| paper | d4-conf-aaai-2007-0100 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
The proposed repeated-stakeholder model is plausible, but neither named result has computational character. Asking how to compute the maxmin outcome would add a computational problem the paper does not formulate or analyze, so no continuous computational mirror of these results survives.
fails bit a — no named computational result to mirror
The proposed extension concerns Theorem 1's modest maxmin solution; Proposition 1 derives weak Pareto optimality and balanced contraction. The paper's logical definitions and numerical mapping do not add a computational result.
The strongest plausible mirror is a mass population of stakeholders in a large negotiation: a type \(t\) is a complete belief state \((\rho_t,e_t)\), and \(\mu_t\) is the fraction of stakeholders with that state. The decision is a deal \(D_t\) for each type, with each deal comprehensive and contained in that type’s negotiation set, and with the union of all deals consistent. One could ask for the paper’s modest maxmin solution: maximize the minimum satisfaction \(\frac{\rho_t(\top)-\rho_t(D_t)}{\rho_t(\top)-e_t}\), then return, for each type, the intersection of its deals across all maximizing outcomes. That is a faithful extension of the paper’s bargaining model to many repeated stakeholders.
But the paper supplies no qualifying computational anchor for that question. Theorem 1, proved here, characterizes the modest maxmin solution by inclusion, scale invariance, symmetry, and mutually comparable monotonicity. Proposition 1 derives weak Pareto optimality and balanced contraction. These are axiomatic results, not claims about an algorithm, complexity class, or computed formulation. So there is no named computational result to anchor a continuous question to.
The main weakness of the candidate mirror is that the original maxmin objective makes the masses irrelevant beyond identifying which types are present: repeating a stakeholder type does not change the solution. It is a plausible high-multiplicity version of the paper’s bargaining problem, but I would not present it as an anchored continuous-complexity case.
The proponent’s stakeholder scenario is a plausible high-multiplicity regime. The difficulty is that the paper gives no computational result to mirror. Theorem 1 characterizes the modest maxmin solution through four axioms; Proposition 1 derives weak Pareto optimality and balanced contraction from those axioms. Neither states an algorithm, a complexity bound, or a computational formulation. A continuous version of either result would still be an axiomatic question, outside ChoCo’s scope.
The proposed mass model also exposes a deeper mismatch. If repeated stakeholders have identical complete belief states and receive identical treatment, the paper’s maxmin objective takes the minimum satisfaction over players. In a finite-support population, that becomes the minimum over types with positive mass: changing their mass values does nothing. To make mass affect the objective—for example, by using a mass-weighted aggregate—you would change the solution concept, not continuize Theorem 1.
One could ask how to compute the modest maxmin outcome for a large population. But that would be a newly posed computational problem, not a mirror of a computational claim in the paper. As written, the paper does not specify a finite encoding for its real-valued entrenchment functions or analyze the cost of checking logical consistency. Making that question precise would require additional modeling choices. The same applies to Proposition 1: computing whether an outcome satisfies WPO or BC would add a computational task that the proposition itself does not study.
So the strongest case against a ChoCo mirror is not that repeated stakeholders are implausible. It is that the paper’s contributions are axiomatic, and the most faithful mass extension leaves the masses out of the maxmin calculation. This is strong evidence against a worthwhile *mirror of this paper*. It is not proof that no follow-on computational problem could be interesting; such a problem would need its own motivation and formalization.
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.