Pith. sign in

REVIEW 3 major objections 4 minor 51 references

Beyond Bayesian Nash: Learning Minimax-Regret Equilibria for Adversarial Team Games under Asymmetric Information

T0 review · 3 major / 4 minor · reviewed 2026-07-14 · grok-4.5

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

Pith's one-line read PR-MRE is a regret-based equilibrium that protects the uninformed team against strategic shifts in hidden types while still using a nominal prior, and it can be learned via a robust double-oracle loop.

desk verdict Clean middle-ground equilibrium (PR-MRE) and a usable PSRO meta-solver; theory holds under its threat model, but the Graph CtF claim rests on one seed iteration and an unvalidated SDP relaxation. read the letter →

arxiv 2607.09993 v1 pith:LHKDBEUG submitted 2026-07-10 cs.GT cs.AIcs.LGcs.MA

Naman Aggarwal , Jonathan P. How This is my paper
classification cs.GTcs.AIcs.LGcs.MA
keywords adversarialteamgamesasymmetricinformationminimaxregretPR-MREPSROdistributionshiftgraphcapture-the-flagrobustequilibrium
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

In adversarial team games one side knows a hidden type (for example a flag location) and can condition play on it; the other side only has a prior. Bayesian Nash strategies that trust that prior can be badly exploited when the type distribution is shifted. This paper introduces Probabilistically Robust Minimax-Regret Equilibrium (PR-MRE): the uninformed team minimizes worst-case expected exploitability only over distributions that keep rare types rare under the nominal prior. That yields a tighter performance lower bound than fully distribution-free minimax regret, while still covering adversarial redistribution of mass among typical types. For normal-form games the equilibrium is a robust bilinear program with a tractable semidefinite relaxation; the authors embed that relaxation as the meta-solver in a double-oracle algorithm (PRMRE-PSRO) that expands populations with deep RL best responses. On graph Capture-the-Flag the resulting Blue policies scout both corridors instead of over-committing to the majority flag and keep higher win-rates when the type distribution moves.

What carries the argument

Probabilistically Robust Minimax-Regret Equilibrium (PR-MRE): the Blue operator that minimizes the supremum of expected exploitability over the typicality-preserving threat set S(nominal, δ). For finite normal-form games this is a robust bilinear program; an SDP lift with McCormick cuts and rank-1 projection is used as the meta-solver inside a robust double-oracle / PSRO loop with RL best responses.

What would settle it

On the same Graph CtF seed metagame, run identical PSRO expansions with BNE (or pure MRE) as meta-solver instead of PR-MRE; if Blue win-rate for test mass on the majority flag ≤ 0.6 no longer exceeds the BNE curve, or if the symmetric corridor-scouting visitation pattern disappears, the central empirical claim fails.

Watch

Extended reading notes

Core claim

PR-MRE is the fixed point of a Blue best response that minimizes worst-case expected exploitability over the typicality-preserving set of type distributions together with the informed Red best response. Among the concepts compared, it supplies the tightest uniform lower bound on expected payoff for every distribution inside that set, and the SDP-relaxed double-oracle PRMRE-PSRO produces Blue policies whose worst-case win-rate under type shifts substantially exceeds that of Bayesian-Nash PSRO on the Graph CtF instance.

Load-bearing premise

The argument needs strategic deception to only redistribute mass while keeping types that are rare under the nominal prior still rare, and needs the SDP relaxation of the robust bilinear program to be accurate enough that the learned policies inherit that certificate.

Editorial extensions

If this is right

  • Uninformed-team strategies hedge across high-confidence type subsets rather than specializing to the modal type under the nominal prior.
  • Learned policies exhibit scouting before commitment, lowering exploitability when the realized type is not the majority hypothesis.
  • Performance lower bounds hold for any redistribution of mass that keeps rare types rare (low-probability covers stay low-probability).
  • The same meta-solver can replace the usual Nash meta-solver in PSRO for any normal-form abstraction of an asymmetric-information team game.
  • Fully distribution-free minimax-regret equilibrium is recovered as the special case δ = 0.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same typicality-preserving threat model could be applied to other hidden-parameter multi-agent settings (unknown skill tiers, map layouts) without requiring divergence-ball ambiguity sets.
  • Whether the McCormick SDP relaxation stays tight enough for larger type spaces will decide how far PRMRE-PSRO scales beyond two-flag Graph CtF.
  • A sequential (extensive-form) refinement of PR-MRE that admits local regret decomposition would open a path to CFR-style algorithms for the same robustness objective.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper introduces Probabilistically Robust Minimax-Regret Equilibrium (PR-MRE) for adversarial team games with asymmetric information. Under a typicality-preserving threat model (Definition 3.2), PR-MRE minimizes worst-case expected regret over admissible type distributions S(μ̄,δ) rather than expected payoff or fully distribution-free worst-case regret. Lemmas 1–4 establish uniform performance lower bounds; for normal-form games the equilibrium is cast as a robust bilinear program (PRMRE-EQM-NF) that is lifted to a McCormick-cut SDP with rank-1 projection (Appendix A.2). The SDP is used as a meta-solver inside a robust double-oracle / PSRO loop (PRMRE-PSRO, Algorithms 1–2) whose best responses are obtained by MAPPO on GNN policies. On a 2v2 Graph Capture-the-Flag instance the method is claimed to produce Blue policies with better worst-case win-rate under type-distribution shifts than BNE-PSRO, together with scouting-like visitation patterns.

Significance. If the claims hold, the work supplies a clean intermediate point on the risk-neutral / distributionally-robust / fully worst-case spectrum for asymmetric-information team games, together with an explicit programming formulation and a population-based learning algorithm. The normal-form bilinear program, the SDP lift, and the lower-bound certificates (Lemmas 1–4) are mathematically self-contained and correctly derived from standard regret algebra and water-filling. The Graph CtF experiments illustrate a concrete behavioral consequence (scouting before commitment) that is of practical interest for multi-agent path-finding and reachability games. These contributions are novel relative to existing ATG and PSRO literature and would be of interest to the algorithmic game-theory and multi-agent RL communities, provided the empirical link between the SDP meta-solver and the claimed robustness certificate is tightened.

major comments (3)
  1. The central empirical claim (Figs. 6–9, §4.1) attributes improved worst-case win-rates under distribution shift to PR-MRE. That attribution rests on the McCormick-cut SDP + rank-1 projection (Appendix A.2) producing a meta-equilibrium whose regret vector is close to the true robust bilinear optimum of PRMRE-EQM-NF. The paper never reports duality gap, recovered rank, residual after projection, or any other fidelity metric on the actual 2×4 Graph CtF seed metagame that carries the claim; only the tiny synthetic 1×5×3 / 2×5×3 matrices are solved. Without such diagnostics it is impossible to know whether the mixture fed to the MAPPO best-response is a genuine PR-MRE mixture or an uncontrolled approximation, and therefore whether the robustness curves can be credited to the proposed concept.
  2. Empirical support is limited to a single expansion iteration on a hand-crafted 2×4 seed metagame (Figs. 6–7) with no multi-seed statistics, no error bars, and no comparison against a pure MRE (δ=0) baseline or against DRO-MaxMin on the same structured ambiguity set. The performance-robustness curve (Fig. 8) and the visitation densities (Fig. 9) are therefore suggestive but insufficient to establish that PRMRE-PSRO systematically discovers more robust strategies than risk-neutral PSRO on graph-structured ATGs.
  3. The typicality-preserving threat model (Definition 3.2) is presented as the appropriate model of strategic deception, yet the paper supplies no external justification or sensitivity analysis for the choice of δ, nor does it examine how the recovered Blue mixture changes when the low-probability covers are misspecified. Because Lemma 4’s “tightest lower bound” certificate holds only inside S(μ̄,δ), the practical value of the certificate depends on whether that set is a realistic description of the adversary’s power; this assumption remains untested outside the synthetic matrices.
minor comments (4)
  1. Notation for the low-probability cover set oscillates between Θℓ, Θl and L; a single consistent symbol would improve readability.
  2. Figure 3’s caption refers to “δ3” and “δ4” which appear to be typographical errors for θ3 and θ4.
  3. The relationship between the BR_MMR stopping criterion used in Algorithm 2 and the exact PR-MRE best-response operator (Definition 3.3) is only sketched; a short formal statement would clarify that the training loop is still approximating the same fixed point.
  4. Several references to “Appendix A.1” for proofs of Lemmas 1 and 4 are correct, but the main text could briefly indicate that the proofs are elementary rearrangements so that a reader need not consult the appendix for the core argument.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: PR-MRE, lower-bound certificates, and PSRO meta-solver are self-contained constructions, not fits or self-citation chains.

full rationale

The paper's load-bearing claims are definitional equilibrium constructions and derived certificates, not predictions forced by fitted inputs. PR-MRE (Def. 3.3) is defined as minimizing worst-case expected regret over the typicality-preserving set S(μ̄,δ); Lemma 4 then states that this arginf yields the tightest V_S^(-) lower bound for every μ in S. That is ordinary optimality of an optimization problem (the subtracted term in the lower bound is exactly the objective being minimized), not a circular reduction of an independent empirical claim to its inputs. Lemmas 1–3 follow algebraically from the exploitability definition (Def. 2.4) and the threat model (Def. 3.2). The normal-form programs (PRMRE-BR-NF / PRMRE-EQM-NF) and the McCormick-cut SDP lift with rank-1 projection (A.2) are standard reformulations of a robust bilinear program; they do not smuggle an ansatz via self-citation, nor import a uniqueness theorem from the authors. References are external (Harsanyi, CFR, PSRO, minimax-regret mechanism design). Empirical Graph CtF results compare learned policies under distribution shift against BNE-PSRO; they do not fit a parameter on a subset and re-label a related quantity as a prediction. Algorithmic self-consistency of the double-oracle loop (meta-equilibrium on the current population, BR_MMR stopping relative to that population) is ordinary iterative approximation, not definitional circularity. Score 0 is therefore the correct finding.

Assumptions & free parameters 3 free parameters · 4 assumptions · 3 invented entities

The central claim rests on a new threat model and equilibrium definition, standard game-theoretic best-response fixed points, and an unproven but standard SDP relaxation of a bilinear program. Free parameters are the confidence level δ and the nominal prior; invented entities are the PR-MRE concept itself and the typicality-preserving set S.

free parameters (3)
  • confidence level δ
    Hand-chosen (0.15 in the synthetic examples, 0 for the Graph CtF run); controls how much mass may be placed on rare types and therefore directly shapes the equilibrium.
  • nominal type distribution μ̄
    Taken as given (e.g., (0.8,0.2) or (0.25,0.70,0.05)); all high-probability covers and the set S are defined relative to it.
  • L1-ball radius ρ for DRO-Ball baseline
    Set to 0.5 in the synthetic comparison; affects only the baseline, not PR-MRE itself.
assumptions (4)
  • domain assumption Red team observes the realized type and best-responds type-conditionally (Definition 2.1).
    Standard asymmetric-information modeling choice for the Capture-the-Flag setting; load-bearing for the three-player Nature–Red collusion view.
  • ad hoc to paper Admissible distribution shifts preserve all high-probability covers of the nominal (Definition 3.2).
    The paper’s own threat model; not derived from a more primitive principle and not validated outside the examples.
  • ad hoc to paper McCormick inequalities plus largest-eigenvector rank-1 projection yield a usable approximation of the robust bilinear program.
    Standard relaxation technique (A.2) but tightness is not proved; used as the meta-solver for all subsequent claims.
  • standard math Existence of a fixed point of the PR-MRE best-response operators.
    Assumed via the usual compact-convex best-response arguments for finite normal-form games; not re-proved.
invented entities (3)
  • Probabilistically Robust Minimax-Regret Equilibrium (PR-MRE)
    purpose: Interpolate between risk-neutral BNE and fully distribution-free MRE by minimizing weighted regret over high-confidence type subsets.
    Core new solution concept; independent evidence is limited to the synthetic matrices and one Graph CtF seed.
  • Typicality-preserving threat model / set S(μ̄,δ)
    purpose: Define the class of strategic distribution shifts against which the robustness certificate is claimed.
    Postulated geometry of admissible shifts; not derived from empirical deception data.
  • PRMRE-PSRO meta-solver
    purpose: Embed the SDP relaxation inside double-oracle / PSRO with MARL best responses.
    Algorithmic vehicle for the concept; no external re-implementation yet.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond Bayesian Nash: Learning Minimax-Regret Equilibria for Adversarial Team Games under Asymmetric Information." pith.science (2026). https://pith.science/paper/LHKDBEUG

@misc{pith2026260709993,
  author       = {Pith},
  title        = {Pith review of: Beyond Bayesian Nash: Learning Minimax-Regret Equilibria for Adversarial Team Games under Asymmetric Information},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LHKDBEUG}},
  note         = {Machine review of arXiv:2607.09993}
}
read the original abstract

Adversarial team games (ATGs) with asymmetric information, such as adversarial path-finding, goal search, and reachability games on graphs, require strategies that are robust to hidden opponent types, such as a hidden goal flag, and to deception. Under asymmetric information, deception is seen as strategic shifts in the type distribution such that the omniscient opponent can collude with Nature and condition its play on the observed type. Existing risk-neutral solution concepts, such as Bayesian Nash equilibrium (BNE), are sensitive to distribution shifts, while distributionally robust approaches provide guarantees only within a prescribed ambiguity set. To address these limitations, we introduce Probabilistically Robust Minimax-Regret Equilibrium (PR-MRE), a novel equilibrium concept that combines the distribution-free robustness of minimax-regret reasoning with probabilistic information from a nominal type distribution. PR-MRE minimizes worst-case regret over a high-confidence subset of the type space, providing protection against strategic redistribution of probability mass while avoiding the conservatism of fully distribution-free approaches. We show that, for normal-form Bayesian games, PR-MRE can be formulated as a robust bilinear program and derive a tractable semidefinite relaxation. We then adapt this relaxation into a novel meta-solver within a robust double-oracle framework, PRMRE-PSRO, enabling population-based learning of approximate PR-MRE strategies via deep reinforcement learning best responses. Experiments on graph-structured adversarial team games demonstrate that PR-MRE discovers strategies with substantially improved worst-case performance across hidden types compared to risk-neutral equilibrium solutions, resulting in more robust behavior under strategic distribution shifts.

Figures

Figures reproduced from arXiv: 2607.09993 by the authors.

Figure 1
Figure 1. Graph-Structured Capture-the-Flag (Graph CtF) with Blue Team’s Imperfect Information about [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Graph Capture-the-Flag with two Red flag hypotheses – [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Illustration of PR-MRE, nominal distribution [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Comparison of PRMRE (δ = 0.15) with BNE, DRO-Ball, MaxMin, DRO-MaxMin and MRE for the synthetic three-type game in [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]
Figure 5
Figure 5. Figure 5: Comparison of PRMRE (δ = 0.15) with BNE, DRO-Ball, MaxMin, DRO-MaxMin and MRE for the synthetic three-type game in [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: PRMRE-PSRO on the Graph CtF game under asymmetric information with nominal type distribu [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]
Figure 7
Figure 7. Figure 7: BNE-PSRO on the Graph CtF game under asymmetric information with nominal type distribution [PITH_FULL_IMAGE:figures/full_fig_p020_7.png]
Figure 8
Figure 8. Figure 8: Performance-Robustness Curve for PRMRE − and BNE −trained Blue team policies for a nominal distribution of (µ(θ0), µ(θ1)) = (0.8, 0.2). The x-axis on the above plot shows the probability mass for the hypothesis θ0 under the test distribution, and the y-axis denotes the…
Figure 9
Figure 9. Figure 9: Behavioral comparison of BNE-PSRO (left) and PRMRE-PSRO (right) under the minority flag [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Comparison of PRMRE (δ = 0.15) with BNE, DRO-Ball, MaxMin, DRO-MaxMin and MRE for the synthetic three-type game in [PITH_FULL_IMAGE:figures/full_fig_p028_10.png]
Figure 11
Figure 11. Figure 11: Comparison of PRMRE (δ = 0.15) with BNE, DRO-Ball, MaxMin, DRO-MaxMin and MRE for the synthetic three-type game in [PITH_FULL_IMAGE:figures/full_fig_p029_11.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

51 extracted references · 7 canonical work pages

  1. [1]

    Leslie Lamport , title =

  2. [2]

    The Knowledge Engineering Review , volume =

    Intelligent Agents: Theory and Practice , author =. The Knowledge Engineering Review , volume =

  3. [3]

    Artificial Intelligence , volume =

    Collaborative Plans for Complex Group Action , author =. Artificial Intelligence , volume =

  4. [4]

    Logics of programs: axiomatics and descriptive power

    David Harel. Logics of programs: axiomatics and descriptive power. 1978

    1978

  5. [5]

    Clarkson

    Kenneth L. Clarkson. Algorithms for Closest-Point Problems (Computational Geometry). 1985

    1985

  6. [6]

    A More Perfect Union

    Barack Obama. A More Perfect Union. 2008

    2008

  7. [7]

    The fountain of youth

    Joseph Scientist. The fountain of youth. 2009

    2009

  8. [8]

    Sam Anzaroot and Andrew McCallum , title =

Show all 51 references
  1. [9]

    Proceedings of the 20th International Colloquium on Automata, Languages and Programming , series =

    Maintaining Discrete Probability Distributions Optimally , author =. Proceedings of the 20th International Colloquium on Automata, Languages and Programming , series =. 1993 , publisher =

    1993

  2. [10]

    Donald E. Knuth. The Art of Computer Programming, Vol. 1: Fundamental Algorithms. 1997

    1997

  3. [11]

    Anisi , title =

    David A. Anisi , title =

  4. [12]

    SIAM review , volume=

    Guaranteed minimum-rank solutions of linear matrix equations via nuclear norm minimization , author=. SIAM review , volume=. 2010 , publisher=

    2010

  5. [13]

    INFORMS J

    Ahmadi, Amir Ali and Zhang, Jeffrey , title =. INFORMS J. on Computing , month = may, pages =. 2021 , issue_date =. doi:10.1287/ijoc.2020.0960 , abstract =

    2021 doi

  6. [14]

    Ahmadi, Amir Ali and Hall, Georgina , title =. Math. Oper. Res. , month = nov, pages =. 2019 , issue_date =. doi:10.1287/moor.2018.0962 , abstract =

    2019 doi

  7. [15]

    Chen, Tong and Lasserre, Jean-Bernard and Magron, Victor and Pauwels, Edouard , title =. Comput. Optim. Appl. , month = jan, pages =. 2022 , issue_date =. doi:10.1007/s10589-021-00325-z , abstract =

    2022 doi

  8. [16]

    Matrix rank minimization with applications , author=

  9. [17]

    arXiv preprint arXiv:2403.02227 , year=

    Policy space response oracles: A survey , author=. arXiv preprint arXiv:2403.02227 , year=

    arXiv

  10. [18]

    A unified game-theoretic approach to multiagent reinforcement learning , year =

    Lanctot, Marc and Zambaldi, Vinicius and Gruslys, Audr\=. A unified game-theoretic approach to multiagent reinforcement learning , year =. Proceedings of the 31st International Conference on Neural Information Processing Systems , pages =

  11. [19]

    Albrecht and Filippos Christianos and Lukas Sch\"afer , title =

    Stefano V. Albrecht and Filippos Christianos and Lukas Sch\"afer , title =. 2024 , url =

    2024

  12. [20]

    Algorithms for optimization , author=

  13. [21]

    , title =

    Harsanyi, John C. , title =. Management Science , volume =. 1967 , doi =. https://doi.org/10.1287/mnsc.14.3.159 , abstract =

    1967 doi

  14. [22]

    Bayesian

    John C. Harsanyi , journal =. Games with Incomplete Information Played by "Bayesian" Players, I-III. Part II. Bayesian Equilibrium Points , urldate =

  15. [23]

    Aghassi, Michele and Bertsimas, Dimitris , title =. Math. Program. , month = jun, pages =. 2006 , issue_date =

    2006

  16. [24]

    arXiv preprint arXiv:2507.15325 , year=

    Strategically Robust Game Theory via Optimal Transport , author=. arXiv preprint arXiv:2507.15325 , year=

    arXiv

  17. [25]

    Proceedings of the 34th International Conference on Machine Learning , pages =

    Deep Decentralized Multi-task Multi-Agent Reinforcement Learning under Partial Observability , author =. Proceedings of the 34th International Conference on Machine Learning , pages =. 2017 , editor =

    2017

  18. [26]

    Implicit Ensemble Training for Efficient and Robust Multiagent Reinforcement Learning , author=

  19. [27]

    2025 , url =

    Zhengdao Shao and Liansheng Zhuang and Yihong Huang and Houqiang Li and Shafei Wang , title =. 2025 , url =. doi:10.1109/TNNLS.2024.3457509 , timestamp =

    2025 doi

  20. [28]

    IEEE Transactions on Neural Networks and Learning Systems , keywords =

    Distributed Policy Space Response Oracles in Two-Player Zero-Sum Games. IEEE Transactions on Neural Networks and Learning Systems , keywords =. doi:10.1109/TNNLS.2025.3550827 , adsurl =

    2025 doi

  21. [29]

    Proceedings of the 20th Conference on Uncertainty in Artificial Intelligence , pages =

    Hyafil, Nathanael and Boutilier, Craig , title =. Proceedings of the 20th Conference on Uncertainty in Artificial Intelligence , pages =. 2004 , isbn =

    2004

  22. [30]

    2025 , eprint=

    Optimism as Risk-Seeking in Multi-Agent Reinforcement Learning , author=. 2025 , eprint=

    2025

  23. [31]

    IEEE Transactions on Intelligent Transportation Systems , volume=

    Risk-averse equilibria for vehicle navigation in stochastic congestion games , author=. IEEE Transactions on Intelligent Transportation Systems , volume=. 2022 , publisher=

    2022

  24. [32]

    IEEE Transactions on Automatic Control , volume=

    Dynamic games with asymmetric information: Common information based perfect bayesian equilibria and sequential decomposition , author=. IEEE Transactions on Automatic Control , volume=. 2016 , publisher=

    2016

  25. [33]

    2026 , eprint=

    Computing Perfect Bayesian Equilibria, with Application to Empirical Game-Theoretic Analysis , author=. 2026 , eprint=

    2026

  26. [34]

    Extensive-form game solving via blackwell approachability on treeplexes , year =

    Chakrabarti, Darshan and Grand-Cl\'. Extensive-form game solving via blackwell approachability on treeplexes , year =. Proceedings of the 38th International Conference on Neural Information Processing Systems , articleno =

  27. [35]

    2022 , eprint=

    A Marriage between Adversarial Team Games and 2-player Games: Enabling Abstractions, No-regret Learning, and Subgame Solving , author=. 2022 , eprint=

    2022

  28. [36]

    Proceedings of the 37th International Conference on Neural Information Processing Systems , articleno =

    McAleer, Stephen and Farina, Gabriele and Zhou, Gaoyue and Wang, Mingzhi and Yang, Yaodong and Sandholm, Tuomas , title =. Proceedings of the 37th International Conference on Neural Information Processing Systems , articleno =. 2023 , publisher =

    2023

  29. [37]

    International Conference on Learning Representations (ICLR) , year =

    Reevaluating Policy Gradient Methods for Imperfect-Information Games , author =. International Conference on Learning Representations (ICLR) , year =

  30. [38]

    Advances in neural information processing systems , volume=

    The surprising effectiveness of ppo in cooperative multi-agent games , author=. Advances in neural information processing systems , volume=

  31. [39]

    Proceedings of the 20th international conference on machine learning (ICML-03) , pages=

    Planning in the presence of cost functions controlled by an adversary , author=. Proceedings of the 20th international conference on machine learning (ICML-03) , pages=

  32. [40]

    Journal of Economic Theory , volume=

    Minimax regret and strategic uncertainty , author=. Journal of Economic Theory , volume=. 2010 , publisher=

    2010

  33. [41]

    , title =

    Renou, Ludovic and Schlag, Karl H. , title =. Games and Economic Behavior , volume =. 2011 , issn =. doi:10.1016/j.geb.2010.05.010 , url =

    2011 doi

  34. [42]

    Games and Economic Behavior , volume =

    Anagnostides, Ioannis and Kalogiannis, Fivos and Panageas, Ioannis and Vlatakis-Gkaragkounis, Emmanouil-Vasileios and Mcaleer, Stephen , title =. Games and Economic Behavior , volume =. 2026 , doi =

    2026

  35. [43]

    Proceedings of the 38th International Conference on Neural Information Processing Systems , articleno =

    Kalogiannis, Fivos and Yan, Jingming and Panageas, Ioannis , title =. Proceedings of the 38th International Conference on Neural Information Processing Systems , articleno =. 2024 , isbn =

    2024

  36. [44]

    Expert Systems with Applications , volume =

    Xiao, Jinheng and Qiu, Chen and Xu, Yingying and Zhang, Jiajia and Qi, Shuhan and Wang, Xuan , title =. Expert Systems with Applications , volume =. 2026 , issn =. doi:10.1016/j.eswa.2025.129496 , url =

    2026 doi

  37. [45]

    2024 , editor =

    Zhang, Youzhi and An, Bo and Zeng, Daniel Dajun , booktitle =. 2024 , editor =

    2024

  38. [46]

    Computational results for extensive-form adversarial team games

    Celli, Andrea and Gatti, Nicola. Computational results for extensive-form adversarial team games. Proc. Conf. AAAI Artif. Intell

  39. [47]

    Proceedings of the 39th International Conference on Machine Learning , pages =

    A Marriage between Adversarial Team Games and 2-player Games: Enabling Abstractions, No-regret Learning, and Subgame Solving , author =. Proceedings of the 39th International Conference on Machine Learning , pages =. 2022 , editor =

    2022

  40. [48]

    International Conference on Learning Representations (ICLR) , year=

    Efficiently Computing Nash Equilibria in Adversarial Team Markov Games , author=. International Conference on Learning Representations (ICLR) , year=

  41. [49]

    Advances in neural information processing systems , volume=

    Regret minimization in games with incomplete information , author=. Advances in neural information processing systems , volume=

  42. [50]

    and McAleer, Stephen and Yang, Yaodong and Wang, Jun , title =

    Slumbers, Oliver and Mguni, David Henry and Blumberg, Stefano B. and McAleer, Stephen and Yang, Yaodong and Wang, Jun , title =. Proceedings of the 40th International Conference on Machine Learning , articleno =. 2023 , publisher =

    2023

  43. [51]

    Advances in Neural Information Processing Systems , volume=

    Team-PSRO for Learning Approximate TMECor in Large Team Games via Cooperative Reinforcement Learning , author=. Advances in Neural Information Processing Systems , volume=

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.