Non-equilibrium steady states as game changers

20.04.2017 14:15 – 15:00

Predicting the long-term evolution of species and strategies in changing environments is a longstanding challenge in evolutionary dynamics. In evolutionary game theory, where the payoff a player receives is a function of its own strategy and the strategy of the others, the long-term evolution leads to one of the evolutionarily stable states. In the great majority of cases, this payoff function also depends on the environmental state, which may change over time. We generalise game theory to account for this. We find anomalous, sometimes counterintuitive, long-term behaviours which are markedly different from traditional games defined by constant payoffs. Intricately, these anomalous stationary states are sensitive to the covariance of the payoffs. In contrast to evolutionarily stable states of games with constant payoffs, where coexisting species necessarily receive equal payoffs, anomalous stable states can be unfair, meaning that, on average, two coexisting species may receive different payoffs. Moreover, environmental noise can induce transitions between different games. As a tool to reveal these transitions, we introduce a classification for evolutionary games with payoff stochasticity, which contains the traditional games for vanishing payoff variance. Our framework, developed here analytically, robustly predicts the long-term evolution of species and strategies in fluctuating environments.

Lieu

Bâtiment: Ecole de Physique

Salle 234, 24 quai Ernest-Ansermet

entrée libre