| paper | d4-journals-mss-2009-0016 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
Theorems 2 and 3 give existence characterizations, but neither states an algorithm, a complexity bound, or hardness. The proposed population versions require new type aggregation rules and cannot supply the missing computational result.
fails bit a — no named computational result to mirror
The proposed mirror concerns Theorems 2 and 3; it leaves the convexity, homotheticity, and monotonicity extension results as alternative-space results, along with the paper's general extension framework.
I cannot make a positive case that meets your anchor requirement: the paper has no named computational complexity result. Its named theorems characterize existence and rationalizability; they do not establish that a computational problem is in P, NP-hard, or otherwise give an algorithmic complexity result.
The closest anchors are Theorem 2, proved in the paper, which says that for an algebraic closure operator \(F\) satisfying C7, a relation \(R\) has a complete extension \(R'=F(R')\) if and only if \(F(R)\cap P^{-1}(R)=\varnothing\); and Theorem 3, also proved there, which says that, under its stated conditions on \(F\), a choice function \(K\) is \(F\)-rationalizable if and only if \(R_v\cap P_v^{-1}=\varnothing\). These are mathematical characterizations, not computational results: the paper gives no complexity analysis or algorithm for deciding the conditions.
A population of consumers with choice-behavior types and fractional masses could motivate a new population-level rationalizability question, but that would be an extrapolation, not a continuous version of a computational problem established by this paper. And the paper’s continuous alternative spaces in its convexity and homotheticity sections continuize the alternatives, not the population. So there is no defensible per-anchor continuous problem to hand a prover under the requested standard.
The proponent is right about the paper’s main weakness as a ChoCo candidate: it proves existence characterizations, not computational complexity results. Theorem 2 characterizes when a relation has a complete extension with a specified closure property; Theorem 3 characterizes when a choice function is \(F\)-rationalizable. Neither gives a running time, a hardness result, or an algorithmic problem for a population. That leaves no named computational result here to mirror.
Theorem 2 could be turned into a decision problem by asking whether a given finite relation has the required extension. But its input is one relation over alternatives, not a society. To supply a distribution over types, one must first decide what types are and how their relations combine into the input relation. If each type is checked separately, mass has no effect beyond support. If relations are combined by unanimity, majority, or another rule, that aggregation is an additional model choice absent from the paper. A majority relation derived from voter rankings could make a coherent continuous-population problem, but it would be an aggregation problem built around Theorem 2—not a computational result of this paper.
Theorem 3 offers the best route to a population interpretation: consumers could be types, each with a choice function, and their choices could be aggregated across menus. But the theorem’s \(K\) is one set-valued choice function, not a distribution of choices. Aggregating a population’s choices produces shares or probabilities; converting those into \(K\) requires a further rule, such as a threshold or support. Alternatively, requiring the same relation to rationalize each type makes the condition depend on which types occur, not their masses. The more natural fractional-choice version may be worth studying, but it changes the object whose rationalizability Theorem 3 characterizes.
The paper’s convex and homothetic sections do not rescue the mirror: they continuize the alternative space \(X\), while the programme’s object is a continuous population. The strongest negative case is therefore about fit, not impossibility. One can construct plausible projects nearby, especially population aggregation of preferences or choice behavior. I cannot honestly rule out that the authors would recognize one as worthwhile. But it requires adding the population model and its aggregation rule; it is not a continuous mirror of a computational question this paper establishes.
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.