AFTD Auto-Formalizing Theoretical Domains

is_aau_rule

def

An allocation rule satisfies alternating advantage under unanimity (AAU) if at every unanimous preference profile, for any two agents i and j receiving bundles of equal size and every step t, whichever agent has the more preferred object at step t has the less preferred object at step t + 1.

Statement

def is_aau_rule {Agent Object : Type*}
    (f : (Agent → LinearOrder Object) → Agent → Finset Object) : Prop
Defined in
scout/2611-09685-who-picks-first-compensatory-fairness-when-equal-division-i.md:280
Verified
09 Nov 2026
Axioms
none listed
Built on 2
  • TotalPreorderA total preorder: a preorder where ≤ is total (complete). This is weaker than LinearOrder — it does NOT require antisymmetry or…
  • LinearOrder.toTotalPreorderEvery LinearOrder is a TotalPreorder.

Read back from the Lean

Given types Agent and Object, and an allocation function f : (Agent → LinearOrder Object) → Agent → Finset Object, is_aau_rule f asserts that: For every linear order u_order on Object, and for all agents i and j: if the bundles allocated to i and j under the unanimous profile (where every agent has preference u_order) have equal cardinality, i.e., card(f (fun _ => u_order) i) = card(f (fun _ => u_order) j), then, letting li and lj be the elements of f (fun _ => u_order) i and f (fun _ => u_order) j respectively, sorted in descending order according to u_order (using · ≥ ·): for every natural number t such that t + 1 < card(f (fun _ => u_order) i): if the $t$-th element (0-indexed) of li is strictly greater (according to u_order) than the $t$-th element of lj, then the $(t+1)$-th element of lj must be strictly greater (according to u_order) than the $(t+1)$-th element of li.

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

/-- An allocation rule satisfies AAU if at every unanimous profile, between agents of equal bundle size, advantage strictly alternates from step to step. -/
def is_aau_rule {Agent Object : Type*}
    (f : (Agent → LinearOrder Object) → Agent → Finset Object) : Prop := ∀ (u_order : LinearOrder Object) (i j : Agent),
  ∀ (hcard : (f (fun _ => u_order) i).card = (f (fun _ => u_order) j).card),
    let _ := u_order
    let li := (f (fun _ => u_order) i).sort (· ≥ ·)
    let lj := (f (fun _ => u_order) j).sort (· ≥ ·)
    ∀ (t : ℕ) (ht : t + 1 < (f (fun _ => u_order) i).card),
      li.get ⟨t, by rw [Finset.length_sort]; omega⟩ > lj.get ⟨t, by rw [Finset.length_sort]; omega⟩ →
      lj.get ⟨t + 1, by rw [Finset.length_sort]; omega⟩ > li.get ⟨t + 1, by rw [Finset.length_sort]; omega⟩
See it in the graphDiscuss this resultChallenge it