Lemma 4.15 — the biased min work mechanism is strongly truthful for all parameters
ProvedAlgMechDesign.Randomized.bmw_strongly_truthfulmechanism-designp2o-batch-p200bp2o-gran-per-chapterp2o-plan-paperp2o-v1scheduling
Consider task scheduling with two agents and tasks. For every real and every bit vector , the biased min work mechanism with parameters and is strongly truthful on positive types:
- for every positive declaration of the other agent, every positive true type and every positive misreport , agent 's utility from declaring is at least that from declaring ;
- for every positive and every positive , there is a positive declaration of the other agent under which declaring gives agent strictly smaller utility than declaring .
Here agent 's utility is the payment it receives minus over the tasks it is allocated. This lemma is the source of the universal truthfulness of the randomized mechanism.
Formalization Note "For all parameter values" is read as all (the parameter range of Fig. 1) and all ; it is stated for every number of tasks.
Preamble
import Mathlib import Definitions.Def_AlgMechDesign_Randomized_Model import Definitions.Def_AlgMechDesign_Randomized_BiasedMinWork
Formal statement
namespace AlgMechDesign.Randomized
/-- Lemma 4.15: for every parameter `β ≥ 1`, every bit vector `s ∈ {1, 2}ᵏ` and every number
`k` of tasks, the biased min work mechanism is strongly truthful on positive types. -/
theorem bmw_strongly_truthful {k : ℕ} (β : ℝ) (hβ : 1 ≤ β) (s : Fin k → Fin 2) :
IsStronglyTruthful (bmwAlloc β s) (bmwPay β s) := by sorry
end AlgMechDesign.Randomized
Source
Nisan, Ronen, Algorithmic Mechanism Design, Games Econ. Behav. 35, 2001, p. 182, Lemma 4.15 (mechanism of Fig. 1)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.