| paper | d4-conf-atal-2011-0283 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
Theorem 7 is a representation characterization, and Corollary 19 establishes a finite model property; neither states a computational result. The proposed mass-based realization project may be worthwhile, but it is a follow-on question with no computational result in this paper to continuize.
fails bit a — no named computational result to mirror
No computational result is mirrored: Theorem 7 gives a structural correspondence, while Corollary 19 gives a finite model property.
The strongest conceptual hook is Theorem 7, proved in this paper: an effectivity function \(E\) corresponds to a strategic game exactly when it is truly playable—equivalently, when it is playable and \(E(\varnothing)\) has a complete nonmonotonic core. In that case, there is a game \(G\) with \(E=E^\alpha_G\). This is a structural representation result, not a computational result. Corollary 19 gives the finite model property for TPCL, but no complexity bound; it does not supply the required computational anchor.
A plausible *follow-on* population question would keep the outcome set finite and make the agents continuous. Let \(T\) be a finite set of complete agent types, with rational masses \(\mu_t\) summing to \(1\). Each type \(t\) has a finite action set \(A_t\), and outcomes lie in a finite set \(S\). A coalition is described by its type masses \(q_t\in[0,\mu_t]\); its action choices are mass allocations \(y_{t,a}\), while the complement chooses \(z_{t,a}\). Given an aggregate outcome rule \(g\), define \(X\subseteq S\) to be enforceable by that coalition when
\[
\exists y\ \forall z,\qquad g(y+z)\in X,
\]
where \(y\) and \(z\) have row sums \(q_t\) and \(\mu_t-q_t\), respectively. One could ask whether a finitely encoded candidate effectivity map \(E(q)\) is induced by such a rule \(g\), and request \(g\) as a witness when it is. This is a genuine continuum-population representation question, with mass as the coalition variable and an outcome rule as the object to construct.
The regime could be a large anonymous population with many interchangeable agents in a small number of roles or action types. It is a plausible extension of the paper’s game/effectivity correspondence, but it is not anchored in a named computational result of the paper. The paper keeps the player set finite: in Proposition 4, proved here, the obstruction arises with one player and infinitely many outcomes. That tempting continuity is in the outcome space, which is outside the programme’s scope. So the positive case is a credible research extension, but under the stated anchor rule this paper yields no qualifying computational mirror, and I cannot honestly place the proposed extension in the programme’s tractability/hardness trichotomy.
The central problem is that none of the paper’s named results is a computational result. Theorem 7 characterizes when an effectivity function is represented by a strategic game; it gives no algorithmic problem, input encoding, objective, or complexity claim to continuize. Corollary 19 establishes a finite model property for a logic, but gives no complexity bound. A continuous-population version of either theorem could be an interesting new structural result, but it would not supply the computational mirror ChoCo is looking for.
The proposed realization question makes that gap concrete. A continuum game with finitely many agent types and actions is a plausible model: coalitions can be specified by mass vectors \(q\), and one can ask whether an outcome rule \(g\) induces a given \(E(q)\). But “finitely encoded” does the hard work here. \(E\) ranges over uncountably many mass vectors and, in the paper’s setting, subsets of an outcome set that may be infinite. To make realization a complexity problem, one must choose a representation for \(E\) and \(g\)—for example, a semialgebraic description or a circuit—and bound the witness representation. Those choices define a new computational problem; the paper’s theorem neither supplies nor selects them. The scenario may be worth studying, but it is an adjacent nonatomic game-form project, not a mirror of a computational result in this paper.
The paper’s specific obstruction reinforces the mismatch. Proposition 4 uses one player and an infinite outcome set; its issue is a non-principal filter over outcomes, not population multiplicity. Replacing finite players with masses does not address that feature. Restrict to finite outcomes and the paper itself shows that every playable effectivity function is truly playable (Corollary 6); retain infinite outcomes and the proposed computational question still needs a new encoding for the outcome and effectivity objects. This is not an argument that a dull answer is uninteresting. It is why no population regime in the paper yields a defined computational question to classify.
The same problem affects Corollary 19. A population version would need a new language for coalitions described by masses, and a finite representation of its models and semantics. That could lead to a decision-complexity project, but it is not a complexity result already present in the finite model property.
So the negative case is strong against calling this paper a continuous computational mirror: it has no computational result to anchor one, and the best proposed continuation must add its own formal input model. It is weaker as a claim that *no worthwhile research project* could grow from the paper. An encoded realization problem for anonymous nonatomic games could be worthwhile; the honest objection is that its worth would belong to that new problem, not to a continuization of this paper’s computational landscape.
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.