| paper | d4-journals-jet-2015-0047 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
Theorem 1 gives an equilibrium-existence characterization, but no named computational result to mirror. The proposed atomless jury also removes juror-level pivotality, while a finite-committee limit remains a different analytical game-theory question.
fails bit a — no named computational result to mirror
In the atomless model, an individual report moves no mass, and truthful signal frequencies identify the state; the pivotality and conflict mechanism behind Theorem 1 disappears.
fatal: True
The proposed mirror targets Theorem 1's TS-equilibrium existence and conflict result; the paper's other equilibrium analysis is not covered, and none of it supplies a named computational result.
The closest named anchor is Theorem 1, proved in this paper. It gives an exact condition for the truthful-communication, sincere-voting equilibrium (TS equilibrium) under a non-unanimous voting rule and the no-partisans assumption. If hawks have critical mass, TS exists exactly when \(q_D\le\hat q_D(q_H)\), where \(\hat q_D\) is the conditional, probability-weighted average posterior guilt across profiles where a dove’s truthful guilty-signal report triggers conviction but an innocent-signal report triggers acquittal. If doves have critical mass, the corresponding condition is \(q_H\ge\hat q_H(q_D)\), with the threshold averaged over the analogous pivotal profiles for a hawk. In either case, at equality the theorem guarantees at least one profile where hawks and doves disagree. This is a substantial equilibrium-existence result, but it does not classify a problem’s computational complexity or give an algorithmic result. The paper has no named computational result of the kind required for a formal anchor.
The best positive lead is a large forecasting or risk committee: many analysts with one of a few error-cost cutoffs and one of finitely many signal types. A candidate mirror, Atomless TS Existence, would take a finite state and signal space, the paper’s rational prior and signal probabilities, a population measure \(\nu\) over preference cutoffs \(q\), and a conviction quota \(\kappa\). In state \(\omega\), the mass of analysts with preference \(q\) and signal \(s\) is \(\nu(q)P(s\mid\omega)\). Analysts publicly report a signal and vote; conviction occurs when the mass voting to convict reaches \(\kappa\). The question is whether truthful reports and sincere votes form a perfect Bayesian equilibrium with positive-probability ex post disagreement among preference types. A solution would give an equilibrium assessment and establish whether such disagreement occurs.
I would expect this atomless question to be tractable, but largely trivial. With truthful reports, the exact mass distribution of signals identifies the state: the paper’s distinct variants have distinct signal distributions. The posterior is then zero or one, so all preference types strictly prefer the same action. And an individual of zero mass cannot change the aggregate reports or vote. Truthful communication is therefore weakly incentive-compatible, while the theorem’s interesting cutoff and conflict result disappear.
That is the weakest point in the positive case: atomlessness removes the pivotality that drives Theorem 1. A study of how its cutoff \(\hat q\) behaves as finite committees grow could be a worthwhile asymptotic project, especially in a large committee with a small number of types. The theorem gives that project a clear starting point. But the atomless mirror above does not preserve the paper’s central strategic effect, and the paper supplies no named computational anchor for it.
Theorem 1 is the only plausible anchor, and it is an equilibrium characterization, not a computational result. Its condition averages posterior beliefs over the finite signal profiles at which a juror’s report can change the outcome. The paper poses no algorithmic problem, gives no complexity classification, and studies neither manipulation nor optimization. One can invent a problem about computing \(\hat q\), but that would algorithmize the paper’s finite-game formula; it would not, by itself, give the theorem a continuous computational mirror.
The obvious high-multiplicity setting is a large jury with many jurors in each preference group. That is a sensible setting, so multiplicity itself is no objection. But the atomless version loses the mechanism Theorem 1 studies. Under truthful reporting, the exact masses of the signal types reveal the state: conditional on each variant, the signal distribution is distinct, and with infinitely many independent signals the observed distribution identifies that variant. The posterior is then \(0\) or \(1\), so every juror with \(q\in(0,1)\) agrees on the action. And a single juror’s report changes no mass, so it cannot shift anyone’s aggregate evidence. Truth-telling is weakly compatible with equilibrium for this uninteresting reason; the theorem’s positive-probability conflict disappears.
A stronger proposal would keep finite juries, scale their sizes, and study the limit of the theorem’s cutoff as preference-group proportions converge. That preserves individual deviations at every finite size and may be a worthwhile asymptotic game-theory project. But it does not rescue an atomless mirror of the strategic result: under truthful play, conflict profiles become negligible as the signal frequencies concentrate on the true state. The incentives may still depend on rare pivotal profiles, so a nontrivial limiting cutoff is possible. Studying that cutoff means analyzing a sequence of finite Bayesian games, potentially conditioning on pivotal events whose probability vanishes. It is not the mass-transfer or continuum-population computation that motivates ChoCo.
One could instead put a continuum of identical finite committees in a population. That preserves each committee’s pivotal jurors, but the continuum merely averages outcomes across committees; it does not change or simplify the equilibrium question inside one committee. Other ways to preserve the strategic effect—giving speakers positive atoms, or allowing coordinated mass deviations—change the individual-deviation game that underlies Theorem 1.
So the negative case is strongest as a scope judgment: the paper offers a plausible large-committee asymptotic question, but its central result relies on finite-agent pivotality, and its natural continuum limit removes that force. The case is weakest against broader economic theory: the limiting behavior of the rare pivotal profiles could be a good analytical project. I would not present that possibility as a computational mirror without a separate, concrete algorithmic problem.
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.