| paper | d4-conf-ijcai-2009-0274 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
The paper’s named theorems characterize efficiency and neutrality of voting rules and correspondences; they do not state algorithms, complexity bounds, or hardness results. A distribution over complete CP-net types gives a credible population model, but a mass-weighted version of Theorem 4 remains axiomatic and cannot satisfy the computational-result gate.
fails bit a — no named computational result to mirror
The closest proposed analogue is a mass-weighted version of Theorem 4; it covers an axiomatic efficiency claim, while none of the paper’s dichotomy theorems yields a computational mirror.
There is a plausible high-multiplicity regime here: a large community voting on several linked issues, with many residents sharing the same complete CP-net conditional preference table. Each type’s mass would be its fraction of the population, and local majority choices could be computed from those masses. The authors would likely recognize that as a natural population version of their voting setting.
But the paper has no numbered computational-complexity result to anchor a ChoCo mirror. The nearest candidates are axiomatic. Theorem 3, proved here, classifies the efficient or neutral seat-by-seat correspondences outside the two-binary-issue case: only dictatorships, anti-dictatorships, and the correspondence that always returns all alternatives qualify. Theorem 4, also proved here, says that sequential composition of two majority correspondences on two binary issues is Pareto efficient. Theorems 5 and 6 give the corresponding dichotomies for correspondences and rules. In these theorems, “efficient” means socially efficient, as defined in the paper—not polynomial-time tractable.
So there is no qualifying lead anchor and no honest per-anchor computational problem or complexity-class prediction to give. A mass-weighted version of Theorem 4 would be a plausible axiomatic extension, but it would not supply the computational result this programme requires. The positive case is therefore limited: the population model is credible, while the paper’s results do not establish a computational mirror.
The high-multiplicity setting is credible: a large community could contain many residents with the same complete CP-net, and fractions of residents are meaningful. I would not argue that the population model itself fails.
The problem is that the paper’s results ask a different kind of question. Theorem 4 says that sequential majority is Pareto efficient on two binary issues. Its natural mass version replaces the two voter counts in the proof with masses and reaches the same equality. That may be a sensible axiomatic result, but it is not a computational question. Theorems 3, 5, and 6 likewise classify which correspondences have efficiency or neutrality properties across all legal profiles. Extending their domain from finite profiles to distributions would extend the axioms’ scope, not produce the complexity landscape ChoCo studies.
The best computational rescue would be to invent a mass bribery or campaigning problem: specify costs for changing CP-net types, then ask how much population must change to make a target alternative win. That could be interesting, but the paper supplies neither such an intervention nor a cost model. Different choices—changing one conditional preference, an entire table, or a voter’s whole CP-net—define different problems. The paper’s own sequential procedure is closer to direct weighted aggregation, and its output can be represented compactly. The proposed intervention problem would therefore be a follow-on attached to the paper’s setting, not a computational mirror of one of its results.
So the strongest negative case is about fit with the programme, not about whether a continuum is sensible: none of the named results anchors a computational mirror. I cannot honestly claim that no worthwhile continuous problem could be built from this setting; a mass intervention problem might be worth studying on its own. But the proponent’s credible population scenario does not make this paper’s axiomatic contribution a ChoCo computational result.
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.