| paper | d4-conf-aaai-2013-0309 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
Theorem 3 bounds the quality of strong-equilibrium outcomes but gives no algorithm, complexity claim, or result about computing an equilibrium. The proposed mass game is a possible game-theoretic extension, but its equilibrium analysis is not a computational result of this paper. No named anchor meets bit (a).
fails bit a — no named computational result to mirror
The proposed extension concerns only Theorem 3's MAGNET CCC bound on arbitrary networks; it leaves the gold-rush and CCC results, and the complete-graph and line-network bounds, unaddressed.
The paper contains no named computational-complexity result of the kind required for an anchor. Its numbered results concern equilibrium existence and price-of-anarchy bounds; they do not classify a problem as NP-hard, polynomial-time solvable, FPT, or similar. So there is no compliant computational anchor here.
The strongest candidate for a continuous analogue, if the anchor requirement were broadened, is Theorem 3, proved in this paper: in the MAGNET CCC game on an arbitrary collaboration network, the strong price of anarchy is at most \(2\). The proof compares the winning consortium with an optimal eligible consortium using their overlap, sizes, and average qualities.
A plausible mirror would use a large research workforce with finitely many types, each defined by quality and collaboration block. Type \(t\) has rational mass \(\mu_t\); links are specified between blocks, with agents in a block interchangeable. A consortium is a vector of selected masses \(x_t\), required to have connected support and to meet a per-capita threshold \(\sum_t q_t x_t \ge \theta\sum_t x_t\). The MAGNET protocol runs on these masses: types may split across proposals, and appeals add mass when the resulting consortium meets the paper’s connectivity, consistency, and improvement conditions. Members of a winning consortium share the fixed grant equally, so a member’s payoff is inversely proportional to its mass.
The resulting question could be handed to a prover: given rational type masses and qualities, a block collaboration graph, a threshold, and a grant, characterize the strong-equilibrium outcomes of this continuum MAGNET game and determine the worst equilibrium’s average quality relative to the best eligible consortium. The natural conjecture is that the bound in Theorem 3 survives when the paper’s overlap and average-quality arguments are expressed in masses. That would be an equilibrium-quality analysis, though—not a computational-complexity result, and not a result established by this paper.
The high-multiplicity scenario is plausible: a large grant programme with recurring cohorts of researchers who share quality and collaboration access. The main weakness is that the paper’s arbitrary individual network does not generally collapse to a small set of interchangeable blocks. Restricting to block networks makes the continuum model coherent, but narrows the setting; moreover, equilibrium existence in the continuum would need separate treatment, since the paper’s finite descent argument does not directly carry over.
The decisive problem is that this paper has no computational-complexity result to continuize. Theorem 3 is a strong price-of-anarchy bound: every strong-equilibrium outcome has average quality within a factor of \(2\) of the optimum. Lemma 4 establishes equilibrium existence. Neither gives an algorithm, a running time, or a complexity classification. The other named results are also equilibrium-quality bounds. So the proposed Theorem 3 mirror is an equilibrium-analysis question, not the computational question this programme is set up to pursue.
A large workforce with a finite set of quality and collaboration types is a plausible high-multiplicity scenario. A block network could preserve connectivity while letting types split across proposals. But the proponent’s eligibility condition, \(\sum_t q_t x_t \ge \theta\sum_t x_t\), changes the paper’s model: the paper requires a *total* quality threshold, not an average-quality threshold within each consortium. If population size is \(n\), values are \(q_t\), and the threshold scales as \(T_n\), the faithful mass condition is \(\sum_t q_t x_t \ge T_n/n\), while average quality remains the objective. Correcting this makes the mirror more faithful, but it still asks for a price-of-anarchy theorem.
The continuous game would also need its own equilibrium model. Individuals have zero mass, so unilateral deviations cannot change a consortium; strong deviations would need to be defined for measurable groups of positive mass. And the paper’s equilibrium-existence proof uses finite descent, which does not directly establish termination or existence when mass can be split arbitrarily. Those are real modelling and proof obligations, but they do not rule out a worthwhile atomless-game extension.
That distinction sets the honest limit of the negative case. One could make a worthwhile continuous game-theoretic mirror of Theorem 3. But proving its equilibrium-quality bound would not fill ChoCo’s stated gap in computational questions and algorithms. One could instead ask how hard it is to compute a worst equilibrium in the new game; that may be a computational project, but it is a new question, not a computational result of this paper being mirrored. So the case against is strong as a verdict about ChoCo’s scope, but weak as the universal claim that no worthwhile continuous mirror exists at all.
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.