| paper | d4-conf-atal-2017-0297 |
| authors | — |
| venue | ? |
| filed under | unclassified |
| judged by | gpt-6-luna / xhigh (triple__luna__xhigh__c6r1) |
| judge confidence | high |
| authors would recognise it | unclear |
The numbered propositions and corollaries establish delay and welfare results for a two-agent, two-task scenario; they do not assert an algorithm, complexity bound, hardness result, or other computational result. The paper describes algorithm extensions and experiments, but attributes the polynomial-time claim for FMC_TA to prior work. The proposed population model is a plausible new scheduling problem, but it is not anchored in a named computational result of this paper.
fails bit a — no named computational result to mirror
No anchored result survives. The paper presents FMC_TA+ and FMC_TAH+ and evaluates them experimentally, but gives no numbered computational result for those algorithms.
I can’t give an anchor-based positive case under your rule: the paper has no numbered result asserting a computational complexity or algorithmic classification. Proposition 1 proves a delay bound for a two-agent, two-task scenario; Proposition 2 derives the social-welfare difference between cooperative and individual execution; Corollaries 1 and 2 give conditions for cooperation to improve welfare. These are proved here, but they are analytic results, not complexity results. The polynomial-time statement about FMC_TA is attributed to prior work [1], and the experiments on FMC_TA+ and FMC_TAH+ are not named computational results.
There is a plausible *unanchored* mirror. In a finite-horizon continuous LEP-H scheduling problem, each type \(t\) could specify responders’ skills, staging area, shift-start availability, and identical utility and cost parameters; \(\mu_t\) would be that type’s fraction of the workforce. The instance would include the paper’s tasks, with rational arrival times, locations, skill workloads, capability functions, and interruption costs. A solution would be an aggregate schedule assigning mass of each type to tasks, skills, and time periods, respecting workforce, travel, and task-completion constraints, while maximizing discounted team utility. A plausible regime is a large regional police or rescue force pooled across a small number of staging areas and skill categories, so the workforce is much larger than the number of types.
That is close to the authors’ interest in skill-aware cooperation, but it is not a result they establish: the paper’s fractional allocations concern portions of tasks, not a continuous population of agents. I would expect much of the dynamic scheduling difficulty to survive in the aggregate model because task order and travel remain; that is a conjecture for a new problem, not a hardness conclusion supported by this paper.
The strongest negative is that this paper supplies no named computational result for ChoCo to mirror. Proposition 1 is a delay bound for two officers and two tasks; Proposition 2 and Corollaries 1 and 2 are welfare calculations for the same small scenario. They are analytical results, not complexity claims. The paper’s new contributions are algorithm extensions and empirical comparisons. Its polynomial-time statement about FMC_TA is attributed to earlier work [1]. Recasting the propositions with population fractions would yield new delay and welfare formulas, but not a mirror of this paper’s computational contribution.
The proposed LEP-H population model is more promising than those anchors. A large workforce pooled across a few stations and skill groups is a plausible high-multiplicity regime, and fractional task allocations in the paper do not already make its population continuous. But making the proposed aggregate schedule faithful requires tracking each cohort’s changing location, active task and progress, availability, and interruption costs. One can repair the model by representing a workforce as mass over complete feasible trajectories, with shared task completion and timing constraints. That is a coherent new optimization problem; it is not a continuous version of a result this paper establishes. The paper’s algorithm first clears a market over task fractions, then constructs individual schedules heuristically. It gives no complexity result for optimization over aggregate trajectories, or for the pricing problem such a formulation might produce.
So the case against is strong only in a narrower sense: this paper is a poor anchor for a continuization programme organized around computational results. I cannot honestly turn that into the universal claim that no worthwhile mirror exists. A trajectory-based model could be worth studying, even though it would require specifying a new computational problem rather than mirroring one of this paper’s results.
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.