Zero Sum Games
Thread and Full Text View
Fred, every steady state of a replicator dynamics is Nash and vice-versa. When we ask whether the stability of a steady state leads to a refinement of Nash, then every ESS is an asymptotically stable steady state, though the converse is [View full text and thread]
|02/11/2001 11:46 AM by Rodrigo; ESS|
I have been working on lattice based models of evolutionary games. The learning rules are always switch-if-better (still an undergrad so anything else heap big complicated) when I run simple coordination games with the coordination payoffs are equal I get a chaotic fluctuation around (1/2,1/2) from a wide set of initial conditions. Solving the replicator dynamics indicates that either pure strategy should be an ESS. What is going on? Does the use of random walkers on a lattice deviate in its predictions constantly from the asymtotically stable solutions in the equations? I appreciate any time anybody has.
Fred [Manage messages]