Zero-Sum Game vs Coordination Game
Zero-Sum Game and Coordination Game are two Game Theory & Information concepts in AP Economics that students often mix up. A zero-sum game is a situation where one player's gain exactly equals another player's loss, so the total is unchanged. A coordination game is one where players do best by making the same choice, so it has two or more Nash equilibria and the problem is agreeing on one. Here is how they compare side by side.
Poker among friends is roughly zero-sum: winnings equal losses. Many real economic interactions, like voluntary trade, are positive-sum (both gain), which is why framing economics as zero-sum is usually a mistake.
Interests here are mostly aligned: both players want to match, and mismatching hurts them both. That produces several Nash equilibria instead of one, which makes the question not what is best but which outcome the players will settle on. Because any of them is stable once expectations point that way, conventions, past practice, and focal points do the real work; drivers keeping to one side of the road is the everyday case. Some coordination games have equilibria that are not equally good, so a group can be stuck in a worse one that nobody can escape alone. That is the difference from a prisoner's dilemma, where there is a single equilibrium and it is bad because the players' incentives genuinely conflict.
Zero-Sum Game vs Coordination Game: Pure Conflict Against Common Interest
| Zero-Sum Game | Coordination Game | |
|---|---|---|
| What the players want | Opposite outcomes, since one gain is the other's loss | The same outcome, since mismatching hurts both |
| Pure-strategy Nash equilibria | Often none, so the solution is a mix | Two or more, one for each way of matching |
| Value of announcing your move | Negative, it hands the opponent a best reply | Positive, an agreement enforces itself |
| What makes it hard | Staying unpredictable | Choosing which of several good outcomes to land on |
| Effect of a shared cue or habit | None, no convention helps you win | Decisive, it picks out one equilibrium |
| Standard examples | Matching pennies, hide and seek, splitting a fixed prize | Driving on the same side, both installing the same app |
Saying what you plan to do ruins you in one game and rescues you in the other
Cheap talk is the cleanest test between these two, because its value flips sign. Picture hide and seek. The hider picks the attic or the basement, the seeker picks one place to search, a match pays the seeker 1 and the hider negative 1, and a miss pays the reverse. Announcing your choice here defeats you. A hider who says attic and means it gets caught, so the message cannot be believed and cannot be acted on, which is why the only stable play is to use each hiding place half the time. Now picture two people who need to be on the same video platform. Both on the first pays 5 each, both on the second pays 4 each, and being on different platforms pays nothing. Say that you will be on the first at eight and the message enforces itself: once your partner believes you, joining you is their best reply, and once they join, you had no reason to lie. The same sentence is worthless in one game and settles the other outright. When a question lets the players talk before choosing, work out which world you are in before deciding whether the talk matters.
Most exam games are neither, and the mixed cases are where marks are lost
Pure conflict and pure common interest are the two ends of a scale, and the games set most often sit between them. A prisoner's dilemma holds a common interest, since both cooperating beats both defecting, alongside a private conflict, since each side would rather cheat. Treat it as zero-sum and you conclude that cooperation is impossible. Treat it as coordination and you conclude that a short conversation fixes everything. Both readings fail, because talk is worthless when defecting is a dominant move: your partner agrees to cooperate and then defects anyway. A second in-between case is a couple who want an evening together but disagree about where to spend it. Coordination governs the payoffs, since mismatching is the worst result for both, and there is still conflict over which match happens, so a message is believable about wanting to meet and self-serving about the venue. Before labeling any game, check two things separately: whether the joint total varies across cells, and whether the players rank the matched outcomes the same way.
Frequently asked questions
Is a coordination game the opposite of a zero-sum game?
A coordination game sits at the opposite end of the scale from a zero-sum game, since its players share an interest in matching while zero-sum players are in complete conflict. Calling them exact opposites is fine as a definition but hides the fact that most games fall in between, with a prisoner's dilemma holding conflict and common interest at once. The difference worth remembering in practice is that conventions and communication solve coordination games and do nothing at all in zero-sum ones.
Why does a coordination game have more than one Nash equilibrium?
Coordination games reward matching, so every way of matching becomes a candidate equilibrium. If your partner is on one platform, joining them beats standing alone, and the identical argument holds for the other platform. That gives at least two stable outcomes, plus a mixed one in which both randomize and sometimes miss each other. The difficulty is not finding an equilibrium but agreeing on which, which is why habits, defaults and prior announcements carry so much weight in these games.
Can players in a coordination game end up at the worse equilibrium?
Landing on the worse equilibrium of a coordination game is common and fully rational. Put a 70 percent chance on your partner choosing the option worth 4 and your expected payoff from joining them is 2.8, against 1.5 from holding out for the option worth 5. Each of you acting on those beliefs makes them true, and the pair stays stuck on the inferior match until something moves expectations, such as an announcement, a default setting or a deadline.
Get AP Econ exam tips in your inbox
Occasional emails with study tips, new interactive graphs, and exam-season reminders. Free, no spam.
No spam. Unsubscribe anytime. Read our privacy policy.
Keep track of what you have studied
A free EconLearn account adds progress tracking, your quiz history, and achievements. Studying here is free either way, and there is nothing to pay for as a student.
Create a free accountAlready have one? Sign in
Last updated