Lemma 3.14 (corrected: the extended co-measure copies γ) — extensions preserve γ-potential, nonstrategic, μ-normalized and (μ,γ)-harmonic games
ProvedDecompGames.Duplication.lemma_3_14Let be a finite set of players; player has a finite strategy set , and is the set of strategy profiles. A game is a vector of payoff functions ; is the space of games. For let be a strictly positive measure on , the product measure, , and let be a co-measure vector, a strictly positive function of .
Let be a collapsing map, the extended game and the extended co-measure vector (the extending map applied to ), and let be a strictly positive extension of : for all and . Duplicating the strategy of player into (Definition 3.13) is the case , , all other 's the identity. For every game :
- (i) if is -potential, then is -potential;
- (ii) if is nonstrategic, then is nonstrategic;
- (iii) if is -normalized, then is -normalized;
- (iv) if is -harmonic, then is -harmonic.
Each class of the decomposition is mapped into the corresponding class of the extended game; with uniqueness of the decomposition this gives Theorem 3.15.
Differences from the source:
- generalization: standing assumptions n ≥ 2 players and |Sⁱ| ≥ 2 strategies for every player (p. 7) → any finite type of players ι and any finite strategy sets S i, of any size; reason: the result holds without these assumptions (with an empty Sⁱ the space of profiles is empty and every game is 0); every instance of the source is an instance of the Lean statement
- encoding: μⁱ a strictly positive finite measure on Sⁱ → μ i : S i → ℝ, the point masses, with 0 < μ i a for all a (IsMeasureVec); reason: a measure on a finite set is determined by its point masses; μⁱ(Sⁱ) is mass μ i = ∑ₐ μ i a
- encoding: γⁱ a co-measure, a strictly positive measure on S⁻ⁱ (Definition 2.1) → γ i : (∀ j, S j) → ℝ positive with γ i (Function.update s i a) = γ i s (IsCoMeasureVec); reason: functions on S⁻ⁱ are exactly the functions on S that do not depend on the coordinate sⁱ
- encoding: statements over s⁻ⁱ ∈ S⁻ⁱ and profiles (tⁱ, s⁻ⁱ) → statements over full profiles s : ∀ j, S j, with (tⁱ, s⁻ⁱ) written Function.update s i tⁱ; reason: the coordinate sⁱ of s is overwritten, so the quantifier ranges over the same subprofiles
- encoding: g nonstrategic: for each i there is ℓⁱ : S⁻ⁱ → ℝ with gⁱ(sⁱ, s⁻ⁱ) = ℓⁱ(s⁻ⁱ) (2.4) → gⁱ(Function.update s i a) = gⁱ(s) for all i, s, a (IsNonstrategic); reason: a function on S is of the form ℓⁱ(s⁻ⁱ) exactly when it does not depend on sⁱ
- generalization: a single duplication: Šⁱ = Sⁱ ∪ {s⁰ⁱ}, ǧ(s⁰ⁱ, s⁻ⁱ) = g(s¹ⁱ, s⁻ⁱ), μ̌ⁱ(s⁰ⁱ) + μ̌ⁱ(s¹ⁱ) = μⁱ(s¹ⁱ), μ̌ = μ elsewhere (Definition 3.13) → any family of collapsing maps πʲ : S'ʲ → Sʲ, with ǧʲ(s') = gʲ(π(s')) (extend π g) and μ' strictly positive with ∑_{b ∈ (πʲ)⁻¹(a)} μ'ʲ(b) = μʲ(a) (IsMeasureExtension); reason: Definition 3.13 is the case S'ⁱ = Sⁱ ∪ {s⁰ⁱ}, πⁱ(s⁰ⁱ) = s¹ⁱ, πⁱ = id on Sⁱ, πʲ = id for j ≠ i, where the fibre condition is exactly (i)–(ii) of the extension of μ; the Lean statement also covers several duplications at once
- encoding: μ̌ an extension of μ (a measure, strictly positive as every measure of Definition 2.2) → μ' with the hypothesis IsMeasureVec μ' (0 < μ' i b); reason: the classes of Definition 2.2 are defined for strictly positive measures
- correction: γ̌ an extension of γ: γ̌ⁱ(s⁻ⁱ) = γ⁻ⁱ(s⁻ⁱ) and, for j ≠ i, γ̌ʲ(s⁰ⁱ, s^{−(i,j)}) + γ̌ʲ(s¹ⁱ, s^{−(i,j)}) = γʲ(s¹ⁱ, s^{−(i,j)}) (Definition 3.13) → γ̌ = Ť(γ), γ̌ʲ(s') = γʲ(π(s')): γ̌ⁱ = γⁱ and γ̌ʲ(s⁰ⁱ, ·) = γ̌ʲ(s¹ⁱ, ·) = γʲ(s¹ⁱ, ·); reason: with the printed sum (i) is false: two players, Sʲ = {sʲ, tʲ}, μ ≡ 1, γ ≡ 1, g¹ = g² = Ψ with Ψ(s¹,s²) = Ψ(t¹,t²) = 1, Ψ(s¹,t²) = Ψ(t¹,s²) = −1 (γ-potential); duplicate s¹ with μ̌¹ = (½, ½, 1) and γ̌²(s¹₀,·) = γ̌²(s¹₁,·) = ½; a γ̌-potential Ψ̌ of ǧ would need Ψ̌(t¹,t²) − Ψ̌(t¹,s²) = 3 ≠ 2 = g²(t¹,t²) − g²(t¹,s²). The proof on p. 40 uses the copy, which also makes γ̌ unique ("the extension of γ"); γ⁻ⁱ in (i) is read as γⁱ
import Mathlib import Definitions.Def_DecompGames_Duplication_Extension
namespace DecompGames.Duplication
open Finset
/-- Lemma 3.14, p. 18, for an extension along any collapsing map `π` (a single duplication is the
case of Definition 3.13), with the extended co-measure `γ̌ = Ť(γ)` (`γ̌ʲ(s') = γʲ(π(s'))`):
(i) `γ`-potential ⇒ `ǧ` is `γ̌`-potential; (ii) nonstrategic ⇒ `ǧ` nonstrategic;
(iii) `μ`-normalized ⇒ `ǧ` is `μ̌`-normalized; (iv) `(μ, γ)`-harmonic ⇒ `ǧ` is `(μ̌, γ̌)`-harmonic. -/
theorem lemma_3_14 {ι : Type} [Fintype ι] [DecidableEq ι] {S S' : ι → Type}
[∀ i, Fintype (S i)] [∀ i, DecidableEq (S i)]
[∀ i, Fintype (S' i)] [∀ i, DecidableEq (S' i)]
(μ : ∀ i, S i → ℝ) (hμ : IsMeasureVec μ)
(γ : ι → (∀ i, S i) → ℝ) (hγ : IsCoMeasureVec γ)
(π : ∀ i, S' i → S i) (μ' : ∀ i, S' i → ℝ) (hμ' : IsMeasureVec μ')
(hext : IsMeasureExtension π μ μ') (g : Game ι S) :
(IsPotential γ g → IsPotential (extend π γ) (extend π g)) ∧
(IsNonstrategic g → IsNonstrategic (extend π g)) ∧
(IsNormalized μ g → IsNormalized μ' (extend π g)) ∧
(IsHarmonic μ γ g → IsHarmonic μ' (extend π γ) (extend π g)) := by sorry
end DecompGames.Duplication
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.