This article investigates an evolutionary game based on the framework of
interacting particle systems. Each point of the square lattice is occupied by a
player who is characterized by one of two possible strategies and is attributed
a payoff based on her strategy, the strategy of her neighbors and a payoff
matrix. Following the traditional approach of evolutionary game theory, this
payoff is interpreted as a fitness: the dynamics of the system is derived by
thinking of positive payoffs as birth rates and the absolute value of negative
payoffs as death rates. The nonspatial mean-field approximation obtained un-
der the assumption that the population is well mixing is the popular replicator
equation. The main objective is to understand the consequences of the inclu-
sion of local interactions by investigating and comparing the phase diagrams
of the spatial and nonspatial models in the four dimensional space of the
payoff matrices. Our results indicate that the inclusion of local interactions
induces a reduction of the coexistence region of the replicator equation and
the presence of a dominant strategy that wins even when starting at arbitrar-
ily low density in the region where the replicator equation displays bistability.
We also discuss the implications of these results in the parameter regions that
correspond to the most popular games: the prisoner’s dilemma, the stag hunt
game, the hawk-dove game and the battle of the sexes.
%0 Journal Article
%1 lanchier2015evolutionary
%A Lanchier, N.
%D 2015
%I Institute of Mathematical Statistics
%J Ann. Appl. Probab.
%K cooperation evolutionary_games game_theory particle_systems spatial_structure
%N 3
%P 1108--1154
%R 10.1214/14-aap1018
%T Evolutionary games on the lattice: Payoffs affecting birth and death rates
%U http://dx.doi.org/10.1214/14-AAP1018
%V 25
%X This article investigates an evolutionary game based on the framework of
interacting particle systems. Each point of the square lattice is occupied by a
player who is characterized by one of two possible strategies and is attributed
a payoff based on her strategy, the strategy of her neighbors and a payoff
matrix. Following the traditional approach of evolutionary game theory, this
payoff is interpreted as a fitness: the dynamics of the system is derived by
thinking of positive payoffs as birth rates and the absolute value of negative
payoffs as death rates. The nonspatial mean-field approximation obtained un-
der the assumption that the population is well mixing is the popular replicator
equation. The main objective is to understand the consequences of the inclu-
sion of local interactions by investigating and comparing the phase diagrams
of the spatial and nonspatial models in the four dimensional space of the
payoff matrices. Our results indicate that the inclusion of local interactions
induces a reduction of the coexistence region of the replicator equation and
the presence of a dominant strategy that wins even when starting at arbitrar-
ily low density in the region where the replicator equation displays bistability.
We also discuss the implications of these results in the parameter regions that
correspond to the most popular games: the prisoner’s dilemma, the stag hunt
game, the hawk-dove game and the battle of the sexes.
@article{lanchier2015evolutionary,
abstract = {This article investigates an evolutionary game based on the framework of
interacting particle systems. Each point of the square lattice is occupied by a
player who is characterized by one of two possible strategies and is attributed
a payoff based on her strategy, the strategy of her neighbors and a payoff
matrix. Following the traditional approach of evolutionary game theory, this
payoff is interpreted as a fitness: the dynamics of the system is derived by
thinking of positive payoffs as birth rates and the absolute value of negative
payoffs as death rates. The nonspatial mean-field approximation obtained un-
der the assumption that the population is well mixing is the popular replicator
equation. The main objective is to understand the consequences of the inclu-
sion of local interactions by investigating and comparing the phase diagrams
of the spatial and nonspatial models in the four dimensional space of the
payoff matrices. Our results indicate that the inclusion of local interactions
induces a reduction of the coexistence region of the replicator equation and
the presence of a dominant strategy that wins even when starting at arbitrar-
ily low density in the region where the replicator equation displays bistability.
We also discuss the implications of these results in the parameter regions that
correspond to the most popular games: the prisoner’s dilemma, the stag hunt
game, the hawk-dove game and the battle of the sexes.},
added-at = {2016-05-18T17:36:57.000+0200},
author = {Lanchier, N.},
biburl = {https://www.bibsonomy.org/bibtex/25784930ecd7a526501ec334d92257c63/peter.ralph},
doi = {10.1214/14-aap1018},
interhash = {ca6df95a6e5a6c0c5ced5587f595a558},
intrahash = {5784930ecd7a526501ec334d92257c63},
journal = {Ann. Appl. Probab.},
keywords = {cooperation evolutionary_games game_theory particle_systems spatial_structure},
month = jun,
number = 3,
pages = {1108--1154},
publisher = {Institute of Mathematical Statistics},
timestamp = {2016-05-18T17:36:57.000+0200},
title = {Evolutionary games on the lattice: Payoffs affecting birth and death rates},
url = {http://dx.doi.org/10.1214/14-AAP1018},
volume = 25,
year = 2015
}