| paper | d4-conf-aaai-2014-0116 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
No numbered result gives an algorithm, complexity classification, or computational formulation used to compute. The brute-force manipulation experiment reports empirical frequencies without a stated guarantee, leaving no named computational result to anchor a continuous-population mirror.
fails bit a — no named computational result to mirror
No result is mirrored; the rule-property propositions and empirical manipulation experiment have no eligible computational anchor.
There is no eligible computational anchor in this paper. Its numbered propositions concern voting-rule properties: Proposition 4 proves monotonicity; Proposition 6 proves that no RDSR is Condorcet-consistent under its stated conditions; Proposition 7 identifies lexicographic scoring rules represented in the Borda family; and Proposition 9 proves a fairness property for strictly decreasing weights. None classifies an underlying problem as tractable or hard, or gives an algorithmic result of the kind your anchor requirement calls for.
The closest positive lead is the manipulation experiment: the authors use brute force to compare how often a single voter can manipulate Borda and interval-weighted RDSRs across generated profiles. A plausible continuous follow-up would take a distribution over ranking types and ask for the minimum mass of one true-preference type that must change its reported rankings to make a more-preferred candidate win under a continuum version of the OWA scores. A large population with relatively few ranking types gives this a sensible high-multiplicity regime.
But that follow-up has no qualifying named result to anchor it, and it changes the experiment’s single manipulator into a positive-mass coalition while extending finite OWA vectors to a continuum weighting function. So it may be a worthwhile new question, but I cannot present it as a continuous mirror of a named computational result in this paper. There are zero anchors, and therefore zero per-anchor prover questions under the requested standard.
The case for supplies no eligible anchor. Propositions 4, 6, 7, and 9 concern rule properties: monotonicity, Condorcet consistency, representation by scoring rules, and fairness under transfers. None gives a complexity classification or an algorithm for a computational problem. The closest candidate is the brute force experiment on single-voter manipulation, but it reports empirical frequencies, not a named computational result of the kind this programme mirrors.
The strongest continuous version of that experiment would give ranking types masses \(\mu_t\), extend the OWA weights to a quantile functional, and ask how much mass must change its report to make a preferred candidate win. That is coherent, but it changes the paper’s manipulation question. In a mass-based continuum, one voter has measure zero and cannot change the score distribution. Giving a positive mass the power to deviate turns the question into coalition manipulation. The paper neither studies that problem nor establishes a computational result about the single-voter version to mirror.
The extension is not uniquely determined by the experiments, either: as the electorate grows, fixed-rank trimming can vanish into average scoring, while proportional trimming or a median limit gives a different rule. Choosing among those limits may produce worthwhile new research, but it would be a new computational study rather than a mirror of a result in this paper.
So the strongest honest negative case is narrow: this paper provides no paper-anchored continuous mirror under the programme’s standard. I cannot establish the universal claim that no continuous population question inspired by it could be worthwhile; the mass-coalition version could be. But that possibility does not supply a computational anchor the paper lacks.
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.