Theorem 1 — every finite game has a correlated equilibrium
ProvedHartSchmeidler.Finite.correlated_equilibrium_existscorrelated-equilibriumfinite-gamesgame-theoryp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be a finite set of players. For every , let be a finite, nonempty set of pure strategies, and let be an arbitrary real payoff function on . There exists a probability vector on satisfying the correlated-equilibrium incentive inequalities:
This is Hart and Schmeidler's finite-game existence theorem, which supplies the finite core for their later infinite-game results.
Formalization Note Each strategy set must be nonempty to admit a probability vector. An empty player set is allowed: its profile set has one element and the incentive conditions have no instances. Finiteness and decidable equality are Lean enumeration requirements; payoffs have no sign or normalization restriction.
Preamble
import Definitions.Def_HartSchmeidler_Finite_Game
Formal statement
namespace HartSchmeidler.Finite
/-- Theorem 1, p. 19: every finite game has a correlated equilibrium in
the sense of condition (1). -/
theorem correlated_equilibrium_exists {ι : Type*} [Fintype ι] [DecidableEq ι]
(S : ι → Type*) [∀ i, Fintype (S i)] [∀ i, DecidableEq (S i)]
[∀ i, Nonempty (S i)] (h : ι → (∀ i, S i) → ℝ) :
∃ p : (∀ i, S i) → ℝ, IsCorrelatedEq h p := by sorry
end HartSchmeidler.Finite
Source
Hart and Schmeidler, Existence of Correlated Equilibria, Math. Oper. Res. 14 (1989), p. 19, Theorem 1 and condition (1); https://doi.org/10.1287/moor.14.1.18
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.