| paper | d4-conf-aaai-2018-0937 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Proposition 12 supplies a natural population version of its welfare bound, but the proposed optimization over arbitrary agendas is a new computational question, not a computational result of the paper. The paper's convergence results give round bounds, but the case offers no concrete continuous question for them; with atomless agents, unilateral moves lose their effect and the strategic process changes.
fails bit a — no named computational result to mirror
The proposed population formulation covers Proposition 12's static plurality Price-of-Truth bound; it does not provide a continuous analogue for the convergence or Dynamic Price of Anarchy results.
The strongest honest mirror is Continuous Price of Truth for plurality. This paper has named computational results about convergence and social welfare, but no theorem classifying a problem as NP-hard, in P, or FPT. Proposition 12 is the best anchor for a population mirror because it gives a growing welfare gap without relying on a zero-welfare denominator.
Proposition 12, proved in this paper, states that there exist an agenda \(\Phi\) and a group of agents with Hamming-distance preferences for which the plurality rule’s Price of Truth is \(\Omega(|\mathcal J(\Phi)|)\), where \(\mathcal J(\Phi)\) is the set of complete, consistent judgment sets. In the proof, writing \(m=|\Phi^+|\) and \(k=|\mathcal J(\Phi)|\le m\), the authors construct a truthful profile containing each of the \(k\) judgment sets equally often. With a suitable lexicographic tie-break, plurality selects \(J_1\), while the socially best outcome is \(J_2\). The resulting ratio is \(\frac{(k-1)m+2-k}{m+k-2}\), which grows linearly in \(k\) in the construction. The paper proves this result here; it is not a cited theorem.
The regime is a large consultation or membership vote on a fixed bundle of logically related issues. Participants sharing a complete judgment set have the same preferences under the paper’s Hamming-distance model, so they are one type. Mass is their fraction of the population. In the Proposition 12 construction, there are \(k\) types, each with mass \(1/k\), and the number of people can be \(n=xk\) for arbitrarily large \(x\). That is a natural high-multiplicity instance: many participants, relatively few distinct judgment types.
Here is the continuous problem I would hand to a prover. Given an agenda \(\Phi\) and a rational distribution \(\mu\) over supported truthful types \(t\in T\subseteq\mathcal J(\Phi)\), compute the continuous Price of Truth for plurality and return a report assignment attaining it. A type’s desired set is its truthful judgment \(t\). Let \(x_{t,K}\ge0\) be the mass of type \(t\) reporting judgment \(K\), with \(\sum_K x_{t,K}=\mu_t\). The reported mass of \(K\) is \(\nu_K=\sum_t x_{t,K}\); plurality selects a \(K\) with maximum \(\nu_K\), using the paper’s lexicographic tie-break. Define proportional welfare by
\[
W_\mu(K)=\sum_{t\in T}\mu_t\frac{|K\cap t|}{|t|}.
\]
The requested value is
\[
\mathrm{PoT}_\infty(\Phi,\mu)=
\frac{\max_x W_\mu(F^{p\ell}(\nu(x)))}
{W_\mu(F^{p\ell}(\mu))},
\]
with positive truthful-baseline welfare as a promise. A solution is the exact ratio and a feasible report assignment \(x\) attaining it. This is the same social objective as in the paper, with counts replaced by proportions. Any feasible judgment \(K\) can be made the plurality outcome by having all mass report \(K\), so the numerator reduces to maximizing \(W_\mu(K)\) over consistent judgments.
For the Proposition 12 construction, the continuous profile is simply \(\mu_t=1/k\) for every judgment type in its support. The plurality tie and welfare calculation are unchanged, so the linear-in-\(k\) price-of-truth gap has a direct continuous counterpart. This is the lead mirror. I expect the exact computation to be tractable when the feasible judgment sets are explicitly listed, by scanning them and evaluating \(W_\mu\). For succinctly represented arbitrary agendas, I expect agenda-driven hardness: maximizing this weighted agreement over consistent judgments can encode hard logical optimization. That would be Class B if established, since the discrete profile embeds exactly by setting \(\mu_t=n_t/n\). Proposition 12 itself proves the welfare gap, not that complexity classification.
This mirror covers Proposition 12’s static plurality welfare result. It does not claim a continuous analogue of the paper’s convergence or dynamic-price-of-anarchy results. Those rely on unilateral agents changing reports; in an atomless population, an individual’s infinitesimal report change generally cannot alter the collective outcome.
The weak point is that the anchor is a welfare bound rather than an algorithmic hardness or tractability theorem, and the continuous construction preserves the discrete calculation almost exactly. It establishes that the population-distribution question is coherent and useful, but it does not by itself show that continuization changes the complexity landscape. The natural next questions are whether exact continuous Price of Truth is tractable in \(m,\tau,L\) for succinct agendas, whether approximation or parameterization helps, and how its value compares with the finite high-multiplicity version for arbitrary rational \(\mu\).
Proposition 12 is the only strong anchor, and the best case against it is that it proves a welfare bound, not a computational result. The paper gives convergence-round bounds and Price of Truth and Dynamic Price of Anarchy results, but it classifies no computational problem and gives no algorithmic, approximation, or parameterized result. The continuous task proposed for Proposition 12 is therefore a new optimization problem prompted by the paper’s welfare metric, rather than a continuization of computational content the paper establishes.
The proposed task is mathematically coherent. Given \(\mu\), its truthful baseline is the welfare of the plurality winner, and its numerator simplifies to \(\max_{K\in\mathcal J(\Phi)} W_\mu(K)\): all mass can report \(K\), making it the plurality outcome. That asks for a highest-welfare consistent judgment. Proposition 12 neither studies computing that judgment nor gives it a hard instance or an algorithmic structure. Its own witness has equal rational masses and a stated welfare ratio. Extending the question to arbitrary succinct agendas could yield an interesting complexity problem, but that would need independent motivation and a precise agenda encoding. Agenda-driven hardness is allowed by the programme; the point is that the proposition itself does not supply it.
The paper’s dynamic results offer little additional support for a continuous strategic mirror. In an atomless population, a single agent’s report has zero mass and cannot change the plurality outcome. Modeling a positive-mass deviation instead makes the actor a coalition or type-level bloc, changing the unilateral improvement process the paper studies.
The negative case weakens at its most important point: Proposition 12’s type construction is a natural high-multiplicity instance, and the continuous Price of Truth formulation is well-posed. The programme permits a discrete-to-continuous identity to count as a result, so the fact that the witness carries over exactly is not grounds to dismiss it. The strongest defensible objection is that the paper supplies no computational finding to make this a compelling ChoCo lead. I cannot honestly defend the universal claim that no worthwhile mirror exists; if a new weighted-agenda optimization question is enough to anchor one, Proposition 12 survives.
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.