pith. sign in

arxiv: 2606.28100 · v1 · pith:BCYPIIMVnew · submitted 2026-06-26 · 💻 cs.GT · math.OC· math.PR· q-bio.PE

Discrete Event Population Updates: finding game theoretic emergent behaviour in queueing systems with simulation

Vincent Knight , Geraint I. Palmer-Liyu , Thomas Hutton This is my paper

Pith reviewed 2026-06-29 01:58 UTC · model grok-4.3

classification 💻 cs.GT math.OCmath.PRq-bio.PE
keywords discrete event simulationevolutionary game theoryqueueing systemsreplicator dynamicsMoran processjockeying modelstrategic behaviorsimulation-based analysis
0
0 comments X

The pith

A single long discrete event simulation can be coupled directly to evolutionary update rules to analyze strategic behavior in queueing systems without closed-form payoffs.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces Discrete Event Population Updates to apply evolutionary game theory to queueing systems that lack closed-form payoff expressions. It does this by running a single long discrete event simulation and feeding the resulting payoffs into standard evolutionary update rules such as replicator dynamics or Moran replacement. This approach is demonstrated on a multi-server jockeying model where no closed forms exist. A sympathetic reader would care because it extends evolutionary analysis to a much wider class of systems that can be simulated but not solved analytically, while also showing substantial computational speedups.

Core claim

The central claim is that Discrete Event Population Updates (DEPU) couple a single long run of a discrete event simulation directly to an evolutionary population update rule, removing the closed-form constraint that previously limited evolutionary game theory in queueing systems. Implementations via Discrete Event Replicator Dynamics and Discrete Event Moran Replacement applied to a multi-server jockeying model reach comparable precision tens of times faster than nesting short simulations inside an outer evolutionary loop, and make systematic parameter sweeps tractable because each operating point costs only one simulation run.

What carries the argument

Discrete Event Population Updates (DEPU), which integrate one long discrete event simulation run directly with evolutionary population update rules such as replicator dynamics or Moran replacement.

If this is right

  • Evolutionary game theory applies to any queueing system that can be built in a discrete event simulator.
  • Each operating point in parameter space requires only a single simulation run.
  • Systematic parameter sweeps over model parameters become computationally tractable.
  • Comparable precision is reached tens of times faster than the standard practice of nesting short simulations inside an evolutionary loop.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could extend to other simulation-based models in operations research that involve population-level strategic interactions.
  • It opens the possibility of hybrid methods that combine discrete event simulation with agent-based evolutionary dynamics for real-time queue management.
  • Direct comparison against additional models with known analytical solutions would test the robustness of the single-run payoff coupling.

Load-bearing premise

That the payoffs obtained from a single long discrete event simulation run can be fed directly into standard evolutionary update rules to produce accurate population dynamics without requiring closed-form expressions or additional validation of the simulation-to-evolution coupling.

What would settle it

Applying the method to a queueing model that admits known closed-form payoffs and observing that the resulting population dynamics diverge from those computed using the closed forms.

Figures

Figures reproduced from arXiv: 2606.28100 by Geraint I. Palmer-Liyu, Thomas Hutton, Vincent Knight.

Figure 1
Figure 1. Figure 1: The multi-node jockeying queueing system. [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Fitness and replicator dynamics derivative for Λ = 10 [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Diagram of the three phase event scheduling approach used in Ciw. [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The emergent behaviour for the example considered in Figure 2(b). (a): [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Examples of runs of the Moran process and large-population convergence. At each Moran iteration [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Diagrammatic representation comparing using DEPU and the using DES to approximate fitnesses [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Emergent behaviour of the model from Section 3.1 under DERD and DEMR, and the effect of [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Diagram showing the relationships between DEPU, numerical solution of the replicator dynamics [PITH_FULL_IMAGE:figures/full_fig_p017_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The queueing model of [20]. Individuals join or baulk based on a threshold strategy ˜n [PITH_FULL_IMAGE:figures/full_fig_p017_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Discrete Event Replicator Dynamics on Naor’s observable M/M/1 queue with Λ = 30 [PITH_FULL_IMAGE:figures/full_fig_p018_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: DERD applied to the two-node model with the extended six-strategy space ( [PITH_FULL_IMAGE:figures/full_fig_p019_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: DEPU applied to the four-floor jockeying model (Λ = 40, [PITH_FULL_IMAGE:figures/full_fig_p021_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Stationary distribution of the four-floor jockeying model under two parameter sweeps. Each [PITH_FULL_IMAGE:figures/full_fig_p023_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Computational efficiency of RD (blue) and DERD (red) on the jockeying example, measured by [PITH_FULL_IMAGE:figures/full_fig_p024_14.png] view at source ↗
read the original abstract

Strategic behaviour in queueing systems has been studied extensively in the behavioural queueing literature, but almost exclusively for systems that admit closed-form expressions for the cost or utility experienced by a strategic user. Evolutionary game theory offers a mature framework for analysing populations whose individual payoffs depend on the composition of the population itself, and would in principle apply to a much wider class of queueing systems; its application has, however, been constrained by the same closed-form requirement. We introduce Discrete Event Population Updates (DEPU), a general algorithmic framework that couples a single long run of a discrete event simulation (DES) directly to an evolutionary population update rule, removing that constraint. We present two implementations: Discrete Event Replicator Dynamics (DERD), which follows an Euler discretisation of the replicator dynamics equation, and Discrete Event Moran Replacement (DEMR), which maintains a finite population updated via Moran-style copying events. Both are applied to a multi-server jockeying model for which no closed-form fitness expressions are available. On the jockeying model considered, DEPU reaches comparable precision tens of times faster than the standard practice of nesting short simulations inside an outer evolutionary loop, and because each operating point then costs only a single simulation run it also makes systematic parameter sweeps tractable. This brings the toolkit of evolutionary dynamics within reach of any system a modeller can build in a discrete event simulator.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper introduces Discrete Event Population Updates (DEPU) as an algorithmic framework coupling a single long discrete event simulation (DES) run directly to evolutionary population update rules, with two implementations: Discrete Event Replicator Dynamics (DERD) following an Euler discretisation of the replicator equation and Discrete Event Moran Replacement (DEMR) using Moran-style copying in a finite population. It applies the method to a multi-server jockeying queueing model without closed-form fitnesses and claims that DEPU achieves comparable precision to nested short simulations inside an outer evolutionary loop but tens of times faster, while also enabling tractable parameter sweeps.

Significance. If the DEPU coupling is shown to be accurate, the work would extend evolutionary game theory to queueing systems that admit simulation but not analytic solution, removing a key barrier in behavioural queueing research. The approach is algorithmic, introduces no free parameters, and avoids circularity or fitted quantities.

major comments (2)
  1. [Abstract and results on the jockeying model] The central claim that payoffs from one long DES run can be fed directly into replicator or Moran updates to produce accurate population dynamics (abstract; results on jockeying model) is load-bearing but rests on an unvalidated assumption. No benchmark against any closed-form model is provided to test for systematic bias from on-the-fly payoff estimation, event interleaving, or time-scale choices, leaving the accuracy of the simulation-to-evolution coupling untested relative to the nested-simulation baseline.
  2. [Methods descriptions of DERD and DEMR] Implementation details, error analysis, and validation data for simulation fidelity are absent (abstract; methods descriptions of DERD/DEMR), so the performance and precision claims cannot be assessed for soundness.
minor comments (1)
  1. [Abstract] Notation for the two implementations (DERD vs. DEMR) could be introduced more explicitly when first mentioned to aid readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments, which highlight important aspects of validation and methodological detail. We respond to each major comment below and agree that revisions are warranted to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and results on the jockeying model] The central claim that payoffs from one long DES run can be fed directly into replicator or Moran updates to produce accurate population dynamics (abstract; results on jockeying model) is load-bearing but rests on an unvalidated assumption. No benchmark against any closed-form model is provided to test for systematic bias from on-the-fly payoff estimation, event interleaving, or time-scale choices, leaving the accuracy of the simulation-to-evolution coupling untested relative to the nested-simulation baseline.

    Authors: We agree that the absence of a benchmark against a closed-form model leaves the coupling unvalidated with respect to potential systematic biases. Although the framework targets systems without closed-form fitnesses, we will add a dedicated validation subsection in the revised manuscript. This will apply DEPU to a simple queueing game (e.g., an M/M/1 pricing game) that admits closed-form fitness expressions, allowing direct comparison of the resulting population dynamics against both the known analytic replicator/Moran trajectories and the nested-simulation baseline. revision: yes

  2. Referee: [Methods descriptions of DERD and DEMR] Implementation details, error analysis, and validation data for simulation fidelity are absent (abstract; methods descriptions of DERD/DEMR), so the performance and precision claims cannot be assessed for soundness.

    Authors: We accept that the current Methods section lacks sufficient implementation detail, error analysis, and fidelity validation to allow full assessment of the performance claims. In the revision we will expand the descriptions of DERD and DEMR to include pseudocode, explicit discussion of time-scale separation assumptions, potential biases from event interleaving and on-the-fly payoff estimation, and additional quantitative comparisons (error metrics and runtime ratios) against the nested-simulation baseline on the jockeying model. revision: yes

Circularity Check

0 steps flagged

No circularity: algorithmic coupling is independent of inputs

full rationale

The paper introduces DEPU as a direct algorithmic coupling of a single long DES run to standard replicator or Moran update rules. No equations, fitted parameters, or derivations reduce the claimed speed or population dynamics to self-referential definitions or renamed inputs. The method is presented as a general framework applicable to any simulatable system, with performance claims based on empirical timing comparisons rather than constructed equivalences. Lack of closed-form validation is a separate empirical concern, not a circularity reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Ledger populated from abstract only. The central claim rests on the domain assumption that simulation-derived payoffs can substitute for closed-form fitness in evolutionary dynamics.

axioms (1)
  • domain assumption Simulation outputs from a single long DES run provide reliable fitness values that can be inserted directly into replicator dynamics or Moran replacement rules.
    This premise is required for the DEPU coupling to produce valid evolutionary trajectories; it is invoked when the abstract states that the framework removes the closed-form requirement.

pith-pipeline@v0.9.1-grok · 5790 in / 1314 out tokens · 23114 ms · 2026-06-29T01:58:31.044332+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

38 extracted references

  1. [1]

    Individual versus social optimization in the allocation of customers to alternative servers.Management Science, 29(7):831–839, 1983

    Colin E Bell and Shaler Stidham Jr. Individual versus social optimization in the allocation of customers to alternative servers.Management Science, 29(7):831–839, 1983

    1983

  2. [2]

    A solution for queues with instantaneous jockeying and other customer selection rules.Naval Research Logistics Quarterly, 17(3):315–325, 1970

    Ralph L Disney and William E Mitchell. A solution for queues with instantaneous jockeying and other customer selection rules.Naval Research Logistics Quarterly, 17(3):315–325, 1970

    1970

  3. [3]

    Congestion tolls for Poisson queueing processes.Economet- rica, 43(1):81–92, 1975

    Noel M Edelson and David K Hildebrand. Congestion tolls for Poisson queueing processes.Economet- rica, 43(1):81–92, 1975. 25

    1975

  4. [4]

    Springer Science & Business Media, 2011

    Walter Gautschi.Numerical analysis. Springer Science & Business Media, 2011

    2011

  5. [5]

    Social stability and equilibrium.Econometrica, 59(3):859–867, 1991

    Itzhak Gilboa and Akihiko Matsui. Social stability and equilibrium.Econometrica, 59(3):859–867, 1991

    1991

  6. [6]

    Two queues before a single server.Biometrika, 45(1/2):1–11, 1958

    Frank A Haight. Two queues before a single server.Biometrika, 45(1/2):1–11, 1958

    1958

  7. [7]

    CRC press, 2016

    Refael Hassin.Rational queueing. CRC press, 2016

    2016

  8. [8]

    Springer, New York, 2003

    Refael Hassin and Moshe Haviv.To Queue or Not to Queue: Equilibrium Behavior in Queueing Systems. Springer, New York, 2003

    2003

  9. [9]

    The price of anarchy in an exponential multi-server.Operations Research Letters, 35(4):421–426, 2007

    Moshe Haviv and Tim Roughgarden. The price of anarchy in an exponential multi-server.Operations Research Letters, 35(4):421–426, 2007

    2007

  10. [10]

    Simulating dynamical features of escape panic.Nature, 407(6803):487–490, 2000

    Dirk Helbing, Ill´ es Farkas, and Tam´ as Vicsek. Simulating dynamical features of escape panic.Nature, 407(6803):487–490, 2000

    2000

  11. [11]

    Computation and simulation of evolution- ary game dynamics in finite populations.Scientific reports, 9(1):6946, 2019

    Laura Hindersin, Bin Wu, Arne Traulsen, and Julian Garc´ ıa. Computation and simulation of evolution- ary game dynamics in finite populations.Scientific reports, 9(1):6946, 2019

    2019

  12. [12]

    Cambridge Univer- sity Press, Cambridge, 1998

    Josef Hofbauer and Karl Sigmund.Evolutionary Games and Population Dynamics. Cambridge Univer- sity Press, Cambridge, 1998

    1998

  13. [13]

    Measuring the price of anarchy in critical care unit interactions.Journal of the Operational Research Society, 68(6):630–642, 2017

    Vincent Knight, Izabela Komenda, and Jeff Griffiths. Measuring the price of anarchy in critical care unit interactions.Journal of the Operational Research Society, 68(6):630–642, 2017

    2017

  14. [14]

    Selfish routing in public services.European Journal of Operational Research, 230(1):122–132, 2013

    Vincent A Knight and Paul R Harper. Selfish routing in public services.European Journal of Operational Research, 230(1):122–132, 2013

    2013

  15. [15]

    On jockeying in queues.Management Science, 12(5):412–436, 1966

    Ernest Koenigsberg. On jockeying in queues.Management Science, 12(5):412–436, 1966

    1966

  16. [16]

    A review of the discrete choice models applied to evacuation modelling.Transportation Research Part C: Emerging Technologies, 70:83–97, 2016

    Ruggiero Lovreglio, Achille Fonzone, and Luigi Dell’Olio. A review of the discrete choice models applied to evacuation modelling.Transportation Research Part C: Emerging Technologies, 70:83–97, 2016

    2016

  17. [17]

    A model for rational abandonments from invisible queues

    Avishai Mandelbaum and Nahum Shimkin. A model for rational abandonments from invisible queues. Queueing Systems, 36:141–173, 2000

    2000

  18. [18]

    Cambridge University Press, Cambridge, 1982

    John Maynard Smith.Evolution and the Theory of Games. Cambridge University Press, Cambridge, 1982

    1982

  19. [19]

    Random processes in genetics

    Patrick Alfred Pierce Moran. Random processes in genetics. InMathematical proceedings of the cam- bridge philosophical society, volume 54, pages 60–71. Cambridge University Press, 1958

    1958

  20. [20]

    The regulation of queue size by levying tolls.Econometrica: journal of the Econometric Society, pages 15–24, 1969

    Pinhas Naor. The regulation of queue size by levying tolls.Econometrica: journal of the Econometric Society, pages 15–24, 1969

    1969

  21. [21]

    Harvard university press, 2006

    Martin A Nowak.Evolutionary dynamics: exploring the equations of life. Harvard university press, 2006

    2006

  22. [22]

    Palmer, Vincent A

    Geraint I. Palmer, Vincent A. Knight, Paul R. Harper, and Asyl L. Hawa. Ciw: An open-source discrete event simulation library.Journal of Simulation, 13(1):68–82, 2019

    2019

  23. [23]

    A game theoretic model of the behavioural gaming that takes place at the ems-ed interface.European Journal of Operational Research, 305(3):1236–1258, 2023

    Michalis Panayides, Vince Knight, and Paul Harper. A game theoretic model of the behavioural gaming that takes place at the ems-ed interface.European Journal of Operational Research, 305(3):1236–1258, 2023

    2023

  24. [24]

    John Wiley & Sons, Inc., 1998

    Michael Pidd.Computer simulation in management science. John Wiley & Sons, Inc., 1998

    1998

  25. [25]

    Bloomsbury Publishing, 2014

    Stewart Robinson.Simulation: the practice of model development and use. Bloomsbury Publishing, 2014. 26

    2014

  26. [26]

    How bad is selfish routing?Journal of the ACM, 49(2):236–259, 2002

    Tim Roughgarden and ´Eva Tardos. How bad is selfish routing?Journal of the ACM, 49(2):236–259, 2002

    2002

  27. [27]

    MIT Press, Cambridge, MA, 2010

    William H Sandholm.Population Games and Evolutionary Dynamics. MIT Press, Cambridge, MA, 2010

    2010

  28. [28]

    Ten simple rules for repro- ducible computational research.PLOS Computational Biology, 9(10):e1003285, 2013

    Geir Kjetil Sandve, Anton Nekrutenko, James Taylor, and Eivind Hovig. Ten simple rules for repro- ducible computational research.PLOS Computational Biology, 9(10):e1003285, 2013

    2013

  29. [29]

    Why imitate, and if so, how? A boundedly rational approach to multi-armed bandits

    Karl H Schlag. Why imitate, and if so, how? A boundedly rational approach to multi-armed bandits. Journal of Economic Theory, 78(1):130–156, 1998

    1998

  30. [30]

    A conservative index heuristic for routing problems with multiple heterogeneous service facilities.Mathematical Methods of Operations Research, 92:511– 543, 2020

    Rob Shone, Vincent A Knight, and Paul R Harper. A conservative index heuristic for routing problems with multiple heterogeneous service facilities.Mathematical Methods of Operations Research, 92:511– 543, 2020

    2020

  31. [31]

    Containment of socially optimal policies in multiple-facility markovian queueing systems.Journal of the Operational Research Society, 67(4):629–643, 2016

    Rob Shone, Vincent A Knight, Paul R Harper, Janet E Williams, and John Minty. Containment of socially optimal policies in multiple-facility markovian queueing systems.Journal of the Operational Research Society, 67(4):629–643, 2016

    2016

  32. [32]

    Comparisons between observable and unobservable m/m/1 queues with respect to optimal customer behavior.European Journal of Operational Research, 227(1):133–141, 2013

    Rob Shone, Vincent A Knight, and Janet E Williams. Comparisons between observable and unobservable m/m/1 queues with respect to optimal customer behavior.European Journal of Operational Research, 227(1):133–141, 2013

    2013

  33. [33]

    Reinforcement learning: An introduction.A Bradford Book, 2018

    Richard S Sutton. Reinforcement learning: An introduction.A Bradford Book, 2018

    2018

  34. [34]

    Evolutionary stable strategies and game dynamics.Mathematical Biosciences, 40(1–2):145–156, 1978

    Peter D Taylor and Leo B Jonker. Evolutionary stable strategies and game dynamics.Mathematical Biosciences, 40(1–2):145–156, 1978

    1978

  35. [35]

    Ciw: 3.2.4

    The Ciw library developers. Ciw: 3.2.4

  36. [36]

    Coevolutionary dynamics: from finite to infinite populations.Physical review letters, 95(23):238701, 2005

    Arne Traulsen, Jens Christian Claussen, and Christoph Hauert. Coevolutionary dynamics: from finite to infinite populations.Physical review letters, 95(23):238701, 2005

    2005

  37. [37]

    Deciding which queue to join: Some counterexamples.Operations Research, 34(1):55–62, 1986

    Ward Whitt. Deciding which queue to join: Some counterexamples.Operations Research, 34(1):55–62, 1986

    1986

  38. [38]

    Wilkinson, Michel Dumontier, IJsbrand Jan Aalbersberg, Gabrielle Appleton, Myles Axton, Arie Baak, Niklas Blomberg, Jan-Willem Boiten, Luiz Bonino da Silva Santos, Philip E

    Mark D. Wilkinson, Michel Dumontier, IJsbrand Jan Aalbersberg, Gabrielle Appleton, Myles Axton, Arie Baak, Niklas Blomberg, Jan-Willem Boiten, Luiz Bonino da Silva Santos, Philip E. Bourne, et al. The FAIR guiding principles for scientific data management and stewardship.Scientific Data, 3:160018, 2016. 27

    2016