Nash Equilibrium

Two firms could both keep prices high and split a comfortable market. Instead they hammer each other down to razor margins and stay there — each one certain that raising its own price alone would just hand customers to the other. Two countries could both stay lightly armed. Instead they both stockpile weapons they’d rather not pay for, each unwilling to be the one who disarmed while the other didn’t. In both cases the players are stuck — not because anyone forced them, and not because they failed to think it through, but because given exactly what the other side is doing, no one can improve by moving alone. That stuck point has a name, and it is the single most important idea in all of strategy: the Nash equilibrium.

A Nash equilibrium is a combination of strategies — one for each player — where no player can raise their own payoff by changing only their own move, holding everyone else’s fixed. It is the resting point of a strategic situation: the outcome that, once reached, no one has a private reason to walk away from. In 1950 a 21-year-old named John Nash proved that if you allow players to randomise their choices, every finite game has at least one such resting point — a result so foundational it earned a Nobel Prize and rewired economics, political science, and evolutionary biology around it. The equilibrium is where strategic reasoning finally lands: after all the “if I do this, they’ll do that, so then I’d…” the chain stops at the cell no one wants to leave.

This is an advanced strategy course, and it assumes you’ve already met game theory basics — payoff matrices, dominant strategies, and the prisoner’s dilemma. From there it builds the equilibrium concept floor to ceiling. We’ll define it precisely as a point of no unilateral regret, then show it is really the place where every player’s best response meets every other’s — the still point of a hall of mirrors. We’ll see why a game can have one equilibrium, several, or none in pure strategies, and why Nash’s theorem still guarantees a mixed-strategy equilibrium exists once you’re allowed to randomise. Along the way we’ll work the three games that teach the whole model: the prisoner’s dilemma, whose one equilibrium is worse for everyone than the outcome they can’t reach (the deep lesson that stable ≠ good); coordination games with several equilibria, where the real problem is which one you land on (driving on the right, the stag hunt, focal points); and matching pennies, where the only equilibrium is to become genuinely unpredictable. Finally we’ll connect the model to incentives and leverage: an equilibrium is what a system of self-interested players settles into, so if you dislike the outcome, the move is to change the payoffs — not to scold the players. And we’ll be honest about where the model lies to you: equilibria need not be efficient or fair, real people don’t always reach them, and a Nash equilibrium is not the same thing as a dominant-strategy outcome. By the end you’ll be able to find the resting point of a strategic situation, predict where rational play will get stuck, and know which lever actually moves it.