveal that efficient outcomes are not automatic when several self-consistent equilibria compete.
Varieties and the selection problem
Coordination games come in flavours. In a pure coordination game the players are indifferent among the matched outcomes and care only about matching. In games with conflicting preferences, like Battle of the Sexes, both still prefer coordinating but disagree on which equilibrium to land on. In Pareto-ranked games, like the stag hunt, one equilibrium beats another for everyone yet is riskier to attempt. What unites them is that Nash equilibrium alone cannot say which will be played, because every equilibrium is self-consistent. Game theorists proposed refinements to break the tie: Harsanyi and Selten's payoff dominance picks the outcome best for all, while risk dominance picks the safer one. The two often disagree, and which principle actually governs behaviour is an empirical question rather than a theorem.
What the experiments show
Laboratory work turned selection into data. Van Huyck, Battalio and Beil's 1990 minimum-effort, or weakest-link, game gave groups an efficient equilibrium worth far more to everyone, yet players, unsure whether others would risk it, drifted toward the safe but inefficient one. It was a clean demonstration that coordination failure is real and distinct from any clash of interests. Cooper, DeJong, Forsythe and Ross showed that cheap-talk communication before play often rescues the efficient outcome, though a lone announcement helps less than messages flowing both ways. The broad finding is that history, group size, observability and the riskiness of the good equilibrium all shape whether coordination succeeds. Efficient equilibria are attainable but fragile; strategic uncertainty, not opposed interests, is usually what sinks them.
Where focal points fail
Schelling's focal points, the option that somehow stands out, coordinate people who cannot communicate, and Mehta, Starmer and Sugden confirmed in 1994 that shared salience lets strangers match on labels far better than chance. But salience is brittle. Crawford, Gneezy and Rottenstreich found in 2008 that adding even a tiny payoff asymmetry to an otherwise symmetric game, making one option worth slightly more to one player, collapses coordination: the label's pull and the payoff's pull point in different directions, players begin second-guessing, and mismatches multiply. The lesson for design is that a focal point is reliable only when nothing else competes for attention. Introduce the faintest reason to prefer one equilibrium over another, and the shared intuition that made coordination effortless can quietly evaporate.
Related but distinct
A coordination game is easy to confuse with a prisoner's dilemma, since both involve two parties and a jointly good outcome, but the incentives differ at the root. In a prisoner's dilemma each player has a dominant strategy to defect, so the efficient outcome is not an equilibrium at all; the trouble is a conflict between private and collective interest. In a coordination game the efficient outcome is an equilibrium, because nobody wants to deviate once everyone is there, and the only trouble is getting everyone to the same one. It also differs from pure conflict, or zero-sum, games, where one player's gain is exactly another's loss and there is nothing to coordinate on. Naming which structure you face decides whether the remedy is trust, a signal, or enforcement.
Examples
Everyone benefits from adopting the same file format, but the group must coordinate on which one to make it work.
Six friends all want to meet and none minds where; someone suggests the park, someone the cafe, and with no one naming a single spot they all stay home.
Separated at a festival with dead phones, a couple both head for the main stage, not because it is the best spot, but because each expects the other to expect it.
An orchestra tunes to the oboe's A not because that pitch is uniquely correct, but because every player expects the others to treat it as the shared reference.
A currency holds value only while everyone expects others to accept it; when that shared expectation cracks, people rush to switch and the coordination propping it up unwinds fast.
First described in Game theory; Schelling (1960).
Key references
- Crawford, V. P., Gneezy, U., & Rottenstreich, Y. (2008). The power of focal points is limited: Even minute payoff asymmetry may yield large coordination failures. American Economic Review, 98(4), 1443-1458. doi.org/10.1257/aer.98.4.1443
- Mehta, J., Starmer, C., & Sugden, R. (1994). The nature of salience: An experimental investigation of pure coordination games. American Economic Review, 84(3), 658-673. econpapers.repec.org/RePEc:aea:aecrev:v:84:y:1994:i:3:p:658-73
- Cooper, R., DeJong, D. V., Forsythe, R., & Ross, T. W. (1992). Communication in coordination games. Quarterly Journal of Economics, 107(2), 739-771. doi.org/10.2307/2118488
- Van Huyck, J. B., Battalio, R. C., & Beil, R. O. (1990). Tacit coordination games, strategic uncertainty, and coordination failure. American Economic Review, 80(1), 234-248. ideas.repec.org/a/aea/aecrev/v80y1990i1p234-48.html
- Schelling, T. C. (1960). The Strategy of Conflict. Harvard University Press. www.hup.harvard.edu/books/9780674840317