Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Lemma 7, p. 11 — Σ_{i=1}^{k−1} ψ(i)(α_i − β_i) ≤ N/k

Proved
AdWordsMSVV.Tradeoff.lemma_7

by mikedeng1 · Oct 7, 2026 · Mathlib 0df444a (Lean v4.33.1)

adwordsonline-algorithmsp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Fix k≥1k\ge1k≥1, an adwords instance with NNN bidders, unit budgets and nonnegative bids cb,tc_{b,t}cb,t​, a monotonically decreasing tradeoff function ψ:{1,…,k}→R≥0\psi:\{1,\dots,k\}\to\mathbb R_{\ge0}ψ:{1,…,k}→R≥0​ with ψ(k)=0\psi(k)=0ψ(k)=0, a run σ\sigmaσ of the discrete algorithm with tradeoff function ψ\psiψ, and any allocation τ\tauτ. For 1≤i≤k−11\le i\le k-11≤i≤k−1 let

αi=∑q: type(q)=iOPT(q),βi=∑q: slab(q)=iALG(q),\alpha_i=\sum_{q:\ \mathrm{type}(q)=i}\mathrm{OPT}(q),\qquad \beta_i=\sum_{q:\ \mathrm{slab}(q)=i}\mathrm{ALG}(q),αi​=q: type(q)=i∑​OPT(q),βi​=q: slab(q)=i∑​ALG(q),

where OPT(q)\mathrm{OPT}(q)OPT(q) is the bid of the bidder τ\tauτ assigns qqq to, type(q)\mathrm{type}(q)type(q) is that bidder's type at the end of the run, and ALG(q)\mathrm{ALG}(q)ALG(q), slab(q)\mathrm{slab}(q)slab(q) are as in Lemma 6. Then

∑i=1k−1ψ(i)(αi−βi)≤Nk.\sum_{i=1}^{k-1}\psi(i)(\alpha_i-\beta_i)\le\frac Nk .i=1∑k−1​ψ(i)(αi​−βi​)≤kN​.

Applied to Theorem 8's ψk(i)=∑j=ik−1yj∗\psi_k(i)=\sum_{j=i}^{k-1}y^*_jψk​(i)=∑j=ik−1​yj∗​ (which satisfies these hypotheses), through Δ⋅y∗=∑i(αi−βi)ψk(i)\Delta\cdot y^*=\sum_i(\alpha_i-\beta_i)\psi_k(i)Δ⋅y∗=∑i​(αi​−βi​)ψk​(i) this is the bound on the error term in the proof of Theorem 8.

Formalization Note The two sums are the paper's opening observations in the proof of Lemma 7, taken here as the definitions of αi\alpha_iαi​ and βi\beta_iβi​. The paper states the lemma for a generic monotonically decreasing ψ:[1…k]→R+\psi:[1\dots k]\to\mathbb R^+ψ:[1…k]→R+; the added hypothesis ψ(k)=0\psi(k)=0ψ(k)=0 (true for Theorem 8's ψk\psi_kψk​ and for §3's ψk\psi_kψk​) replaces the paper's bound on the slab-kkk term, which rests on its simplifying assumption that bidders of type jjj spend exactly j/kj/kj/k: for a tradeoff function positive at slab kkk the slab-kkk money can exceed N/kN/kN/k by one bid per bidder, which the paper absorbs into its small-bid assumption.

Preamble
import Mathlib
import Definitions.Def_AdWordsMSVV_Tradeoff_Setting
Formal statement
namespace AdWordsMSVV.Tradeoff
theorem lemma_7 {N M : ℕ} (k : ℕ) (hk : 1 ≤ k) (ψ : ℕ → ℝ)
    (hψ : AntitoneOn ψ (Set.Icc 1 k)) (hψ0 : ∀ i ∈ Set.Icc 1 k, 0 ≤ ψ i) (hψk : ψ k = 0)
    (bid : Fin N → Fin M → ℝ) (hbid : ∀ b t, 0 ≤ bid b t)
    (σ : Fin M → Option (Fin N)) (hσ : IsRun ψ k bid σ)
    (τ : Fin M → Option (Fin N)) :
    let α : ℕ → ℝ := fun i =>
      ∑ t : Fin M, match τ t with
        | none => 0
        | some b => if typeOf k bid σ b = i then bid b t else 0
    let β : ℕ → ℝ := fun i =>
      ∑ t ∈ Finset.univ.filter (fun t : Fin M => slabQ k bid σ t = i), algRev bid σ t
    ∑ i ∈ Finset.Icc 1 (k - 1), ψ i * (α i - β i) ≤ (N : ℝ) / k := by sorry
end AdWordsMSVV.Tradeoff
Source
Mehta, Saberi, Vazirani, Vazirani, AdWords and generalized on-line matching, J. ACM (2007), DOI 10.1145/1284320.1284321, p. 11, Lemma 7 and the opening lines of its proof
Human review
  • Endorsed by Shuze Chen · Oct 7, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 7, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me