Zero-Sum Game vs Mixed Strategy
Zero-Sum Game and Mixed Strategy 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 mixed strategy is a plan to randomize over your moves with fixed probabilities, used when always making the same predictable choice would be exploited. 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.
A pure strategy commits you to one move every time. A mixed strategy assigns a probability to each move, so the opponent cannot predict you and cannot tailor a response. Games where any predictable pattern gets punished, such as a penalty kick or a pitcher choosing pitches, often have no equilibrium in pure strategies at all, and the only stable outcome is both sides randomizing. The equilibrium probabilities have a surprising property: each player mixes in the way that leaves the opponent indifferent between their own options. John Nash proved that every finite game has at least one equilibrium once mixed strategies are allowed, which is why a game that looks unsolvable in pure strategies still has an answer.
Zero-Sum Game vs Mixed Strategy: Structure Against Strategy
| Zero-Sum Game | Mixed Strategy | |
|---|---|---|
| What the term describes | The payoff structure of the whole game | One player's rule for choosing a move |
| Who it applies to | Every player at once, before anyone chooses | A single player inside a given game |
| How you spot it | Whatever one player wins, the other loses | A player assigns probabilities instead of committing to one move |
| Where the numbers come from | Read straight off the payoff table | Solved for, so the opponent is left indifferent |
| Does one imply the other | A zero-sum game with a saddle point is solved in pure strategies | Mixed equilibria appear in non-zero-sum games too |
| Question that tests it | Do the two payoffs add to the same total in every cell? | Is there no pure-strategy Nash equilibrium here? |
Your probabilities are solved from your opponent's payoffs, not your own
Take a kicker and a keeper who each choose Left or Right. The kicker's payoff is 2 when both go Left, 8 when the kicker goes Left and the keeper goes Right, 9 when the kicker goes Right and the keeper goes Left, and 3 when both go Right. The keeper receives the negative of each, so the table is zero-sum. No pair of pure moves is stable, because whatever the kicker settles on the keeper wants to match, and once the keeper matches the kicker wants to switch. Solve it by finding the mix that leaves the other side nothing to gain by switching. Let q be the chance the keeper guards Left. The kicker's payoff from Left is 2q plus 8 times one minus q, which simplifies to 8 minus 6q, and from Right it is 9q plus 3 times one minus q, which is 3 plus 6q. Setting those equal gives 12q equals 5, so q is 5 over 12, near 42 percent, and the kicker's expected payoff either way is 5.5. Now the kicker's own mix. Let p be the chance the kicker goes Left. The keeper's payoff from guarding Left is 7p minus 9 and from guarding Right is negative 5p minus 3, and equating them gives p equal to one half. Notice which payoffs each calculation used: your probabilities came out of the other player's column.
Zero-sum does not force mixing, and mixing is not confined to zero-sum
Both halves of that claim get tested. Start with a zero-sum game that needs no randomizing at all. Row chooses Top or Bottom, Column chooses Left or Right, and Row's payoffs are 4 and 6 across the top row and 1 and 3 across the bottom. Top beats Bottom for Row whatever Column does, and once Row is playing Top, Column holds Row down to 4 by choosing Left. That pair is stable, both players use one move with certainty, and the game is still zero-sum. Now run it the other way with a couple deciding between a cafe and a gym. Matching at the cafe pays 3 to the first person and 2 to the second, matching at the gym pays 2 and 3, and turning up in different places pays nothing to either. Cell totals of 5, 5, 0 and 0 rule out zero-sum immediately, yet the game still has a mixed equilibrium: the first person goes to the cafe with probability 0.6 and the second with probability 0.4, and neither can do better alone. So the presence of randomizing tells you nothing about whether the totals are fixed.
Frequently asked questions
Does every zero-sum game require a mixed strategy?
No, a zero-sum game needs mixing only when it has no pure-strategy Nash equilibrium, which is the situation in matching pennies or a penalty kick. A zero-sum game with a saddle point, where one player has a move that is best against everything and the other has a clear reply to it, is solved entirely in pure strategies. The habit of pairing the two ideas comes from the examples textbooks choose, not from the definition of either term.
How do you find the mixed strategy in a zero-sum game?
Solving for a mixed strategy starts from the other player's payoffs rather than your own. Write your probability as p, express the opponent's expected payoff from each of their pure moves as a function of p, set those two expressions equal, and solve. The p that comes out leaves the opponent indifferent, which is what stops them exploiting you. Repeat the step with the roles reversed for their probabilities, then check by confirming each player earns the same expected payoff whichever move they pick.
Can a mixed strategy exist in a game that is not zero-sum?
Mixed strategies appear in plenty of games where the payoff totals vary from cell to cell. The couple choosing between a cafe and a gym is the standard case: cell totals of 5, 5, 0 and 0 rule out zero-sum, and the game still has a mixed equilibrium alongside its two pure ones. Coordination games, entry games and the chicken game all behave this way. Randomizing is a response to being predictable, and predictability is punished in many games besides those with a fixed total.
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