payoffVsPure_averageStrategy
Average strategy's payoff against a pure column. Let $G$ be a finite two-player zero-sum game with $M \ge 1$ row actions and $N$ column actions, given by a payoff matrix $A$ with $0 \le A_{ij} \le 1$, and fix a pure column action $j$. Let $T > 0$ and let $p_0, \dots, p_{T-1}$ be probability distributions over the $M$ row actions, each presented as non-negative weights $p_t(i)$ summing to $1$, and let $\bar{p}$ be their average strategy, the mixed row strategy with weights $\bar{p}_i = \frac{1}{T}\sum_{t=0}^{T-1} p_t(i)$. Then the expected payoff of $\bar{p}$ against column $j$ satisfies \[ \sum_{i} \bar{p}_i \, A_{ij} \;=\; \frac{1}{T} \sum_{t=0}^{T-1} \sum_{i} p_t(i) \, A_{ij}. \]
Statement
set_option relaxedAutoImplicit false in set_option autoImplicit false in open Real Finset BigOperators in lemma payoffVsPure_averageStrategy {M N T : ℕ} [NeZero M] (G : ZeroSumGame M N) (hT : 0 < T) (strategies : Fin T → Fin M → ℝ) (h_nonneg : ∀ t i, 0 ≤ strategies t i) (h_sum : ∀ t, ∑ i : Fin M, strategies t i = 1) (j : Fin N) : payoffVsPure G (averageStrategy hT strategies h_nonneg h_sum) j = (∑ t : Fin T, ∑ i : Fin M, strategies t i * G.payoff i j) / T
- Source
- TCSlib, `payoffVsPure_averageStrategy`
- Lean proof
- By Karim Abdel Sadek, Mark Bedaywi, from github.com/Shilun-Allan-Li/tcslib at
c8a02591b1e5(Copyright (c) 2026 Karim Abdel Sadek and Mark Bedaywi. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.25.0.payoffVsPure_averageStrategy← payoffVsPure_averageStrategy (TCSlib/LearningTheory/Minimax/HedgeInteraction.lean:132): verbatim
- Verified
- 07 Oct 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 3
- ZeroSumGameA finite two-player zero-sum game with $M$ row actions and $N$ column actions is a record consisting of a payoff matrix…
- averageStrategyGiven $T$ distributions $p_0, \dots, p_{T-1}$ over $\mathrm{Fin}\,n$ (each presented as non-negative weights summing to $1$) and $T > 0$,…
- payoffVsPureGiven a game $G$, a mixed row strategy $p$, and a pure column $j$, the expected payoff against column $j$ is…
- Used by 1
- hedge_constructionHedge yields an approximate minimax pair. Let $G$ be a finite two-player zero-sum game with $M \ge 2$ row actions and $N \ge 1$ column…
Lean source view module on GitHub
Choose a tactic — a marked line — to see the goals Lean had before and after it, as in Lean’s infoview. Step with ← → or j k.
set_option relaxedAutoImplicit false inset_option autoImplicit false inopen Real Finset BigOperators in/-- The payoff of the averaged row strategy against a pure column `j` equals the time average of the per-round payoffs `Σ_i strategies t i · A(i, j)` against that column. [FS99, §5]. -/lemma payoffVsPure_averageStrategy {M N T : ℕ} [NeZero M] (G : ZeroSumGame M N) (hT : 0 < T) (strategies : Fin T → Fin M → ℝ) (h_nonneg : ∀ t i, 0 ≤ strategies t i) (h_sum : ∀ t, ∑ i : Fin M, strategies t i = 1) (j : Fin N) : payoffVsPure G (averageStrategy hT strategies h_nonneg h_sum) j = (∑ t : Fin T, ∑ i : Fin M, strategies t i * G.payoff i j) / T := by simp only [payoffVsPure, averageStrategy] calc ∑ x : Fin M, (∑ t : Fin T, strategies t x) / ↑T * G.payoff x j = ∑ x : Fin M, (∑ t : Fin T, strategies t x * G.payoff x j) / ↑T := by congr 1 ext x rw [← Finset.sum_mul] ring _ = (∑ x : Fin M, ∑ t : Fin T, strategies t x * G.payoff x j) / ↑T := by rw [Finset.sum_div] _ = (∑ t : Fin T, ∑ i : Fin M, strategies t i * G.payoff i j) / ↑T := by rw [Finset.sum_comm]