For thirteen days in October 1962 the world sat at the edge of nuclear war. Neither the US nor the USSR wanted to fire, but neither wanted to back down. That is a game of nerve, and Schelling used it to explain how threats work when carrying them out would be insane.
Deterrence and brinkmanship
The Cuban Missile Crisis is often modeled as Chicken: two sides on a collision course, each hoping the other swerves, disaster if neither does. The model is contested, but it captures the core puzzle of deterrence, how to make a threat credible when executing it would be catastrophic for you too.
In Chicken, the worst outcome for both is when...
A threat is credible only if the threatener would actually...
Predict, then play
The static Chicken game has two pure equilibria and a 10% symmetric mixing rate. Brinkmanship adds a prior commitment: choose how much genuine disaster risk to create, then watch the rival respond.
Why did that happen?
Take a guess, then reveal the reason
Both sides prefer to be the one who holds firm while the other swerves, so the pure equilibria are lopsided, and neither wants to be the swerver. A flat threat to hold no matter what is not credible, because if the other truly will not swerve, holding means catastrophe. Schelling’s insight was that you do not need a credible threat of certain disaster; you need a credible risk of it. By committing to a course that raises the chance of an accidental slide into war, you make backing down the other side’s best move, even though nobody chose the disaster outright.
The model, and its limits
Write the crisis as Chicken: mutual backing-down , one holds and one swerves , mutual holding a catastrophe . There are two pure equilibria and a symmetric mixed equilibrium. In that static mixed equilibrium, each side independently randomizes to make the other indifferent, giving a 10% probability of holding firm. Brinkmanship is a different object: in a sequential confrontation, one side deliberately changes the probability that escalation becomes uncontrollable in order to alter the rival’s choice. The mixed-Nash probability is not automatically the optimal brinkmanship risk.
Schelling argued that a threat of certain mutual destruction is not credible (you would not push the button), but a threat that creates a genuine risk of accidental escalation can be. By taking steps that partly remove your own control, a blockade, an alert posture, you commit to a probability of disaster, making the opponent’s best response to concede. Caveat: modeling the crisis as static Chicken is contested; Dixit and others argue a dynamic, sequential-bargaining account fits the actual back-and-forth better. Treat Chicken as the entry point, not the last word.
Work it out
Brinkmanship as a committed risk
In the static mixed equilibrium, a side holds firm with probability chosen to make the opponent indifferent between holding and swerving. Setting gives . This is a symmetric equilibrium mixing rate, not a unilateral commitment choice.
Brinkmanship operationalizes this: rather than threaten certain war (incredible), a leader takes actions that raise the chance of an uncontrolled slide, a naval blockade, forces on hair-trigger alert, so that the rival, facing a genuine probability of disaster, prefers to yield. Kennedy’s blockade did exactly this, creating pressure without a first strike. The credibility came from the risk, not from a promise to end the world on purpose.
Key idea
Deterrence in a game of nerve works not by threatening certain catastrophe, which is not credible, but by committing to a genuine risk of it. That is brinkmanship, and Chicken is a useful, if contested, model of the Cuban Missile Crisis.
Check yourself
Sources
- Schelling, T. C. (1960). The Strategy of Conflict. Harvard University Press.
- Schelling, T. C. (1966). Arms and Influence. Yale University Press.