| paper | d4-conf-wine-2008-0054 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | yes |
NP-hard→NP-hard discrete result → its continuous mirror
Answer: \(\mathrm{NP}\text{-hard}\) under polynomial-time oracle reductions · discrete side: hardness · adjudication: proved_with_minor_gaps
full proof & adversarial review →
Theorem 3.1
statement extracted from the paper’s text layer
Given \(n\) candidates and a finite support of trader-offer types \(t=(A_t,π_t,q_t)\), each with a two-entry bidding matrix, binary-rational mass μ_t, limit price π_t, and per-trader order cap \(q_t\), where \(\sum_t μ_t=1\), compute the maximum per-capita worst-case profit. Choose accepted order volumes \(x_t\) and reserve \(r\) to maximize \(\sum_t π_t x_t-r\), subject to \(r\ge\sum_t x_t\mathbf{1}[A_t\bullet M_σ>0]\) for every permutation σ and \(0\le x_t\leμ_t q_t\).
A society is a distribution μ over complete trader-offer types \((A_t,π_t,q_t)\). The organizer chooses accepted order volume \(x_t\) per unit of population mass and reserve \(r\), maximizing per-capita revenue minus the worst-case payout over all candidate rankings.
Since accepted order quantities are already divisible, the mirror mainly pools capacity by offer type and may add limited structural novelty.
fatal: False
The mirror covers the two-entry fixed-reward organizer hardness in Theorem 3.1. It leaves aside the proportional-betting pricing results and the maximum-entropy outcome-distribution results, which concern pricing or outcome-space distributions rather than a continuous trader population.
I’d make the positive case around one lead mirror: Fixed-Reward Permutation Market Organizer\(_\infty\).
The anchor is Theorem 3.1, proved in this paper: the organizer’s optimization problem for fixed-reward betting is NP-hard even when each trader’s bidding matrix has only two nonzero entries. The paper reduces from MAX-SAT, using MAX-2-SAT for the two-entry restriction, and invokes the standard equivalence between optimization and separation. The underlying problem is to choose which orders to accept to maximize worst-case profit across all outcome permutations.
In the mirror, a type \(t\) is a complete trader offer: a two-entry bidding matrix \(A_t\), limit price \(\pi_t\), and per-trader order cap \(q_t\). The society is a distribution \(\mu\) over the \(\tau\) supported offer types, with \(\sum_t \mu_t=1\). The variable \(x_t\) is accepted order volume per unit of population mass for type \(t\); it satisfies \(0\le x_t\le\mu_tq_t\). For permutation \(\sigma\), let \(M_\sigma\) be its candidate-position matrix. Each accepted unit of type \(t\)’s order pays out when at least one of its two specified pairs occurs. The organizer chooses accepted volumes \(x\) and a reserve \(r\) to solve
\[ \max_{x,r}\ \sum_t \pi_t x_t-r \]
subject to
\[ r\ge \sum_t x_t\,\mathbf{1}[A_t\bullet M_\sigma>0] \quad\text{for every }\sigma\in\mathcal S_n, \qquad 0\le x_t\le\mu_tq_t. \]
An instance gives \(n\), the supported types and their matrices, binary rational masses, prices and caps. A solution is an optimal accepted-volume vector and reserve; equivalently, one can ask whether the maximum per-capita profit reaches a given threshold.
A plausible regime is a large election-forecast or horse-race market with standardized tickets and limit-order tiers. Many participants may submit the same sparse ticket at the same price and cap, so the number \(N\) of traders can greatly exceed the number \(\tau\) of distinct offers. That is a natural high-multiplicity society: mass records the fraction of traders in each offer class. The paper already permits fractional accepted orders for each named trader. The mirror retains those order decisions and makes the trader population itself a distribution over types.
I expect the hardness to transfer, in Class B. When named traders share a complete type, their accepted volumes can be summed without changing revenue or any permutation’s payout; setting \(\mu_t\) to each type’s population fraction gives the same optimization after normalizing profit by population size. Duplicating every trader in a hard instance preserves those fractions while making \(N/\tau\) arbitrarily large. The reduction’s combinatorics remain in the candidates, positions and clause-like tickets, so this is not a hardness argument that depends on one-person targeting. Further questions include the approximation guarantees available in \(n\) and \(\tau\), whether useful parameterized algorithms exist, and whether the two-entry restriction admits a sharper boundary between hard and tractable cases.
The weakest point is the regime assumption. Participants with distinct prices or caps become distinct types, and in markets without standardized offers \(\tau\) could approach \(N\). The hardness anchor also does not establish hardness when \(\tau\) is fixed independently of \(n\). The mirror is strongest for markets with genuinely repeated offer classes; in that setting it asks the paper’s organizer problem with its original payoffs and objective, over a high-multiplicity trader population.
The strongest negative case is about fit and marginal value, not formalizability. Theorem 3.1 is the paper’s clear complexity anchor: it proves NP-hardness for the fixed-reward organizer problem. The other computational results concern pricing proportional bets and reconstructing a distribution over outcomes; neither is a population-continuity result.
For Theorem 3.1, a trader’s relevant type can be the complete offer \((A,\pi,q)\). But the paper already lets the organizer accept fractional order quantities. If \(n_t\) traders share a type, their decisions aggregate exactly to one quantity \(X_t\in[0,n_tq_t]\); with \(\mu_t=n_t/N\), the normalized quantity \(x_t=X_t/N\) has cap \(\mu_tq_t\). Revenue and every permutation’s worst-case payout aggregate the same way. So the proposed mirror is sound, but the population mass mostly records pooled order capacity in a model that already treats orders as divisible. In many real markets, exact offers may also be sparse only because prices and caps are discretized; where offers vary freely, the support can grow with the number of traders.
That is a reason to question how much new computational terrain this mirror opens, but it does not defeat the stronger version. A market with standardized tickets, price tiers and caps could plausibly have many traders per offer type. The type definition preserves everything the organizer’s problem uses, and rational masses map back to finite high-multiplicity instances. Neither the existing fractional order variables nor the exact aggregation makes the population regime incoherent.
I therefore cannot make a strong case that no worthwhile mirror exists in any scenario. The refined fixed-reward mirror survives: it is well posed, anchored in a proved complexity result, and has a credible repeated-offer regime. The honest negative case is weak; it is that this mirror may offer less new modeling insight than cases where population continuity changes the computational structure. That is not enough to rule it out under the programme’s standard.
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.