| paper | d4-journals-scw-2001-0024 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
Propositions 1–3 characterize freedom rankings through axioms; none states an algorithmic or complexity result. The proposed comparison over a distribution of rankings is a new population interpretation, and its support-based count does not turn any proposition into a computational result of this paper.
fails bit a — no named computational result to mirror
No proposition is covered by a computational mirror; the proposed support-based comparison is an extension and leaves all three axiomatic characterizations untouched.
The paper has no named computational result of the kind required for an anchor. Its numbered results are Propositions 1–3, proved in the paper, and each is an axiomatic characterization. Proposition 1 says that \(\succeq\) satisfies SND, COM, and IMON if and only if \(A\succeq^*B\) exactly when \(|\max(A)|\ge|\max(B)|\). Propositions 2 and 3 characterize two lexicographic variants. None asserts that an underlying problem is in P, hard, FPT, or another complexity class.
The closest positive foothold would be a new population version: let each type be a complete preference ranking \(R\), and let \(\mu_R\) be the mass of people with that ranking. For a feasible set \(A\), define \(S_\mu(A)\) as the alternatives in \(A\) that are top-ranked over \(X\) by some positive-mass type. Given \(X,A,B\) and a rational distribution \(\mu\) supported on \(\tau\) rankings, ask whether \(|S_\mu(A)|\ge|S_\mu(B)|\). A solution is the yes/no comparison. This preserves the paper’s count of distinct meaningful alternatives, but the masses matter only through whether they are positive; the question reduces to the support of \(\mu\) and is easy to compute.
That is a possible extension, not a defensible mirror of a named computational result. The paper treats reasonable preferences as a normative set of possibilities, not as a population whose members have measurable shares. Its results are axiomatic, and continuizing those axioms would be outside ChoCo’s computational scope. So there is no eligible anchor and no per-anchor continuous problem to hand to a prover. The weak point is decisive: the proposed population interpretation adds a model the authors do not study, while the paper supplies no computational result for it to mirror.
The proponent is right about the main obstacle: Propositions 1–3 are axiomatic characterizations, not computational results. The paper gives no algorithmic problem, input structure, or complexity claim to continuize. The closest direct question is their comparison of \(S_\mu(A)\) and \(S_\mu(B)\), where \(S_\mu(A)\) contains alternatives in \(A\) ranked best over \(X\) by at least one positive-mass type.
That question puts the wrong meaning on \(\mu\). In the paper, the reference orderings describe what a reasonable person might prefer; they are possibilities, not a population whose preferences have frequencies. Giving them masses changes that normative reference set into an empirical society. And under the paper’s cardinality rule, those masses then contribute nothing except whether they are positive. If \(X=\{a,b\}\), give \(a\succ b\) mass \(1-\varepsilon\) and \(b\succ a\) mass \(\varepsilon\). For every \(\varepsilon>0\), \(b\) counts as relevant; at \(\varepsilon=0\), it disappears. A group whose share tends to zero retains its full effect until the limit, then loses it all. The direct mirror therefore treats the society as a support list, not as a continuous aggregate.
The stronger redesign would make freedom sensitive to mass—for example, measure the share of people whose feasible set contains a personally meaningful option. But that replaces the paper’s count of distinct relevant alternatives with a population-level welfare measure. Adding constraints and costs to choose among feasible sets could yield a computational optimization problem, but the paper supplies none of that structure. Those may be worthwhile questions inspired by its topic; they are not continuous mirrors of its results.
The same issue disposes of all three propositions as candidate anchors. Proposition 1 uses the number of relevant alternatives; Propositions 2 and 3 compare their counts across successive levels. Replacing their reference set with population support makes each rule depend on which types are present, not how much mass they carry. Making them mass-sensitive requires changing the rules’ meaning. Neither route yields a computational problem from the paper that continuization can carry forward.
The negative case is strongest for this article, not for every possible project one could build around opportunity sets. A separate menu-design problem might be valuable, but it would need new decision variables and constraints before there is anything here to continuize.
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.