Abstract:Matrix games are the most basic model in game theory, and yet robustness with respect to small perturbations of the matrix entries is not fully understood. In this paper, we introduce value positivity and uniform value positivity, two properties that refine the notion of optimality in the context of polynomially perturbed matrix games. The first concept captures how the value depends on the perturbation parameter, and the second consists of the existence of a fixed strategy that guarantees the value of the unperturbed matrix game for every sufficiently small positive parameter. We provide polynomial-time algorithms to check whether a polynomially perturbed matrix game satisfies these properties. We further provide the functional form for a parameterized optimal strategy and the value function. Finally, we translate our results to linear programming and stochastic games, where value positivity is related to the existence of robust solutions. Funding: This research was supported by Fondation CFM pour la Recherche, the H2020 European Research Council [Grant ERC-CoG-863818 (ForM-SMArt)], the Austrian Science Fund [Grant 10.55776/COE12], ANID Chile [Grant ACT210005], and Agence Nationale de la Recherche [Grant ANR-21-CE40-0020].

New characterizations and a concept of potential for each multinomial (probabilistic) value

Some Notes on Possibilistic Randomisation with T-norm Based Joint Distributions in Strategic-form Games

A Functional Inequality and a New Class of Probabilities in the N-Person Red-and-Black Game

Random partitions, potential, value, and externalities

Characterising Probabilistic Alternating Simulation for Concurrent Games

An exposition of possibility and probability

Non-commutative probability, joint distributions, conditioning, and the associated polymorphisms

Learning Mixed Multinomial Logits with Provable Guarantees

On Characterizations of Potential and Ordinal Potential Games

Testing high-dimensional multinomials with applications to text analysis

Uniqueness of Convex-Ranged Probabilities and Applications to Risk Measures and Games

A generalization of the symmetrical and optimal probability-to-possibility transformations

Probabilistic values for simplicial complexes

Grothendieck-type characterization of representable multilinear mappings

Value-Positivity for Matrix Games

Necessary players and values

Ultrametric and Tree Potential

A Note on Equivalent Conditions for Majorization

Negative Probability

Multiplicative weights in monotropic games

Explaining Probabilistic Models with Distributional Values