AFTD Auto-Formalizing Theoretical Domains

normal_form_subgame_last

def

Given action types A : Fin (r + 1) → Type and payoff function u : Fin (r + 1) → ((j : Fin (r + 1)) → A j) → ℝ for an (r+1)-player game, the r-player subgame payoff function obtained by fixing player Fin.last r to pure action s : A (Fin.last r) assigns to each player i : Fin r and action profile a : (j : Fin r) → A j.castSucc the payoff u i.castSucc (Fin.snoc a s).

Statement

def normal_form_subgame_last {r : ℕ} {A : Fin (r + 1) → Type}
    (u : Fin (r + 1) → ((i : Fin (r + 1)) → A i) → ℝ) (s : A (Fin.last r)) :
    Fin r → ((i : Fin r) → A i.castSucc) → ℝ
Defined in
arXiv:2611.06076 Lemma 2.1
Verified
05 Nov 2026
Axioms
none listed
Used by 1
  • one_player_extension_approx_nashIn an (r+1)-player normal-form game with payoff functions u : Fin (r + 1) → ((j : Fin (r + 1)) → A j) → ℝ taking values in [0, 1]…

Read back from the Lean

This defines a function normal_form_subgame_last taking: - an implicit natural number r, - an implicit family of types A : Fin (r + 1) → Type representing the action spaces of r + 1 players, - a utility function u : Fin (r + 1) → ((i : Fin (r + 1)) → A i) → ℝ assigning to each player in Fin (r + 1) a real-valued payoff given a joint action profile, and - a fixed action s : A (Fin.last r) for the last player (Fin.last r). It produces the payoff profile for the first r players (indexed by Fin r, where player i : Fin r has action space A i.castSucc): for each player i : Fin r and joint action profile a : (j : Fin r) → A j.castSucc, their payoff is u i.castSucc (Fin.snoc a s), which is player i.castSucc's payoff under the full action profile where the first r players play a and the last player plays s.

Written by a model that saw only the Lean, never the English above. If the two disagree, that is worth a challenge.

Lean source view module on GitHub

/-- Subgame obtained by fixing the last player's pure action. -/
def normal_form_subgame_last {r : ℕ} {A : Fin (r + 1) → Type}
    (u : Fin (r + 1) → ((i : Fin (r + 1)) → A i) → ℝ) (s : A (Fin.last r)) :
    Fin r → ((i : Fin r) → A i.castSucc) → ℝ :=
  fun i a => u i.castSucc (Fin.snoc a s)
See it in the graphDiscuss this resultChallenge it