§1.2, p. 8 — leaving to ADX2 while cross-targeting earns (3), unprofitable iff (No Leave & Snipe)
ProvedAdDataSharing.Base.leave_snipe_deviationcross-targetingonline-advertisingp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Assume the standing assumptions and suppose allows cross-targeting. Start from full cross-targeting, where both advertisers share data with and each earns
If advertiser deviates to but continues to cross-target its rival on ("leave and snipe"), it earns
and (3) (2) holds if and only if
This is one of the two conditions that make full cross-targeting an equilibrium (Lemma 2).
Formalization Note. The payoffs do not depend on the snipe decisions of advertisers on ; the statement holds for all of them.
Preamble
import Mathlib import Definitions.Def_AdDataSharing_Base_Model
Formal statement
namespace AdDataSharing.Base
/-- §1.2, p. 8, comparing (3) to (2): from full cross-targeting, the deviation to ADX2
while still cross-targeting the rival ("leave and snipe") earns (3), and it is not profitable
if and only if (No Leave & Snipe) holds. -/
theorem leave_snipe_deviation (P : Params) (hP : P.Standing) (s s' : Bool) :
payoffOpen P ⟨.adx2, true⟩ ⟨.adx1, s⟩ =
P.α * P.M + (1 - P.α) * P.C - 2 * (1 - P.α) * P.w + P.η₁ * (P.C - P.w) ∧
payoffOpen P ⟨.adx1, s'⟩ ⟨.adx1, s⟩ = (1 + P.η₂) * (P.C - P.w) ∧
(payoffOpen P ⟨.adx2, true⟩ ⟨.adx1, s⟩ ≤ payoffOpen P ⟨.adx1, s'⟩ ⟨.adx1, s⟩ ↔
NoLeaveSnipe P) := by sorry
end AdDataSharing.Base
Source
Johnson, Jungbauer, Preuss, Online Advertising, Data Sharing, and Consumer Control, SSRN 3898798 (Nov. 2, 2022), p. 8, §1.2, display (3) and comparing (3) to (2), (No Leave & Snipe)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.