| paper | d4-conf-aaai-2013-0068 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | no |
The paper reports empirical findings and describes market designs, but contains no numbered theorem, lemma, corollary, or proposition asserting an algorithmic or hardness result. A continuous model of manipulation or market informativeness would introduce a new computational problem rather than mirror a computational result proved here.
fails bit a — no named computational result to mirror
No result is covered by a continuous mirror. The paper's findings about price predictivity, insider information, manipulation, and market-maker comparisons are empirical.
I don’t find a qualifying anchor in this paper. It contains no numbered theorem, lemma, corollary, or proposition asserting a computational result such as polynomial-time solvability or hardness. Its main findings—price predictivity, insider information, limited evidence of manipulation, and differences between LMSR and BMM—are empirical. The paper describes market-making algorithms, but does not give a numbered complexity result about the underlying problem.
A large course could provide a plausible population regime: students might be grouped by their information and trading or rating behavior, with each group’s mass representing its share of the class. But a computational question about optimizing trades or ratings in that model would be a new formalization, not a continuous mirror of a named result in this paper. So there are no anchor-specific continuous problems to state here.
The attached case offers no continuous question anchored to a result, so there is no anchor to defeat. Its main findings are empirical: prices predict later ratings, in-class traders contribute information, manipulation appears limited, and BMM prices were more stable than LMSR prices. The paper describes market-making algorithms, but gives no computational problem with a complexity result. That is the decisive gap for a continuous *mirror*.
The strongest rescue is to formalize manipulation as a minimum-mass control problem: given a distribution of students by rating eligibility, information, trading budget, and action costs, find the least mass whose reports and trades can move a market to a target. A large online course could make that population plausible; treating prices or costs as type-level properties is no objection. But the paper reports observed manipulation and asks whether it occurs. It does not define or analyze an optimal attack problem. That formulation may be a worthwhile new project, but it is not a continuous version of a computational result in this paper.
The same distinction applies to optimizing market-maker parameters or modeling how much insider mass makes prices informative. Those could be useful questions, but they introduce computational objectives absent from the study. So the case against a mirror is strong for this paper, though it cannot rule out a separate computational project on robust instructor-rating markets.
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.