Pith. sign in

REVIEW 2 major objections 5 minor 44 references

Pure Nash equilibria in multi-strategy network games with conformists, rebels and stubborn agents exist only when conformists have enough local support and vanish almost surely on large random graphs.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 07:45 UTC pith:3FPVXT2P

load-bearing objection Solid multi-strategy extension of CRS network games: clean existence/non-existence theorems, elementary local obstruction, and almost-sure failure on random graphs under transparent hypotheses. the 2 major comments →

arxiv 2607.10997 v1 pith:3FPVXT2P submitted 2026-07-13 econ.TH

Network games with three types of players

Shan Pei , Wenjie Cao , Boyu Zhang This is my paper
classification econ.TH MSC 91A4391A0605C80
keywords network gamespure Nash equilibriumconformistsrebelsstubborn agentspotential gamesrandom graphsmulti-strategy games
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper studies network games in which three kinds of players interact over many strategies: conformists who copy the most common neighbor strategy, rebels who pick the least common, and stubborn agents locked into fixed strategies. The central question is when a pure-strategy Nash equilibrium exists and what it looks like. On general graphs the authors give two sufficient conditions: either there are no edges between conformists and rebels, or every conformist already has at least half its neighbors among other conformists and stubborn agents of one fixed strategy. They also isolate a simple local configuration (a conformist–rebel pair whose remaining neighbors are evenly distributed stubborn agents) that immediately blocks every equilibrium, and they prove that this configuration appears with probability approaching one in large random networks. On complete graphs, lines, rings, trees and stars they supply sharp necessary-and-sufficient criteria and describe the exact frequencies of each strategy at equilibrium. The unifying picture is that equilibria survive when conformists enjoy local majority support from like-minded or stubborn neighbors, and collapse when conformist–rebel edges become abundant.

Core claim

In an m-strategy CRS network game a pure Nash equilibrium exists whenever there are no conformist–rebel edges or every conformist has at least half its neighbors in the set of conformists and stubborn agents of one common strategy; yet on large random networks satisfying mild degree and type conditions such an equilibrium fails to exist almost surely, because a simple local obstruction appears with probability one.

What carries the argument

The local structure M*: a conformist–rebel edge whose remaining neighbors are stubborn agents perfectly balanced across all m strategies. Its presence forces an immediate best-response cycle and therefore precludes any pure Nash equilibrium; second-moment arguments show it occurs almost surely in large random graphs.

Load-bearing premise

The almost-sure non-existence proof assumes that vertices whose degrees are one more than a multiple of the number of strategies still induce a linear number of edges; if that residue class is too sparse the second-moment argument no longer applies.

What would settle it

Generate large Erdős–Rényi graphs with the paper’s type probabilities and check whether the fraction of instances that still possess a pure Nash equilibrium (found by best-response dynamics) stays bounded away from zero as n grows; if it does, Theorem 2 fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper studies pure-strategy Nash equilibria (PNE) in an m-strategy network game with three player types: conformists (who match the modal neighbor strategy), rebels (who match the least common), and stubborn agents (who fix a strategy). On arbitrary networks it gives two sufficient conditions for existence (no CR edges, or every conformist has at least half its neighbors in C ∪ S_a) and proves, via a local obstruction M* and a second-moment argument, that PNE almost surely fails to exist in large random graphs under mild degree and type conditions. For complete graphs, lines, rings, trees (m > Δ), and stars it supplies necessary and sufficient conditions and characterizes equilibrium strategy frequencies. The unifying message is that PNE is supported when every conformist has enough C/S neighbors and is blocked by abundant CR edges.

Significance. The work cleanly unifies coordination, anti-coordination and fixed-behavior motives in a multi-strategy setting that previous literature treated only pairwise or for binary actions. The potential-game reductions (Results 1–2), the elementary local obstruction (Lemma 1), the second-moment non-existence theorem (Theorem 2), and the complete characterizations on classical architectures (Theorems 3–5, Corollaries 1–2) are all self-contained and machine-checkable from the payoff definitions. The random-graph claim is falsifiable and is already supported by the ER simulations of Figure 1. These contributions supply a systematic existence theory for heterogeneous multi-strategy network games that was previously missing.

major comments (2)
  1. Theorem 2 rests on the technical hypothesis that the induced subgraph on V_n* = {i : d_i ≡ 1 (mod m), d_i ≤ K} still contains at least αn edges. While the authors correctly note that this holds for Poisson, normal, scale-free and exponential families (and for the ER simulations of Figure 1), the hypothesis is not automatic for every degree sequence with bounded second moment. A short remark clarifying the precise range of degree distributions for which the linear-edge condition is known to hold, or an alternative argument that avoids the residue-class restriction, would make the almost-sure claim fully robust.
  2. For trees with m ≤ Δ the paper leaves open the existence question (explicitly flagged after Theorem 5 and illustrated by Figure 3B). Because trees are a core architecture and the m > Δ case is already settled, even a partial characterization (e.g., for stars or for bounded-degree trees) would substantially strengthen the special-structure section; otherwise the limitation should be stated more prominently in the abstract and introduction.
minor comments (5)
  1. Result 2 is labelled “RS game” but the surrounding sentence twice writes “CS game”; the typo should be corrected.
  2. In the model section the payoff collection is introduced as U but then written as V; notation should be made consistent.
  3. Figure 1 caption: “exsitence” → “existence”.
  4. The discussion of weighted payoffs (Section 5) is useful but could briefly note whether the local obstruction M* survives under strategy-specific weights.
  5. A few references (e.g., Cao et al. 2026) are cited as arXiv preprints; final versions or DOIs should be supplied if available.

Circularity Check

1 steps flagged

No significant circularity: central existence/non-existence claims are proved directly from the payoff definition and elementary potential-game or second-moment arguments; self-citations supply only standard two-type building blocks.

specific steps
  1. self citation load bearing [Section 3.1, Result 1 and its use in Theorem 1(i)]
    "Result 1. For any network, the m-strategy game with conformists and stubborn agents (CS game) has at least one PNE. This result was introduced in Cao et al. (2024), showing that the m-strategy CS game is an exact potential game … For condition (i), a CRS game without CR edges can be decomposed into a CS game and an RS game. Result 1 and Result 2 show that both … possess a PNE."

    Existence of PNE for the pure CS subgame is asserted solely by citation to a prior paper by overlapping authors (Cao, Zhang) rather than re-derived; that fact is then invoked as one of the two sufficient conditions of Theorem 1. The dependence is real but minor: the multi-strategy CRS claims do not reduce to it by construction, the potential-game argument is elementary and standard, and the complementary RS case is proved in full inside the present manuscript.

full rationale

The paper’s load-bearing results (Theorem 1 sufficient conditions, Lemma 1 local obstruction M*, Theorem 2 almost-sure non-existence via second-moment method on V_n*, and the complete/line/ring/tree/star characterizations) are derived self-containedly from the CRS payoff functions (Eq. 1) and Definition 1 of PNE. Result 2 (RS potential) is proved in full; Result 3 (bipartite CR) is proved in full; the random-graph argument never invokes external data fits or uniqueness theorems. The sole self-citation of substance is Result 1 (CS potential game), taken from Cao et al. (2024) by overlapping authors and used to obtain the “no CR edges” half of Theorem 1. That citation is ordinary reuse of a prior special case, not a circular reduction of the multi-strategy claims, and the potential-game method itself is standard and independently verifiable. No fitted parameters are renamed as predictions, no ansatz is smuggled, and no uniqueness result is imported to force the present conclusions. Hence circularity is negligible.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 2 invented entities

The paper is pure theory. It rests on standard game-theoretic definitions (pure Nash, best response, exact potential games) and on elementary graph-theoretic notions. No free parameters are fitted; the only modeling choices are the three payoff rules and the independent type assignment in the random-graph theorem. The local structure M* is an invented diagnostic device, not an ontological claim.

axioms (4)
  • domain assumption Payoffs of conformists equal the number of neighbors sharing their strategy; payoffs of rebels equal the number of neighbors with a different strategy; stubborn agents have constant payoff 0.
    Eq. (1) and Definition 1; standard linear best-response formulation used throughout the network-games literature.
  • standard math A finite undirected simple graph with no self-loops; pure strategies only.
    Section 2; classical setting for pure-strategy network games.
  • domain assumption In the random-graph model, degrees satisfy a uniform second-moment bound, the induced subgraph on V_n* retains linearly many edges, and types are drawn i.i.d. with full support.
    Hypotheses (i)–(iii) of Theorem 2; needed for the second-moment argument that M* appears a.s.
  • standard math Exact potential games admit pure Nash equilibria (Monderer–Shapley).
    Invoked for CS and RS games (Results 1–2) and for the reduction arguments in Theorems 1 and 4.
invented entities (2)
  • Local structure M* no independent evidence
    purpose: A minimal forbidden configuration (conformist–rebel edge plus balanced stubborn neighbors) that precludes any pure Nash equilibrium on the whole network.
    Definition 2 and Lemma 1; purely combinatorial diagnostic, no independent empirical claim.
  • CRS game no independent evidence
    purpose: Unified multi-strategy model containing conformists, rebels and stubborn agents.
    Section 2; natural extension of earlier two-type models, introduced for the present analysis.

pith-pipeline@v1.1.0-grok45 · 17373 in / 2522 out tokens · 24150 ms · 2026-07-14T07:45:48.165361+00:00 · methodology

0 comments
read the original abstract

In this paper, we analyze a multi-strategy network game with three types of players, conformists, rebels, and stubborn agents. Conformists adopt the strategy that is most common among their neighbors, rebels adopt the least common, and stubborn agents adhere to a fixed strategy. We study the existence and structure of pure strategy Nash equilibrium (PNE). On arbitrary networks, we establish sufficient conditions for PNE existence, and we prove that in large random networks PNE almost surely fails to exist. For several specific network architectures, such as complete network, lines, rings, trees, and stars, we derive necessary and sufficient conditions for PNE existence and fully characterize the equilibrium strategy frequencies. Collectively, these results offer a unified perspective that PNE is likely to exist when every conformist has more conformist and stubborn neighbors, and fails when the network game has numerous conformist-rebel edges.

Figures

Figures reproduced from arXiv: 2607.10997 by Boyu Zhang, Shan Pei, Wenjie Cao.

Figure 1
Figure 1. Figure 1: Estimated probability of PNE existence as the number of players varies across [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Examples of PNE in 3-strategy networked CRS games. (A) A PNE on a line. [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Examples of CRS games on tree networks. (A) A tree admitting a PNE. (B) [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Examples of PNE in 3-strategy CRS games on star networks. (A) A PNE with [PITH_FULL_IMAGE:figures/full_fig_p015_4.png] view at source ↗

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

44 extracted references · 1 linked inside Pith

  1. [1]

    Cuesta, and Angel Sánchez, Triadic influence as a proxy for compatibility in social relationships, Proceedings of the National Academy of Sciences, vol

    Miguel Ruiz-García, Juan Ozaita, María Pereda, Antonio Alfonso, Pablo Brañas-Garza, José A. Cuesta, and Angel Sánchez, Triadic influence as a proxy for compatibility in social relationships, Proceedings of the National Academy of Sciences, vol. 120, no. 13, pp. e2215041120, 2023

    2023

  2. [2]

    Pei, Dynamic games on arbitrary networks with two types of players, Journal of Mathematical Economics, vol. 113, pp. 102990, 2024

    2024

  3. [3]

    Boyu Zhang, Zhigang Cao, Cheng-Zhong Qin, and Xiaoguang Yang, Fashion and homophily, Operations Research, vol. 66, no. 6, pp. 1486--1497, 2018

    2018

  4. [4]

    Wenjie Cao, Hegui Zhang, Gang Kou, and Boyu Zhang, Discrete opinion dynamics in social networks with stubborn agents and limited information, Information Fusion, vol. 109, pp. 102410, 2024

    2024

  5. [5]

    Shapley, Potential games, Games and Economic Behavior, vol

    Dov Monderer and Lloyd S. Shapley, Potential games, Games and Economic Behavior, vol. 14, no. 1, pp. 124--143, 1996

    1996

  6. [6]

    Information Fusion , volume=

    Discrete opinion dynamics in social networks with stubborn agents and limited information , author=. Information Fusion , volume=. 2024 , publisher=

    2024

  7. [7]

    Games and economic behavior , volume=

    Potential games , author=. Games and economic behavior , volume=. 1996 , publisher=

    1996

  8. [8]

    Operations Research , volume=

    Fashion and homophily , author=. Operations Research , volume=. 2018 , publisher=

    2018

  9. [9]

    Proceedings of the National Academy of Sciences , volume =

    Triadic influence as a proxy for compatibility in social relationships , author =. Proceedings of the National Academy of Sciences , volume =

  10. [10]

    Dynamic games on arbitrary networks with two types of players , journal =

  11. [11]

    Opinion fluctuations and disagreement in social networks

    Acemo g lu, D., Como, G., Fagnani, F., Ozdaglar, A., 2013. Opinion fluctuations and disagreement in social networks. Math. Oper. Res. 38, 1--27

    2013

  12. [12]

    Evolutionary dynamics on any population structure

    Allen, B., Lippner, G., Chen, Y.T., Fotouhi, B., Momeni, N., Yau, S.T., Nowak, M.A., 2017. Evolutionary dynamics on any population structure. Nature 544, 227--230

    2017

  13. [13]

    Evolutionary dynamics of higher-order interactions in social networks

    Alvarez-Rodriguez, U., Battiston, F., de Arruda, G.F., Moreno, Y., Perc, M., Latora, V., 2021. Evolutionary dynamics of higher-order interactions in social networks. Nat. Hum. Behav. 5, 586--595

    2021

  14. [14]

    Cournot competition in networked markets

    Bimpikis, K., Ehsani, S., Ilkilic, R., 2019. Cournot competition in networked markets. Manage. Sci. 65, 2467--2481

    2019

  15. [15]

    Identification of peer effects through social networks

    Bramoull\'e, Y., Djebbari, H., Fortin, B., 2009. Identification of peer effects through social networks. J. Econom. 150(1), 41--55

    2009

  16. [16]

    Games played on networks, in: The Oxford Handbook of the Economics of Networks

    Bramoull\'e, Y., Kranton, R., 2016. Games played on networks, in: The Oxford Handbook of the Economics of Networks. Oxford University Press

    2016

  17. [17]

    Strategic interaction and networks

    Bramoull\'e, Y., Kranton, R., D'amours, M., 2014. Strategic interaction and networks. Am. Econ. Rev. 104, 898--930

    2014

  18. [18]

    An experimental study of network effects on coordination in asymmetric games

    Broere, J., Buskens, V., Stoof, H., S\'anchez, A., 2019. An experimental study of network effects on coordination in asymmetric games. Sci. Rep. 9, 6842

    2019

  19. [19]

    Peer effects and social networks in education

    Calv\'o-Armengol, A., Patacchini, E., Zenou, Y., 2009. Peer effects and social networks in education. Rev. Econ. Stud. 76, 1239--1267

    2009

  20. [20]

    Dynamic matching pennies on networks

    Cao, Z., Qin, C.Z., Yang, X., Zhang, B., 2019. Dynamic matching pennies on networks. Int. J. Game Theory 48, 887--920

    2019

  21. [21]

    Network games with heterogeneous players

    Cao, W., S\'anchez, A., Zhang, B., 2026. Network games with heterogeneous players. https://arxiv.org/html/2607.05932v1

    Pith/arXiv arXiv 2026

  22. [22]

    Discrete opinion dynamics in social networks with stubborn agents and limited information

    Cao, W., Zhang, H., Kou, G., Zhang, B., 2024. Discrete opinion dynamics in social networks with stubborn agents and limited information. Inf. Fusion 109, 102410

    2024

  23. [23]

    Divide and conquer in two-sided markets: A potential-game approach

    Chan, L.T., 2021. Divide and conquer in two-sided markets: A potential-game approach. RAND J. Econ. 52, 839--858

    2021

  24. [24]

    Learning, local interaction, and coordination

    Ellison, G., 1993. Learning, local interaction, and coordination. Econometrica, 1047--1071

    1993

  25. [25]

    Network games

    Galeotti, A., Goyal, S., Jackson, M.O., Vega-Redondo, F., Yariv, L., 2010. Network games. Rev. Econ. Stud. 77, 218--244

    2010

  26. [26]

    Heterogeneous network games: Conflicting preferences

    Hern\'andez, P., Mu\ noz-Herrera, M., S\'anchez, A., 2013. Heterogeneous network games: Conflicting preferences. Games Econ. Behav. 79, 56--66

    2013

  27. [27]

    Social and economic networks

    Jackson, M.O., 2008. Social and economic networks. Princeton University Press, Princeton

    2008

  28. [28]

    O., Watts, A., 2002

    Jackson, M. O., Watts, A., 2002. On the formation of interaction networks in social coordination games. Games Econ. Behav. 41, 265--291

    2002

  29. [29]

    Diffusion of behavior and equilibrium properties in network games

    Jackson, M.O., Yariv, L., 2007. Diffusion of behavior and equilibrium properties in network games. Amer. Econ. Rev. 97, 92--98

    2007

  30. [30]

    O., Zenou, Y., 2015

    Jackson, M. O., Zenou, Y., 2015. Games on networks. In Handbook of game theory with economic applications. Elsevier. 4, 95--163

    2015

  31. [31]

    S., 1996

    Monderer, D., Shapley, L. S., 1996. Potential games, Games Econ. Behav. 14, 124--143

    1996

  32. [32]

    Evolutionary games and spatial chaos

    Nowak, M.A., May, R.M., 1992. Evolutionary games and spatial chaos. Nature 359, 826--829

    1992

  33. [33]

    A simple rule for the evolution of cooperation on graphs and social networks

    Ohtsuki, H., Hauert, C., Lieberman, E., Nowak, M.A., 2006. A simple rule for the evolution of cooperation on graphs and social networks. Nature 441, 502--505

    2006

  34. [34]

    Dynamic games on arbitrary networks with two types of players

    Pei, S., Cressman, R., Zhang, B., 2024. Dynamic games on arbitrary networks with two types of players. J. Math. Econ. 113, 102990

    2024

  35. [35]

    H., 2023

    Ramazi, P., Roohi, M. H., 2023. Characterizing oscillations in heterogeneous populations of coordinators and anticoordinators. Automatica 154, 111068

    2023

  36. [36]

    Networks of conforming or nonconforming individuals tend to reach satisfactory decisions

    Ramazi, P., Riehl, J., Cao, M., 2016. Networks of conforming or nonconforming individuals tend to reach satisfactory decisions. Proc. Natl. Acad. Sci. U.S.A. 113, 12985--12990

    2016

  37. [37]

    A., Sánchez, A., 2023

    Ruiz-Garcia, M., Ozaita, J., Pereda, M., Alfonso, A., Brañas-Garza, P., Cuesta, J. A., Sánchez, A., 2023. Triadic influence as a proxy for compatibility in social relationships. Proc. Natl. Acad. Sci. U.S.A. 120, e2215041120

    2023

  38. [38]

    Social diversity promotes the emergence of cooperation in public goods games

    Santos, F.C., Santos, M.D., Pacheco, J.M., 2008. Social diversity promotes the emergence of cooperation in public goods games. Nature 454, 213--216

    2008

  39. [39]

    Information gerrymandering and undemocratic decisions

    Stewart, A.J., Mosleh, M., Diakonova, M., Arechar, A.A., Rand, D.G., Plotkin, J.B., 2019. Information gerrymandering and undemocratic decisions. Nature 573, 117--121

    2019

  40. [40]

    Binary opinion dynamics with stubborn agents

    Yildiz, E., Ozdaglar, A., Acemoglu, D., Saberi, A., Scaglione, A., 2013. Binary opinion dynamics with stubborn agents. ACM Trans. Econ. Comput. 1, 1--30

    2013

  41. [41]

    The dynamics of social innovation

    Young, H.P., 2011. The dynamics of social innovation. Proc. Natl. Acad. Sci. U.S.A. 108, 21285--21291

    2011

  42. [42]

    Z., Yang, X., 2018

    Zhang, B., Cao, Z., Qin, C. Z., Yang, X., 2018. Fashion and homophily. Oper. Res. 66, 1486--1497

    2018

  43. [43]

    Leslie Lamport, : a document preparation system, Addison Wesley, Massachusetts, 2nd edition, 1994

    1994

  44. [44]

    In connection with cross-referencing and possible future hyperlinking it is not a good idea to collect more that one literature item in one + +

    + is cited as + ESG96 +. In connection with cross-referencing and possible future hyperlinking it is not a good idea to collect more that one literature item in one + +. The so-called Harvard or author-year style of referencing is enabled by the package natbib . With this package the literature can be cited as follows: enumerate [ ] Parenthetical: + WB96 ...

    1996