| paper | d4-conf-atal-2022-0129 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Theorem 4 gives a tight welfare bound for equilibria, but does not state an algorithm, computational hardness, or a formulation used to compute. The proposed population limit is coherent as an analytic extension, yet it cannot supply the missing computational result.
fails bit a — no named computational result to mirror
The proposed extension concerns only Theorem 4’s multi-unit CCE-POA bound with overbidding; it leaves Theorems 3 and 5–9 and Proposition 1 untouched.
I do not see an eligible ChoCo anchor in this paper. Its named results—Theorems 3–9 and Proposition 1—give price-of-anarchy bounds, equalities, or a model equivalence. None asserts that a computational problem is in a complexity class or gives an algorithmic result of the kind the programme is charting. The paper studies welfare guarantees at equilibria, analytically; its bid distributions and correlated equilibria do not themselves make the population continuous.
The strongest nearby result is Theorem 4, proved in this paper: for every \(\gamma\in(0,1]\), the coarse-correlated price of anarchy of any multi-unit \(\gamma\)-FPA when bidders may overbid is at most
\[
\frac{1}{\gamma(1-e^{-1/\gamma})}.
\]
The bound is tight, with tightness already attained by a single-item \(\gamma\)-HYA. This is a substantial result, but it is not a computational-complexity anchor.
If an explicitly analytic population-limit question were admitted, I would propose this one. Let \(T\) be a finite set of bidder types, where each type \(t\) specifies a complete submodular valuation function \(v_t\); let \(\mu_t\) be its population share. For each \(N\), form an auction with \(n_t(N)\) bidders of type \(t\), with \(n_t(N)/N\to\mu_t\), and \(K_N=\lfloor\kappa N\rfloor\) identical units, for fixed \(\kappa>0\). Run the \(\gamma\)-hybrid auction, allowing overbidding. The question is: for this type distribution and supply ratio, what is the \(\limsup\) of the finite auctions’ CCE price of anarchy as \(N\to\infty\), and when does it attain Theorem 4’s bound? A solution would characterize that limit and establish matching upper and lower bounds. The regime is plausible for large procurement pools with many suppliers sharing a small set of marginal-value schedules, and the authors could recognize it as a high-multiplicity extension of their multi-unit welfare question.
That candidate is a population continuization, but it remains an analytic equilibrium question, so it has no honest Class A/B/C prediction in ChoCo’s computational trichotomy. Its weakest point is that the tightness example in Theorem 4 uses two bidders; it does not establish tightness when both bidders and supply grow proportionally. I would offer this as an adjacent research question, not as evidence that the paper contains a computational result for ChoCo to mirror.
The strongest case against is that this paper has no ChoCo computational anchor. Proposition 1 is a model equivalence, and Theorems 3–9 give price-of-anarchy bounds and equalities. None states the complexity of a computational problem or gives an algorithm. The paper’s results take valuation profiles and equilibria as their objects and establish welfare ratios; making bidder populations continuous would not turn those results into complexity results.
The proposed Theorem 4 limit question is coherent. A large procurement pool with repeated valuation types is plausible, and the question is not defeated by grouping bidders or by whatever answer it gets. But it asks for an analytic limit and tightness result. The theorem already gives the proposed upper bound for every finite auction in the sequence, so the main new work would be to establish when that bound is attained as bidder count and supply grow. That could be worthwhile population-limit analysis, but it does not mirror a computational result in this paper.
A more computational variant—given type masses and a supply ratio, compute the worst CCE price of anarchy—would add a new, instance-sensitive equilibrium-optimization problem. The paper’s theorem instead bounds the worst case over all valuation profiles; it neither formulates that input-output problem nor develops an algorithm for it. Such a project might be worth pursuing, but it would be a new direction, not a continuous mirror of the paper’s contribution.
So the negative case is strong on fit with ChoCo’s computational programme, but weak as the universal claim requested. The proposed large-market limit is a plausible continuous question; the paper simply gives us no computational result to continuize.
The adversarial triple: the proponent anchors on up to three named results; the opponent sees that case and must defeat every anchor; the judge decides which case convinced it. These are the pipeline’s own outputs, generated by tools/triple_run.py — no human edited them. The paper’s own text is not reproduced here beyond the quoted statement above.