By the time you reach secondary school, club matches and test rankings force you to plan while guessing what others will do. Game theory turns this strategic “mind-reading” into mathematics. Schelling simplified real issues—like nuclear arms talks or traffic jams—into neat diagrams and showed how a “credible threat” or “burning one’s bridges” can change negotiations. Aumann used rigorous math to prove that in a “repeated game,” where the same opponents meet again and again, long-run cooperation is often the smartest choice. This insight explains price cartels or why environmental treaties hold. Their ideas reach far beyond the well-known Prisoner’s Dilemma, affecting everything from stopping online flaming to understanding deterrence in world politics.

Aumann and Schelling closed two major gaps in non-cooperative game theory. First, Schelling’s The Strategy of Conflict (1960) demonstrated, with minimal formalism, how option-cutting, credible threats, and focal points steer equilibrium selection in negotiations, laying groundwork for later concepts such as subgame-perfect equilibria and Selten’s credibility criteria. Second, Aumann formalized the Folk Theorem for infinitely repeated, discounted games and, through his common-knowledge framework, derived precise conditions under which cooperative equilibria arise. He further introduced “correlated equilibrium,” an extension of Nash equilibrium that endogenizes strategy dependence via external signals. These insights permeate analyses of cartel sustainability, international security, public-good provision, and even smart-contract design in fintech.

Schelling emphasized the topological properties of negotiation sets and discontinuities along the Pareto frontier in two-player strategic-form games, reconnecting the minimax debate with empirical phenomena. His focal-point notion anticipates behavioral equilibrium selection and foreshadows K–L equilibrium and experimental game-theory findings. Aumann analyzed infinitely repeated games on complete measure spaces using σ-algebras and Banach limits, proving that ε-equilibrium families exhaust the Folk-theorem region as δ→1. By defining common knowledge via intersecting event hierarchies, he derived the Agreement Theorem that constrains posterior beliefs. For correlated equilibrium, he reduced the problem to linear-constraint optimization over Lebesgue vectors, demonstrating how communication devices enlarge the equilibrium set. These results underpin modern information design, mechanism robustness, and dynamic-contract theory.