ZeroSumChance.GameTree.value_prop
theoremverifiedformalizationimported Lean
Soundness of DStrategy: the backward-induction value is a lower bound on the outcome A achieves by following DStrategy, regardless of how B plays. Formally: for every B-strategy SB and game tree t, t.value ≤ t.outcome DStrategy SB.
Statement
set_option autoImplicit true in set_option relaxedAutoImplicit true in theorem ZeroSumChance.GameTree.value_prop (SB : Strategy) {t : GameTree} : t.value ≤ t.outcome DStrategy SB
- Formalizes
- EconCSLib, `ZeroSumChance.GameTree.value_prop`
- Lean proof
- By xbei (from the file's git history), from github.com/gametheoryinlean/EconCSLib at
1a88f809b538(Copyright (c) 2026 EconCSLib contributors. All rights reserved, Apache-2.0). Compiled here against leanprover/lean4:v4.33.1, Mathlib v4.33.1; written for leanprover/lean4:v4.30.0.ZeroSumChance.GameTree.value_prop← ZeroSumChance.GameTree.value_prop (EconCSLib/GameTheory/ExtensiveGame/ZeroSumGameTreeWithChance.lean:160): verbatim
- Verified
- 08 Oct 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 8
- ZeroSumChance.GameTree.outcomeThe realized payoff for player A when A plays
SAand B playsSB. Nature's moves are resolved by their fixed probabilities. - ZeroSumChance.GameTreeA finite binary game tree for a 2-player zero-sum game with Nature. *
Leaf val— terminal node;valis A's payoff (B gets-val). *… - ZeroSumChance.SelectA binary choice:
l= left branch,r= right branch. - ZeroSumChance.PlayerThe two players in a zero-sum game: A is the maximizer, B is the minimizer.
- ZeroSumChance.instInhabitedGameTreeZeroSumChance.instInhabitedGameTree
- ZeroSumChance.GameTree.valueThe backward-induction value of the game tree for player A. * A-node: A maximizes, so we take the max of both children's values. * B-node:…
- ZeroSumChance.GameTree.StrategyA pure strategy is a function that, given a player's decision node, selects one of the two branches.…
- ZeroSumChance.GameTree.DStrategyA's dominant strategy: at each A-node, move to whichever child has the higher value; ties go left.
- ZeroSumChance.GameTree.outcomeThe realized payoff for player A when A plays
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 autoImplicit true inset_option relaxedAutoImplicit true in/-- **Soundness of `DStrategy`**: the backward-induction value is a lower bound on the outcome A achieves by following `DStrategy`, regardless of how B plays. Formally: for every B-strategy `SB` and game tree `t`, `t.value ≤ t.outcome DStrategy SB`. -/theorem ZeroSumChance.GameTree.value_prop (SB : Strategy) {t : GameTree} : t.value ≤ t.outcome DStrategy SB := by induction t with | Leaf r => simp [outcome, value] | Pnode p L R HL HR => match p with | Player.A => rw [value, outcome, DStrategy] split_ifs with h · -- DStrategy picked R (L.value < R.value); goal: max L.value R.value ≤ outcome … R exact (max_le (le_of_lt h |>.trans HR) HR) · -- DStrategy picked L (¬ L.value < R.value, i.e. R.value ≤ L.value) -- goal: max L.value R.value ≤ outcome … L exact (max_le HL (not_lt.mp h |>.trans HL)) | Player.B => rw [value, outcome] cases SB L R · -- B chose L exact min_le_left L.value R.value |>.trans HL · -- B chose R exact min_le_right L.value R.value |>.trans HR | Nnode prob L R HL HR => rw [outcome, value] have hpL : (prob : ℚ) * L.value ≤ prob * outcome DStrategy SB L := mul_le_mul_of_nonneg_left HL prob.2.1 have hpR : (1 - prob) * R.value ≤ (1 - prob) * outcome DStrategy SB R := mul_le_mul_of_nonneg_left HR (by linarith [prob.2.2]) linarith