pith. sign in

arxiv: 2606.27653 · v1 · pith:DSHJMVCLnew · submitted 2026-06-26 · 💻 cs.GT · econ.TH· q-bio.PE

Characterisation of reactive Nash equilibria in repeated additive games

Franziska Lesigang , Christian Hilbe , Nikoleta E. Glynatsi This is my paper

Pith reviewed 2026-06-29 02:47 UTC · model grok-4.3

classification 💻 cs.GT econ.THq-bio.PE
keywords reactive strategiesNash equilibriarepeated gamesadditive gamesequalizer strategiesevolutionary dynamicssocial learning
0
0 comments X

The pith

Symmetric Nash equilibria for reactive strategies in repeated additive games correspond one-to-one with non-empty subsets of actions used against themselves.

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

The paper establishes that when two players employ reactive strategies in repeated additive games, the requirements for symmetric Nash equilibria simplify to a system of linear equalities and inequalities on the strategy parameters. This simplification enables a full characterization of all such equilibria. A direct one-to-one mapping is shown between each non-empty subset S of the action set and a distinct class of equilibria that employ precisely those actions from S in self-play. The analysis recovers equalizer strategies as the special case supported on the full action set. Complementary simulations of social learning dynamics indicate that the observed prevalence of different equilibrium classes depends on how readily they arise and how resistant they are to invasion.

Core claim

For reactive strategies in repeated additive games, the conditions for symmetric Nash equilibria reduce to a system of linear equalities and inequalities in the strategy parameters. There is a one-to-one correspondence between non-empty subsets S of the action set and S-supporting equilibrium classes, which use exactly the actions in S when playing against themselves. Equalizer strategies appear as the equilibria supported on the entire action set.

What carries the argument

S-supporting equilibria, the one-to-one correspondence between each non-empty subset S of the action set and the class of equilibria that employ exactly the actions in S during self-play.

If this is right

  • All symmetric reactive Nash equilibria can be identified by solving the linear system separately for each possible non-empty subset S.
  • Equalizer strategies are recovered exactly as the equilibria supported on the full action set.
  • Evolutionary simulations show that equilibrium class prevalence is set by the combination of generation likelihood and invasion robustness.

Where Pith is reading between the lines

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

  • The linear reduction could permit systematic enumeration of equilibria even when the number of actions grows moderately large.
  • Identifying which S classes are favored under social learning may suggest how interaction rules could be adjusted to stabilize particular outcomes.
  • The correspondence structure offers a template for checking whether similar linear characterizations exist when strategies incorporate slightly more history.

Load-bearing premise

Players restrict themselves to reactive strategies that respond only to the opponent's immediately preceding action, and the stage games possess an additive payoff structure.

What would settle it

A concrete symmetric Nash equilibrium using reactive strategies in an additive game whose supported actions against itself do not match any single subset S, or that fails to satisfy the corresponding linear equalities and inequalities.

Figures

Figures reproduced from arXiv: 2606.27653 by Christian Hilbe, Franziska Lesigang, Nikoleta E. Glynatsi.

Figure 1
Figure 1. Figure 1: Upper bound on the degrees of freedom of [PITH_FULL_IMAGE:figures/full_fig_p018_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Evolutionary dynamics of reactive strategies. (a), [PITH_FULL_IMAGE:figures/full_fig_p019_2.png] view at source ↗
read the original abstract

In this paper, we study reactive strategies in repeated additive games between two players with finitely many actions. Reactive strategies condition only on the opponent's previous action, making them one of the simplest ways players can respond to past interactions. Additive games include important models of cooperation, such as the donation game and games with a punishment option. We show that, for this class of games and strategies, the conditions for symmetric Nash equilibria reduce to a system of linear equalities and inequalities in the strategy parameters, allowing us to characterise all such equilibria. We establish a one-to-one correspondence between non-empty subsets S of the action set and equilibrium classes, which we call S-supporting equilibria. These are equilibria that use exactly the actions in S when playing against themselves. As a special case, we recover the well-known equalizer strategies as the equilibria supported on the entire action set. To assess which equilibrium classes are most evolutionarily relevant, we complement our analytical characterisation with simulations of social learning dynamics. We find that their prevalence is determined by two factors: how likely they are to be generated and how robust they are against invasion.

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

0 major / 3 minor

Summary. The paper claims that for reactive strategies (conditioning only on the opponent's previous action) in repeated additive games (payoffs separable as u(a,b)=x(a)+y(b)), the conditions for symmetric Nash equilibria reduce to a system of linear equalities and inequalities in the strategy parameters. It establishes a one-to-one correspondence between non-empty subsets S of the action set and classes of S-supporting equilibria (those inducing exactly the actions in S against themselves), recovers equalizer strategies as the full-support case, and uses simulations of social learning dynamics to assess which classes are evolutionarily prevalent based on generation probability and invasion robustness.

Significance. If the central algebraic reduction holds, the work supplies a complete, parameter-free characterization of symmetric reactive equilibria in this restricted but practically relevant class, directly leveraging payoff separability and the Markovian structure of reactive strategies. The explicit S-support bijection and recovery of equalizers are clear strengths, as is the link to evolutionary dynamics via simulation. No internal inconsistency or circularity is apparent; the result is a direct consequence of the stated assumptions rather than fitted parameters.

minor comments (3)
  1. [Abstract / §3] The abstract and introduction assert that equilibrium conditions 'reduce to a system of linear equalities and inequalities' but do not exhibit the explicit system or a worked example for a small action set (e.g., |A|=2 or 3). Adding this in §3 or §4 would make the reduction immediately verifiable without altering the central claim.
  2. [§5] Simulations are invoked to rank equilibrium classes by evolutionary relevance, yet the text provides no detail on the update rule, population size, mutation rate, or number of runs. This is a presentation gap that affects reproducibility but not the mathematical characterization.
  3. [§2] Notation for the reactive strategy matrix (rows/columns indexed by actions) and the induced stationary distribution should be stated once with a clear reference to the transition probabilities; minor inconsistencies in indexing appear in the current draft.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of the manuscript, the clear summary of our contributions, and the recommendation for minor revision. No major comments appear in the report.

Circularity Check

0 steps flagged

No significant circularity; derivation is algebraic reduction from stated assumptions

full rationale

The paper's central claim is a direct algebraic characterization: symmetric Nash conditions for reactive strategies in additive games reduce to linear equalities/inequalities because payoffs separate as u(a,b)=x(a)+y(b) and reactive strategies induce Markov chains whose stationary distributions and deviation payoffs are linear in the strategy matrix entries. The S-support correspondence is a definitional partition by the support of the induced action process. No self-citation is load-bearing for the main theorem, no parameters are fitted then renamed as predictions, and no ansatz or uniqueness result is smuggled in. The derivation is self-contained within the restricted strategy class and payoff structure; simulations are presented only as a complement for evolutionary relevance.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The work rests on standard definitions of symmetric Nash equilibrium and the restriction to reactive strategies in additive stage games; no free parameters or invented entities are introduced in the abstract.

axioms (2)
  • standard math Symmetric Nash equilibrium is defined by the usual best-response condition in the repeated game.
    Invoked to define the equilibria being characterized.
  • domain assumption Stage games have additive payoff structure.
    Required for the linear reduction to hold.

pith-pipeline@v0.9.1-grok · 5734 in / 1273 out tokens · 58071 ms · 2026-06-29T02:47:39.302282+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

40 extracted references

  1. [1]

    & Hamilton, W

    Axelrod, R. & Hamilton, W. D. The evolution of cooperation.Science211, 1390–1396 (1981)

    1981

  2. [2]

    Mailath, G. J. & Samuelson, L.Repeated games and reputations(Oxford Univ. Press, Oxford, UK, 2006)

    2006

  3. [3]

    & van Veelen, M

    Garc´ ıa, J. & van Veelen, M. No strategy can win in the repeated prisoner’s dilemma: Linking game theory and computer simulations.Frontiers in Robotics and AI5, 102 (2018)

    2018

  4. [4]

    Press, Princeton, NJ, 2010)

    Sigmund, K.The Calculus of Selfishness(Princeton Univ. Press, Princeton, NJ, 2010)

    2010

  5. [5]

    G., Ohtsuki, H

    Rand, D. G., Ohtsuki, H. & Nowak, M. A. Direct reciprocity with costly punishment: Gen- erous tit-for-tat prevails.Journal of theoretical biology256, 45–57 (2009)

    2009

  6. [6]

    & Schuster, H

    Hauert, C. & Schuster, H. G. Effects of increasing the number of players and memory size in the iterated prisoner’s dilemma: a numerical approach.Proceedings of the Royal Society B 264, 513–519 (1997)

    1997

  7. [7]

    A., Cuesta, J

    Martinez-Vaquero, L. A., Cuesta, J. A. & Sanchez, A. Generosity pays in the presence of direct reciprocity: A comprehensive study of 2x2 repeated games.PLoS ONE7, E35135 (2012)

    2012

  8. [8]

    Stewart, A. J. & Plotkin, J. B. Small groups and long memories promote cooperation.Scien- tific Reports6, 26889 (2016)

    2016

  9. [9]

    Press, W. H. & Dyson, F. J. Iterated prisoner’s dilemma contains strategies that dominate any evolutionary opponent.Proceedings of the National Academy of Sciences109, 10409–10413 (2012)

    2012

  10. [10]

    & Zelen` y, M

    Lev´ ınsk` y, R., Neyman, A. & Zelen` y, M. Should I remember more than you? On the best response to factor-based strategies. Tech. Rep., Jena Economic Research Papers (2010)

    2010

  11. [11]

    E., Akin, E., Nowak, M

    Glynatsi, N. E., Akin, E., Nowak, M. A. & Hilbe, C. Conditional cooperation with longer memory.Proceedings of the National Academy of Sciences(2024)

    2024

  12. [12]

    LaPorte, P., Hilbe, C., Glynatsi, N. E. & Nowak, M. A. Payoff equivalence and complete strategy spaces of direct reciprocity.Proceedings of the National Academy of Sciences123, e2518486123 (2026)

    2026

  13. [13]

    & Hauert, C

    McAvoy, A., Rao, A. & Hauert, C. Intriguing effects of selection intensity on the evolution of prosocial behaviors.PLoS Computational Biology17, e1009611 (2021)

    2021

  14. [14]

    & Glynatsi, N

    Lesigang, F., Hilbe, C. & Glynatsi, N. E. Can I afford to remember less than you? Best responses in repeated additive games.Economics Letters250, 112300 (2025)

    2025

  15. [15]

    Stewart, A. J. & Plotkin, J. B. Collapse of cooperation in evolving games.Proceedings of the National Academy of Sciences USA111, 17558 – 17563 (2014)

    2014

  16. [16]

    S., Nowak, M

    Park, P. S., Nowak, M. A. & Hilbe, C. Cooperation in alternating interactions with memory constraints.Nature Communications13, 737 (2022)

    2022

  17. [17]

    & Sigmund, K

    Hilbe, C., Traulsen, A. & Sigmund, K. Partners or rivals? Strategies for the iterated prisoner’s dilemma.Games and Economic Behavior92, 41–52 (2015)

    2015

  18. [18]

    & Wang, L

    Zhang, F., Zhou, L., Zhang, G. & Wang, L. Direct reciprocity in multi-action repeated games. Journal of Theoretical Biology618, 112312 (2026)

    2026

  19. [19]

    & Wang, L

    Zhang, F., Chen, F., Zhou, L. & Wang, L. Beyond the prisoner’s dilemma: Cooperation in repeated three-action games.Chaos, Solitons & Fractals204, 117747 (2026)

    2026

  20. [20]

    Nowak, M. A. & Sigmund, K. Game-dyamical aspects of the prisoner’s dilemma.Applied Mathematics and Computation30, 191–213 (1989). 15

    1989

  21. [21]

    Nowak, M. A. & Sigmund, K. Tit for tat in heterogeneous populations.Nature355, 250–253 (1992)

    1992

  22. [22]

    Imhof, L. A. & Nowak, M. A. Stochastic evolutionary dynamics of direct reciprocity.Pro- ceedings of the Royal Society B277, 463–468 (2010)

    2010

  23. [23]

    Allen, B., Nowak, M. A. & Dieckmann, U. Adaptive dynamics with interaction structure. American Naturalist181, E139–E163 (2013)

    2013

  24. [24]

    C., Nowak, M

    Boerlijst, M. C., Nowak, M. A. & Sigmund, K. Equal pay for all prisoners.American Mathematical Monthly104, 303–307 (1997)

    1997

  25. [25]

    Hilbe, C., Nowak, M. A. & Sigmund, K. Evolution of extortion in iterated prisoner’s dilemma games.Proceedings of the National Academy of Sciences110, 6913–6918 (2013)

    2013

  26. [26]

    Traulsen, A., Nowak, M. A. & Pacheco, J. M. Stochastic dynamics of invasion and fixation. Physical Review E74, 011909 (2006)

    2006

  27. [27]

    & T˝ oke, C

    Szab´ o, G. & T˝ oke, C. Evolutionary Prisoner’s Dilemma game on a square lattice.Physical Review E58, 69–73 (1998)

    1998

  28. [28]

    & Imhof, L

    Fudenberg, D. & Imhof, L. A. Imitation processes with small mutations.Journal of Economic Theory131, 251–262 (2006)

    2006

  29. [29]

    S., Wang, L

    Wu, B., Gokhale, C. S., Wang, L. & Traulsen, A. How small are small mutation rates?Journal of Mathematical Biology64, 803–827 (2012)

    2012

  30. [30]

    Imitation processes with small mutations

    McAvoy, A. Comment on “Imitation processes with small mutations”.J. Econ. Theory159, 66–69 (2015)

    2015

  31. [31]

    A., Sasaki, A., Taylor, C

    Nowak, M. A., Sasaki, A., Taylor, C. & Fudenberg, D. Emergence of cooperation and evolu- tionary stability in finite populations.Nature428, 646–650 (2004)

    2004

  32. [32]

    & Tirole, J.Game theory(MIT press, 1991)

    Fudenberg, D. & Tirole, J.Game theory(MIT press, 1991)

    1991

  33. [33]

    Do Yi, S., Baek, S. K. & Choi, J.-K. Combination with anti-tit-for-tat remedies problems of tit-for-tat.Journal of Theoretical Biology412, 1–7 (2017)

    2017

  34. [34]

    A., Chatterjee, K

    Hilbe, C., Martinez-Vaquero, L. A., Chatterjee, K. & Nowak, M. A. Memory-n strategies of direct reciprocity.Proceedings of the National Academy of Sciences114, 4715–4720 (2017)

    2017

  35. [35]

    & Baek, S

    Murase, Y. & Baek, S. K. Five rules for friendly rivalry in direct reciprocity.Scientific reports 10, 16904 (2020)

    2020

  36. [36]

    Memory-two zero-determinant strategies in repeated games.Royal Society open science8, 202186 (2021)

    Ueda, M. Memory-two zero-determinant strategies in repeated games.Royal Society open science8, 202186 (2021)

    2021

  37. [37]

    Controlling conditional expectations by zero-determinant strategies

    Ueda, M. Controlling conditional expectations by zero-determinant strategies. InOperations Research Forum, vol. 3, 48 (Springer, 2022)

    2022

  38. [38]

    Li, J.et al.Evolution of cooperation through cumulative reciprocity.Nature Computational Science2, 677–686 (2022)

    2022

  39. [39]

    P., Perc, M., Xu, H

    Fang, Y., Benko, T. P., Perc, M., Xu, H. & Tan, Q. Synergistic third-party rewarding and punishment in the public goods game.Proceedings. Mathematical, Physical, and Engineering Sciences475, 20190349 (2019)

    2019

  40. [40]

    The iterated prisoner’s dilemma: good strategies and their dynamics.Ergodic Theory, Advances in Dynamical Systems77–107 (2016)

    Akin, E. The iterated prisoner’s dilemma: good strategies and their dynamics.Ergodic Theory, Advances in Dynamical Systems77–107 (2016). 16 Figures and tables SConditions {C} (a)p C = (1,0,0) (b)b 1pM(C) +b 2pM(M)−c 2 ≤b 1 −c 1 (c)b 1pD(C) +b 2pD(M)≤b 1 −c 1 {M} (a)p M = (0,1,0) (b)b 1pC(C) +b 2pC(M)−c 1 ≤b 2 −c 2 (c)b 1pD(C) +b 2pD(M)≤b 2 −c 2 {D} (a)p D...

    2016