Proposition 5, p. 21 — under semi-decentralized dispute resolution the equilibrium ransom is π(a) − π(a^n_N) or (1 − δ)(γ(π(a) − π(a^d_N)) + b)
ProvedReviewBlackmail.SemiDec.proposition_5Consider the review-blackmail game under the semi-decentralized mechanism with timeliness , accuracy and penalty , with genuine reviews kept, under the standing assumptions. Let
The equilibrium ransom is the largest ransom at which settling is an equilibrium response and no other equilibrium response pays the seller more; the malicious consumer purchases iff such a ransom exceeds .
- If , then
and the malicious consumer purchases if and only if . 2. If , then
and the malicious consumer purchases if and only if .
A high enough penalty makes the mechanism irrelevant to blackmail; below the threshold the ransom tracks the seller's expected loss from using the mechanism.
Formalization Note "" is stated as IsGreatest of the set of ransoms at which settling prevails, never by its closed form. is added because the threshold divides by (at Lean's would place every in case 1). The conditions involving are kept as printed; they come from Proposition 4, which justifies the game's restriction to sellers who keep genuine reviews ("By Proposition 4, under the assumption that , it suffices to focus on the strategies ", p. 41). Ransoms are not restricted to ; in both cases the greatest element is positive, so the restriction would not change it.
import Mathlib import Definitions.Def_ReviewBlackmail_SemiDec_Setting
namespace ReviewBlackmail.SemiDec
theorem proposition_5 (M : Model) (hM : M.Standing) (hδ : M.δ < 1) :
((1 - M.γ * (1 - M.δ)) / (1 - M.δ) * (M.fut M.a - M.fut (M.post 1 1 .N)) ≤ M.b →
M.bLow < (1 - M.γ * (1 - M.δ)) / (1 - M.δ) * (M.fut M.a - M.fut (M.post 1 1 .N)) →
IsGreatest {r | M.SettlePrevails r} (M.fut M.a - M.fut (M.post 1 1 .N)) ∧
(M.Purchases ↔ M.p < M.fut M.a - M.fut (M.post 1 1 .N))) ∧
(M.bLow ≤ M.b →
M.b < (1 - M.γ * (1 - M.δ)) / (1 - M.δ) * (M.fut M.a - M.fut (M.post 1 1 .N)) →
IsGreatest {r | M.SettlePrevails r}
((1 - M.δ) * (M.γ * (M.fut M.a - M.fut (M.post (M.γ * (1 - M.δ)) 1 .N)) + M.b)) ∧
(M.Purchases ↔
M.p < (1 - M.δ) * (M.γ * (M.fut M.a - M.fut (M.post (M.γ * (1 - M.δ)) 1 .N)) + M.b))) := by sorry
end ReviewBlackmail.SemiDec
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.