| paper | d4-conf-ecai-2020-0054 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
Theorem 4(ii) supports a recognizable continuous-population Kemeny question, but it only identifies a voting rule. The paper gives no named algorithm, complexity bound, or computational formulation used to compute its winners, and explicitly leaves computation for future work; the required computational anchor is absent.
fails bit a — no named computational result to mirror
The proposed question captures weak Kemeny winner determination from Theorem 4(ii); it leaves the strong tie-break, the other rule classifications, and computational complexity untouched.
There is no qualifying computational anchor in this paper. Theorems 3 and 4 classify which voting rules arise; Theorems 1 and 2 characterize limiting winners. None asserts a complexity result or gives a computational formulation that the paper solves. Section 8 explicitly leaves complexity analysis and Monte Carlo computation for future work.
The closest positive lead is Theorem 4(ii), proved in this paper. It says that if \(\Gamma\) excludes OOpt and either includes Ac or includes both As and T, then \(\Gamma\) is \((p\to1)\)-\(\Delta\)-equivalent to \(\{\mathrm{Ud},\mathrm{Ac}\}\): for every nonzero \(v\), both the strong and weak \(p\to1\) winner sets agree. For \(\Gamma=\{\mathrm{Opt},\mathrm{Ac}\}\), the paper explains that the weak winners are the alternatives placed first in total orders maximizing \(v^+(R)\), giving the Kemeny rule. This is a rule characterization, not a computational result, so it cannot honestly serve as the required anchor.
The best continuous-population question suggested by that result is Continuous Kemeny-Winner Determination. Given \(m\) alternatives and a support of \(\tau\) complete rankings, each with binary-rational mass \(\mu_t\geq0\) summing to \(1\), define
\[
v_\mu(x,y)=\sum_{t:\,x\succ_t y}\mu_t.
\]
Assume both orientations of every pair have positive mass, so the paper’s nonzero-input condition holds. For a candidate consensus ranking \(\rho\), maximize
\[
K_\mu(\rho)=\sum_{x\succ_\rho y}v_\mu(x,y).
\]
Return the set of alternatives placed first in at least one maximizing \(\rho\). This is precisely the weak-winner question in the Kemeny case of Theorem 4(ii); the paper’s strong-winner refinement also breaks ties using how many maximizing rankings put each alternative first.
A plausible regime is a large public consultation with a fixed agenda and many participants sharing a relatively small number of complete rankings. A type is a ranking, its mass is the fraction of participants with that ranking, and the decision variable is the consensus ranking. This is a credible population interpretation: the paper itself motivates \(v(x,y)\) as the number of agents preferring \(x\) to \(y\). But the connection also exposes the weakness: the paper already allows real-valued \(v\), and the rule depends on \(\mu\) only through those pairwise totals. The continuous input therefore changes the population interpretation and support representation more than it changes the rule.
I would expect hardness to transfer: discrete profiles embed as \(\mu_t=n_t/n\), preserving the maximizing rankings up to a positive scale. If full pairwise support is needed, mixing in a balanced ranking distribution gives positive mass in both directions while adding the same amount to every consensus ranking’s score. The resulting question is likely hard for reasons in the number of alternatives, not made tractable by continuizing the population. Further questions include computing the strong tie-break, parameterizing by \(m\) or \(\tau\), approximating the winner set, and measuring winner robustness under changes to \(\mu\).
So the positive case is plausible as a population interpretation of the paper’s Kemeny characterization, but it is not an anchor-based computational mirror under the requested standard. The paper contains no named computational result to anchor one to.
The strongest case against a *result-anchored* mirror is that this paper has no named computational result to mirror. Theorems 3 and 4 classify which voting rules arise from the framework; Theorems 1 and 2 characterize limiting winners; the propositions establish properties such as Pareto and monotonicity. None gives a complexity bound or algorithm. Section 8 explicitly leaves complexity analysis and approximate winner computation for future work. Theorem 4(ii), the proponent’s Kemeny anchor, identifies a rule; it does not solve a computational problem. Continuous Kemeny winner determination would therefore be a new computational question attached to that identification, not the continuous counterpart of a computational result in the paper.
The proposed Kemeny question is hard to dismiss on modeling grounds. A large consultation with ranking types and rational masses is a credible high-multiplicity society. The objective is well-defined on that population, and the paper itself motivates weighted relations by voter counts. The proposal does not depend on voter identities, and normalizing the population to unit mass does not make the question degenerate. The paper’s nonzero-input condition can also be met by an appropriate population. The usual objections about hardness or the answer carrying over from discrete profiles cannot disqualify it under the programme’s criteria.
The best further objection is that the Kemeny computation uses the population only through the pairwise totals \(v_\mu(x,y)\). The paper already accepts arbitrary real-valued \(v\), so supplying a distribution over rankings gives a population interpretation to an input the framework already permits; the support size \(\tau\) adds no information to the winner calculation once those totals are formed. That limits the mirror’s distinct computational content. It does not establish that the question is already solved continuously, and the programme explicitly permits a well-posed mirror whose answer is unchanged or whose hardness transfers.
So I can reject the proponent’s proposal as an anchor to a *computational result* of this paper. I cannot honestly turn that into a convincing claim that no worthwhile continuous mirror exists in any scenario. The Kemeny question remains a credible population-level computational problem, and the paper’s own future-work discussion makes it recognizable to the authors. The negative case is strongest under the strict anchoring requirement and weak as a universal claim.
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.