| paper | d4-conf-aaai-1994-0183 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | medium |
| authors would recognise it | unclear |
Theorems 1 and 2 concern computational outcomes of the two-agent auction, so they provide plausible anchors. But the proponent states no continuous question for either; the opponent's possible typed mass-matching model is an extension of the inherited coalition-formation process, not a specified analogue of those theorems.
fails bit b-none — no population with multiplicity to relax
Theorems 1 and 2 cover manager selection and agreement price for a pair of agents. The paper's larger coalition-formation process, including its cited \(O(n^3)\) algorithm, is not established here as a continuous computational result.
This paper has no named computational-complexity result to anchor a continuous mirror. Theorem 1, proved here, says that in the two-agent auction the agent with the higher estimate of the coalition’s value becomes manager. Theorem 2, also proved here, says the agreement price is within \(\delta\) of a quantity determined by the initial offers and the non-manager’s valuation. Both are mechanism properties, not results about a problem being in P, NP-hard, or otherwise computationally classified.
The paper does mention an \(O(n^3)\) decentralized coalition-formation algorithm, but attributes it to Ketchpel (1993) and gives the claim in prose, not as a numbered result of this paper. A continuous model of mass flows among repeated agent types might be conceivable, but it would be an unanchored extension. Under the stated criteria, there is no qualifying anchor and therefore no continuous problem to specify per anchor.
The case for names no continuous anchor, and that is largely right. Theorem 1 and Theorem 2 are numbered results, but they settle pairwise mechanism questions: for a given pair, which agent manages, and how the agreement price relates to the offers and valuation estimates. Apply them to a distribution of pair types and they hold pair by pair. Aggregating their answers does not create a population-level computational problem.
The strongest computational candidate is the \(O(n^3)\) coalition-formation algorithm in Section 4.1. The paper attributes it to Ketchpel (1993); it is not a complexity result established here. A generous mirror would imagine repeated founders, investors, or contractors in a market, grouped by identical offers, preferences, and valuation estimates. One could study mass matching among these types, then apply the two-agent auction to each matched mass. That is a coherent continuous-population model, so claiming that no such scenario can exist would go too far.
But the computational question in that model would be the mass version of the matching dynamics. The paper’s two theorems contribute no new aggregate result to it: they apply separately to each pair after the matching is chosen. If partner values depend on named individuals or coalition histories, those details must be included in the types and can defeat the high-multiplicity regime. If values are homogeneous by type, the matching model can be specified cleanly, but it is an extension of the paper’s inherited algorithm rather than a continuous mirror of its original contribution.
So the strongest case against is that this paper offers little for ChoCo to continuize: its original results concern a two-agent negotiation, and its only stated runtime belongs to prior coalition-formation work. That is a thin target, not proof that no worthwhile continuous model can be built. Someone could find a valuable typed mass-matching problem nearby; the paper itself gives little reason to count it as a mirror of its results.
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.