(17:C:f) — and iff is a saddle point of
ProvedTheoryOfGames.Minimax.good_iff_saddlePointgame-theorymixed-strategiesp2o-batch-b23ap2o-gran-per-chapterp2o-plan-bookp2o-v1saddle-pointzero-sum-games
Let be the matrix of a normalized zero-sum two-person game , the bilinear form (17:2), and , the sets of good strategies (17:B:a), (17:B:b). For all mixed strategies and ,
that is, iff for all , .
In the book's words: both players play well if and only if is a saddle point of . It is (13:D*) applied to , which is legitimate because a saddle point of always exists by the minimax theorem (17:6).
Preamble
import Mathlib import Definitions.Def_TheoryOfGames_Minimax_MixedStrategy
Formal statement
namespace TheoryOfGames.Minimax
/-- (17:C:f), p. 159: for mixed strategies `ξ ∈ S_{β₁}`, `η ∈ S_{β₂}` of the normalized
zero-sum two-person game with matrix `ℋ`, `ξ ∈ Ā` and `η ∈ B̄` if and only if `ξ, η` is a
saddle point of `K(ξ, η)`. -/
theorem good_iff_saddlePoint {β₁ β₂ : ℕ} (H : Fin β₁ → Fin β₂ → ℝ)
(ξ : Fin β₁ → ℝ) (η : Fin β₂ → ℝ)
(hξ : ξ ∈ stdSimplex ℝ (Fin β₁)) (hη : η ∈ stdSimplex ℝ (Fin β₂)) :
(ξ ∈ goodA H ∧ η ∈ goodB H) ↔ IsMixedSaddlePoint H ξ η := by sorry
end TheoryOfGames.Minimax
Source
von Neumann & Morgenstern, Theory of Games and Economic Behavior (60th-anniversary ed., Princeton 2007), p. 159, (17:C:f)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.