| paper | d4-journals-geb-2018-0094 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The paper’s numbered Results 1–5 report experimental findings, and its Appendix A equilibrium characterizations do not state an algorithm, complexity bound, or hardness result. The proposed continuum equilibrium problem is a statable extension, but no computational result of this paper anchors it, so bit (a) fails.
fails bit a — no named computational result to mirror
The proposed question concerns continuum equilibrium reports and assignment masses; it does not carry over the empirical findings on cognitive ability, strategic mistakes, payoff differences, or school segregation in Results 1–5.
There is a plausible continuous population model here, but the paper has no eligible computational anchor under your rule. Its numbered Results 1–5 are empirical findings from a 16-student experiment; none is a numbered theorem, lemma, corollary, or proposition asserting a complexity or algorithmic result. For example, Result 4 reports that high-ability subjects earn more than low-ability subjects under Boston, with smaller payoff differences under Deferred Acceptance; Result 5 reports that low-ability subjects are over-represented at school D and under-represented at school C under Boston. Those are relevant phenomena, but they are not computational results. The equilibrium characterizations in the appendix are unnumbered and likewise make no complexity claim.
The strongest *unanchored* mirror would be a large municipal admissions cohort. A type records a student’s school-payoff vector, hence their true ranking; the population is a rational mass distribution \(\mu\) over finitely many such types. Each school has a capacity share \(q_s\), and priorities come from a uniform lottery over the population. Students choose reported rankings, represented by a distribution \(\sigma_t\) over reports for each type \(t\). The problem would be: given \((\mu,q,u)\) and Boston or Deferred Acceptance, find all equilibrium reporting distributions and their induced assignment masses \(x_{t,s}\), including type-conditional payoffs. The regime is plausible: a city may have many applicants but relatively few preference types. This asks about the underlying strategic school-choice problem, not about fractional outcomes as the point of the exercise.
The weak point is substantial: the paper studies a small finite game and estimates behavior experimentally, while the proposed mirror is a nonatomic equilibrium problem. Individual pivotality disappears, and cognitive ability does not affect rational best responses unless the model adds a behavioral assumption that the paper does not supply. So this is a credible direction for a new question, but I cannot honestly present it as a continuous mirror anchored in a named computational result of this paper or assign it a supported A/B/C classification.
The strongest case against a mirror is that this paper supplies no ChoCo-style computational result. Results 1–5 are empirical findings about behavior, payoffs, and school assignments. Appendix A characterizes equilibria in two fixed games with 16 students; it gives no algorithm or complexity claim. Treating Results 4 and 5 as computational anchors would confuse findings about what subjects did with results about the computational difficulty of a problem.
The proposed continuum game is a plausible new question. A large admissions cohort can sensibly be represented by masses of students with shared payoff vectors, school capacities scaled by population, and lottery priorities. One could ask for equilibrium reports and the resulting allocation masses. So the negative case cannot rest on there being no sensible high-multiplicity regime, or on capacity shares being an impermissible co-modification.
The difficulty is preserving what makes Results 4 and 5 this paper’s findings. Its central comparison concerns whether cognitive ability predicts strategic mistakes. In a standard equilibrium model, adding ability to a type does not by itself specify how ability changes choices: students with the same payoff vector face the same strategic incentives. To model the reported ability differences, one must add a theory of ability-dependent behavior, such as assumptions about learning, beliefs, or response errors. If those behavioral rules are supplied as inputs, the task becomes aggregating given behavior through the mechanisms. If they are to be solved for, the paper does not specify the model or equilibrium concept. Either route could motivate worthwhile research, but neither is a computational result this paper establishes.
That leaves a strong reason not to treat this paper as an anchored ChoCo case, and a weaker reason to reject its subject as a source of future computational questions. I cannot honestly defend the universal claim that no continuous scenario is worthwhile: equilibrium computation for a continuum school-choice population may be a good independent problem. The negative case is strongest at the programme’s entry gate: this paper offers no named computational result to mirror, and its most natural proposed mirror needs additional behavioral theory to retain the paper’s main contribution.
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.