| paper | d4-conf-aaai-2026-0327 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The named theorems establish equilibrium comparisons and how often they occur in random markets; none asserts an algorithm, a running-time bound, or computational hardness. A continuous-population version could motivate analytic equilibrium questions, but no computational result from this paper anchors one.
fails bit a — no named computational result to mirror
No results are covered by a computational mirror; the paper's equilibrium-comparison, abundance, and mechanism-choice results remain outside one.
I don’t see an honest positive case that meets the required anchor rule. The paper has no numbered result classifying a computational problem as tractable, hard, or parameterized, and no algorithmic running-time result.
The closest foothold is Theorem 3.1, proved in this paper. It says that for every \(p\in(0,1)\), some \(\tau>0\) works for all sufficiently large \(n\): in the uniform \((n;3)\) model, with one district using BM and the other any mechanism, the expected market contains at least \(\tau n\) sophisticated students, each of whom strictly prefers a distinct sincere student to become sophisticated and weakly prefers all other sincere students to do so. That is a large-market equilibrium comparative-statics result, not a complexity result. Theorem 4.1, also proved here, gives only an expected constant number of students preferring DA to BM. Theorem 5.1 and Theorem 6.3 give further abundance results about constraint changes and positional heuristics. None is an eligible computational anchor.
Theorem 3.1 makes mass normalization tempting, but a faithful mirror runs into two problems. A type must include every feature the matching problem uses. With the paper’s strict, randomly drawn school priorities, students generally have distinct priority profiles, so the number of complete types can grow with \(n\). Coarsening priorities into tiers would change the model. And changing one student’s sophistication has zero mass in a nonatomic society; changing a positive mass of a type instead models a bloc or coalition and is a different perturbation.
So there are zero eligible anchors, and I can’t responsibly give one continuous computational problem per anchor or assign the paper’s results to Classes A, B, or C. A continuum analysis of equilibrium comparative statics could be a follow-on direction, but it would be an analytic extension rather than the computational mirror this programme asks for. The strongest point for that direction is Theorem 3.1’s linear abundance result; its weakest point is that neither the theorem’s computational content nor its individual-level perturbation survives the move to a high-multiplicity continuum.
The strongest case against a ChoCo mirror is that this paper supplies no computational result to mirror. Theorem 3.1, Theorem 4.1, Theorem 5.1, Theorem 6.3, and Proposition 6.1 establish equilibrium comparisons or show how often certain comparisons arise. None classifies a computational problem, gives an algorithm, or proves a complexity bound. So there is no result here to place in Classes A, B, or C.
The best candidate for a continuous question is a population shift from sincere to sophisticated students, inspired by Theorem 3.1. But the theorem compares the payoff of one identified student when another identified student changes type. In an atomless society, that student has zero mass. Converting a positive mass instead would ask how a bloc-wide change affects equilibrium, which is a different intervention; the theorem establishes no result about it. Preserving the paper’s individual comparisons also means preserving its strict priority orderings over named students. A complete type would then encode a student’s preference list and position at every school, with the type space changing as the market grows. Priority classes or score distributions could support a new continuum model, but they would need their own equilibrium formulation and would not follow from this paper’s result.
The other named results have the same gap. Theorem 4.1 proves a constant-size lower bound on students who prefer DA to BM, not a positive population share; its exhibited effect has no established positive-mass limit. Theorem 5.1 does establish linear abundance, but its result concerns individual students gaining from another student’s constraint change. Turning that into a mass intervention would ask about aggregate eligibility changes and capacity flows. Theorem 6.3 extends the sophistication comparison to positional heuristics, while keeping the same individual-level structure. These results may motivate new continuum questions, but none supplies a computational problem or a continuous-computation target.
The paper therefore looks like a poor *mirror* candidate for this programme: its contributions are qualitative equilibrium comparative statics and random-market frequency results, with strategic incentives tied to named agents and strict priorities. A continuum model of strategic school choice could still be worth studying in its own right. The negative case is weaker on the user’s universal claim: the paper does not rule out a sensible high-multiplicity regime or a useful computational extension. It does, however, give no eligible anchor for one, and the proposed aggregate versions would require choosing a new problem rather than continuizing a result the paper actually proves.
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.