| paper | d4-journals-jet-2023-0131 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The numbered results establish axiomatic impossibilities, mechanism properties, a characterization, and bounds on unanimity; none asserts an algorithm, computational complexity, or a computational guarantee. The proposed continuous mechanism frontier needs additional structure for agents’ incentives, so it is a new problem rather than a computational mirror supplied by this paper.
fails bit a — no named computational result to mirror
No computational result is mirrored. The paper’s axiomatic impossibilities, RDcED characterization, and unanimity bounds remain outside the mirror.
This paper contains no named computational result of the required kind. Theorems 1–6 and Propositions 1–2 establish axiomatic impossibilities, mechanism properties, a characterization, and a bound on \(q\)-unanimity. None classifies a problem as tractable, NP-hard, fixed-parameter tractable, or similar.
So there is no eligible anchor, and I can’t state a continuous computational problem per anchor without stretching the paper’s results beyond what they prove. A continuum model of random assignment may be worth studying, but this paper does not provide a named computational result to mirror.
The appended “case for” offers no affirmative mirror to rebut: it says the paper has no named computational result. That is the strongest case against a *paper-specific* ChoCo mirror. Theorems 1–6 and Propositions 1–2 concern axiomatic compatibility, mechanism characterization, and bounds on \(q\); none asks for an algorithm or classifies computational complexity.
The most tempting salvage is to turn Theorem 6’s frontier into a computational task: given a population distribution over preference types, find the largest \(q\) achievable by a strategy-proof, envy-free mechanism. But that requires a new model. A continuum society needs a way to represent object capacities and each type’s individual lottery, then to define how a single agent’s report affects those objects. If the mechanism sees only aggregate masses, individual incentive constraints may lose their finite-market meaning; if it retains individual-level response rules, those are additional structure the paper does not provide. The resulting mechanism-design problem could be worthwhile, but it would be a new problem inspired by the paper, not a continuous mirror of its result.
The natural replica limit also weakens the most concrete candidate. Under Theorem 2’s ex-post weak non-wastefulness condition, the sharp guarantee is \(2/n\), which tends to zero as the market grows. That is a genuine degeneration of the paper’s headline quantity. Allowing waste avoids that particular bound, but Theorem 6 gives \(17/18\) only for three agents, while Theorem 5 supplies no limiting rate or continuous characterization. Taking a limit of those finite-market questions would need its own analysis; it does not expose a computational landscape already latent in the results.
So the negative case is strong against calling any of these results a continuous computational mirror. It is weaker against the absolute claim that *no* worthwhile continuous question could be inspired by the paper: optimizing a mechanism frontier in a carefully specified replica model remains conceivable. But that possibility is a proposal for new work, not an anchor supplied by this paper.
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.