| paper | d4-journals-geb-2024-0134 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Proposition 1 gives an asymptotic formula for expected welfare gains, Corollaries 1–3 give comparative statics, and Lemma 1 is an existence result. None states an algorithm, computational guarantee, hardness result, or solver formulation used to compute, so the opponent’s plausible mass-matching task cannot supply the missing named computational result.
fails bit a — no named computational result to mirror
The proposed mirror follows Proposition 1’s expected rank-gain quantity and its district comparisons; it does not cover Lemma 1, the empirical preference-estimation method, or the paper’s specific estimates.
The strongest honest case is that this paper has no eligible named computational anchor, so there is no anchor-based continuous problem to propose.
The nearest result is Proposition 1, proved in the paper’s Appendix A. In a random school-choice market with uniformly random preferences and priorities, it approximates district \(i\)’s expected average-rank gain from consolidation by
\(\displaystyle \gamma_i \approx \frac{N+K}{q}\left[ \frac{\log\!\left(\frac{n_i+k_i}{k_i}\right)+(q-1)\log\log\!\left(\frac{n_i+k_i}{k_i}\right)}{n_i} -\frac{\log\!\left(\frac{N+K}{K}\right)+(q-1)\log\log\!\left(\frac{N+K}{K}\right)}{N} \right], \)
where district \(i\) has \(q n_i\) students and \(q(n_i+k_i)\) seats, \(N=\sum_i n_i\), and \(K=\sum_i k_i\). Corollaries 1–3 state that average gains are positive for every district under the paper’s mild conditions, more under-demanded districts gain less, and smaller districts gain more. These are theoretical welfare and comparative-static results, not claims that a computational problem is in P, hard, or belongs to a parameterized class. Lemma 1 establishes that certain latent preference and priority valuations can be chosen consistently with stability and observed rankings; it also gives no computational classification. The paper’s deferred-acceptance calculations and Gibbs sampler do not supply a numbered complexity result either.
A high-multiplicity school market is plausible: a large urban cohort could contain many students sharing complete school rankings and school-priority categories, with seats scaled to the cohort. That makes a continuous population model worth studying. But the paper’s strict individual priorities and student-specific latent utilities make repeated complete types questionable in its own empirical setting; grouping students into priority classes would require a model extension.
So the paper offers a promising setting, but zero eligible anchors and zero continuous problems per anchor. There is no defensible Class A, B, or C expectation to attach to a mirror of its named results. The weakest point is also decisive: the positive case would have to introduce a new computational question instead of continuizing a computational result this paper actually proves.
The proponent is right about the narrow point: the paper has no numbered theorem classifying a problem’s complexity. Proposition 1 still gives a strong anchor for a computational mirror, because it names the quantity to compute: each district’s average rank gain from consolidation. Its formula and Corollaries 1–3 cover random markets; the paper also computes these gains for empirical and simulated markets with heterogeneous preferences.
A continuous version can take a rational mass distribution \(\mu_{i,t}\) over complete student types in each district \(i\). A type records the student’s school ranking and the priority class relevant at each school; schools have capacities measured as population shares. The computational question is to find the student-optimal stable mass matching before and after consolidation, then compute
\[ G_i(\mu)=\frac{1}{\mu_i}\sum_t \mu_{i,t}\bigl(r_t(\nu^i(t))-r_t(\nu^\Omega(t))\bigr), \]
where \(\nu^i\) is the district matching, \(\nu^\Omega\) the consolidated matching, and \(r_t\) the rank in type \(t\)’s preferences. One can ask for exact or approximate algorithms, parameterized by the number of schools, type support \(\tau\), districts, and input bit length \(L\). This retains the paper’s district comparison and rank-welfare measure while making the population the continuous object.
That regime is plausible for large admissions cohorts with recurring preference rankings and priority categories based on exam bands, catchment, or other school criteria. The type records everything the assignment uses, so aggregating identical types does not discard identity-relevant information. Proportional seat capacities are a natural co-modification when population size is normalized. And the paper’s use of large-market analysis is relevant machinery, not an existing computational classification for arbitrary type distributions.
Lemma 1 need not carry a separate mirror; it establishes feasibility of latent valuations, not the main welfare question. The paper’s deferred-acceptance calculations likewise do not classify complexity, which limits how literally this can be called a theorem mirror. Still, that gap does not remove the paper’s central computational object: comparing student-optimal assignments and their welfare under two district regimes. A further policy version could ask which feasible district mergers maximize aggregate or group-specific gains. The strongest caveat is that the paper does not formulate either task as a complexity problem, so this is a new computational articulation of its named welfare result rather than a restatement of a complexity theorem.
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.