| paper | d4-journals-geb-2015-0009 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
Neither anchor asserts an algorithm, a complexity bound, or hardness, so the paper supplies no named computational result to mirror. The proposed maximum-margin LP is a well-posed auxiliary problem, but it selects an aggregation policy rather than computing or characterizing the paper’s equilibria. In the atomless model, an individual voter has no pivotal influence, so that LP does not preserve the paper’s strategic question.
fails bit a — no named computational result to mirror
The proposed LP concerns the strict-separation condition in Proposition 3 and the signal-structure condition in Corollary 3; it does not cover the paper’s equilibrium comparison, Proposition 1, Example 1, or the premise-based result in Proposition 2.
The strongest positive case is an outcome-based, statewise mass-aggregation problem for a very large electorate with a fixed finite set of possible signals. A plausible regime is a mass public vote on a fixed set of allegations: voters receive independent noisy assessments, share the same objective and voting standard, and the number of voters is much larger than the number of signal types. A type is a voter’s signal \(s\), together with the common preferences and other attributes relevant to the problem. In state \(\omega\), the mass of type \(s\) is \(p_\omega(s)\). Each voter still casts a binary ballot; the continuous object is the fraction of each signal type voting yes.
There is a significant limitation to the case: the paper contains no named computational result asserting that a problem is in P, NP-hard, FPT, or another complexity class. Its numbered results concern strategic equilibria and information aggregation. So none can honestly serve as the computational anchor your instructions require.
The closest result is Proposition 3, proved in this paper. It says that for fixed \(\theta\in[1/2,1)\), if there is a vector \(x\in\mathbb{R}^{\Omega}\) such that \(p_\omega\cdot x>0\) for every state with \(f(\omega)=1\), and \(p_\omega\cdot x<0\) for every state with \(f(\omega)=0\), then there is a sequence of symmetric BNEs in the outcome-based game whose probability of error in each state goes to zero as the number of individuals grows. Corollary 3, also proved here, adds that linearly independent signal-distribution vectors \(p_\omega\) guarantee such a sequence; in particular, asymptotic efficiency holds for almost all signal structures. These are useful anchors for a *prospective* mirror, but they are not computational anchors under your rule.
That mirror can be stated precisely as Strict Mass-Signal Separation. The input is an explicitly listed finite state set \(\Omega\), signal set \(S\), decision rule \(f:\Omega\to\{0,1\}\), rational threshold \(\theta\in(1/2,1)\), and rational conditional masses \(p_\omega(s)\). Find a single signal-contingent voting rule \(x_s\in[0,1]\)—the probability that a voter with signal \(s\) votes yes—and a margin \(\delta>0\) such that
\[ \sum_{s\in S}p_\omega(s)x_s\ge\theta+\delta \quad\text{when } f(\omega)=1, \]
and
\[ \sum_{s\in S}p_\omega(s)x_s\le\theta-\delta \quad\text{when } f(\omega)=0. \]
The requested solution is the rule \(x\) and its positive margin; equivalently, maximize the worst-case margin and ask whether its optimum is positive. In the continuous electorate, these inequalities make the correct collective outcome hold in every state with a strict margin. They capture the separator condition in Proposition 3: since each \(p_\omega\) sums to one, the centered rule \(x-\theta\mathbf{1}\) gives its signed separator, and a strict separator can be scaled to keep \(x\) in \([0,1]^S\).
This is a plausible Class A question: with the probability table listed explicitly, it is a linear program and is tractable in the table size and bit length. It may be polynomial in the explicit input while still exponential in \(K\) if the states and signals are listed as all of \(\{0,1\}^K\). Its scope is the paper’s outcome-based asymptotic-efficiency results, not the comparison between premise-based and outcome-based equilibria as a whole.
The main weakness is also the reason I would not present this as an anchor-compliant case. The paper’s strategic incentives depend on pivotality in finite elections. In a truly atomless population, one voter has no effect on the outcome, so the BNE incentives that are central to the paper disappear. Proposition 3 connects finite-game equilibria to an asymptotic limit, but the proposed LP asks about efficient mass aggregation; it does not compute or characterize those equilibria. A further research question is whether statewise asymptotic efficiency in the finite games always has this strict-separation certificate, or whether boundary cases with vanishing margins can achieve it by \(n\)-dependent strategies.
The strongest objection starts with the paper’s subject: it proves no computational result. Proposition 3 and Corollary 3 are equilibrium and asymptotic-efficiency claims, not complexity statements about an input-output problem. That does not rule out a new computational question, but it leaves no computational result here for a continuous mirror to preserve.
The proposed Strict Mass-Signal Separation problem is the best candidate. Given an explicit table of conditional signal probabilities, its maximum-margin version is a linear program. But its solution is a signal-contingent voting policy, not an equilibrium of the paper’s game. In a genuinely atomless electorate, any one voter has zero effect on the outcome, so every unilateral action is payoff-neutral. The strategic incentive conditions that make the paper’s comparison interesting disappear; the LP supplies a policy from outside the game.
That LP also captures only the strict-separation route to asymptotic efficiency in Proposition 3. It does not decide whether the finite games have efficient equilibrium sequences in cases where strategies or margins vary with \(n\). Corollary 3’s linear-independence test is simpler still: with an explicit table, it is a rank calculation. One could make the input succinct and ask for a more difficult separator problem, but then the computational challenge comes from that encoding and structure, not from a complexity question the paper itself poses.
The strongest repair would keep the finite voters and study symmetric equilibria for large populations grouped by signal. That is a sensible high-multiplicity regime, and I would not argue against it on plausibility grounds. But preserving pivotal incentives requires preserving the finite effect of an individual vote. The resulting problem is about finite games at large \(n\); taking the actual population continuum removes that effect. Treating positive-mass signal blocs as strategic players would restore influence only by changing who the players are.
So the negative case is that the clean continuous model strips away the paper’s strategic core, while the clean computational surrogate is a standard LP for a sufficient condition. That makes this a weak case for ChoCo’s computational programme. It does not establish the universal claim that no worthwhile mirror exists: the paper’s large-electorate asymptotics make a mass-based information-aggregation question natural. But the proponent’s LP does not, by itself, show that the paper contains a worthwhile continuous computational 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.