Theorem 2.8, proof (1) ⇒ (2): a potential has zero closed-path sum
ProvedMondererShapley.ClosedPath.closed_path_I_eq_zero_of_potentialgame-theoryp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1potential-games
Suppose is an exact potential for a game with payoff vector . For every finite closed path ,
Along each step, the deviator's payoff change equals the change in , so the path sum records only the difference between the terminal and initial potential values. This is the forward implication in Theorem 2.8.
Formalization Note Players form a finite type, strategy sets are arbitrary, and every path step changes exactly one coordinate to a different value.
Preamble
import Mathlib import Definitions.Def_MondererShapley_ClosedPath_IsPotential import Definitions.Def_MondererShapley_ClosedPath_FinPath
Formal statement
namespace MondererShapley.ClosedPath
variable {ι : Type*} [DecidableEq ι] {Y : ι → Type*}
/-- Theorem 2.8, proof of (1) implies (2), p. 138. -/
theorem closed_path_I_eq_zero_of_potential [Fintype ι]
(u : ι → (∀ i, Y i) → ℝ) (P : (∀ i, Y i) → ℝ)
(hP : IsPotential u P) (γ : FinPath Y) (hγ : γ.IsClosed) : γ.I u = 0 := by sorry
end MondererShapley.ClosedPath
Source
Monderer and Shapley, Potential Games, Games Econ. Behav. 14 (1996), p. 138 (PDF p. 15), Theorem 2.8, proof (1) ⇒ (2); https://doi.org/10.1006/game.1996.0044
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.