| paper | d4-journals-geb-2021-0083 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
Theorem 1 compares equilibrium sets but states no algorithm, complexity bound, hardness result, or computationally used formulation. The numbered experimental findings provide no computational guarantee, and the proposed atomless model would change finite-player Nash incentives rather than transfer the theorem.
fails bit a — no named computational result to mirror
No computational result is covered: Theorem 1 is an equilibrium comparison, Results 1–6 are experimental findings, and the cited mechanism characterizations are axiomatic.
The most plausible continuous-population setting is a large school district: many students, but a relatively small number of preference and academic-priority types. Let a type record a student’s school ranking, utility for each school, and merit category, with \(\mu_t\) the fraction of students of type \(t\). Scale each school’s capacity by the cohort size, and reserve a fraction \(\lambda=1/3\) for lottery admission. Students choose reported rankings; a measurable strategy maps each type to a report. A continuum DA or DA33 procedure, using merit and lottery cutoffs, then gives each type an assignment lottery. One computational question would be to characterize the pure-strategy equilibria and their induced assignments, comparing DA with DA33. That is a recognizable extension of the paper’s incentive question.
The nearest named result is Theorem 1, proved in this paper. With priorities aligned across schools, and students comparing lotteries by either expected utility or stochastic dominance, it says: (a) every pure-strategy Nash equilibrium under DA33 is also an equilibrium under DA; and (b) some DA equilibria are not equilibria under DA33. Thus DA33 has a strict subset of DA’s equilibria. But this is an equilibrium-comparison theorem, not a result that a computational problem is in P, NP-hard, FPT, or otherwise classified. It is therefore not an eligible anchor under the requested standard. The paper contains no named computational result of that kind, so there is no valid anchor-indexed continuous problem to hand to a prover.
The scope of the candidate question would be the paper’s theoretical equilibrium comparison, not its experimental findings about truth-telling, payoffs, or school composition. Its weakest modeling point is that the paper uses strict individual priorities. Grouping many students into a few merit types may require ties or a new tie-breaking rule; retaining exact individual ranks could instead make the number of types grow with the population. The finite-agent equilibrium theorem also need not survive unchanged when one student’s deviation has zero effect on continuum assignment cutoffs. Thus school choice offers a plausible high-multiplicity regime, but the paper supplies no computational anchor to make the positive case strong on the terms set here.
The only anchor raised is Theorem 1, and the strongest objection is that it is not a computational result. It compares equilibrium sets for finite school-choice games; the paper’s other named results are axiomatic characterizations, and its remaining findings are experimental. None classifies the complexity of a problem or gives an algorithm. So the paper offers no computational result to mirror under ChoCo’s stated brief.
The best attempt to rescue the anchor would be to ask for an algorithm that computes or characterizes equilibria when a large district is represented by a distribution over preference and merit types, with capacities scaled to population and a lottery quota of \(\frac{1}{3}\). That is a plausible question. But it is a new computational problem about equilibrium in a continuum game, not a computational version of a problem this paper studies. Modeling it also forces substantive choices: finite merit categories create ties absent from the paper’s strict priorities; preserving distinct merit scores requires a continuum of types rather than finite support. And the proof of Theorem 1(a) uses a positive-probability event in which the finite lottery order matches the common priority order. That event has probability \(1/n!\), so its role disappears in the atomless limit. A continuum counterpart would need a new model and proof.
This case does not establish that school choice has no worthwhile continuous-population model. A large school district is a credible high-multiplicity setting, and a study of continuum-game equilibria could be worthwhile. The negative case is narrower: this paper supplies no computational anchor for that work, and the natural stronger proposal is an adjacent research question rather than a mirror of one of its computational results. The universal claim that no scenario could be worthwhile is therefore weak.
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.