Virtual labs in evolutionary game theory

by Christoph Hauert, Version 2.3, January 2005.


Evolutionary dynamics on graphs

Tutorial on the fixation probability of mutants in structured populations where individuals are arranged on a graph. For a large class of graphs, the fixation probability does not depend on the details of the population structure and is identical to a homogenous population. All these graphs display the same characteristic balance between evolutionary selection and random drift. Nevertheless, the structure of the graph can have significant effects on the fixation probability ranging from complete suppression of selection to guaranteed fixation of advantageous mutants. This complements a research article in Nature (with Erez Lieberman & Martin Nowak, 2005, 433, 312-316).


Origin of cooperators and defectors

Tutorial on the gradual evolution of distinct cooperative and defective behavioral patterns through evolutionary branching into separate trait groups characterized by high and low cooperative investments. This is based on a model that extends the classical Snowdrift game to continuously varying degrees of cooperation. Apart from evolutionary branching, this model exhibits rich dynamics that can be easily explored using this interactive tutorial. This complements a research article in Science (with Michael Doebeli & Timothy Killingback, 2004, 306, 859-862).


Cooperation in structured populations

Interactive tutorial on the effect of spatial structure in Prisoner's Dilemma and Snowdrift games. This complements a research article in Nature (with Michael Doebeli, 2004, 428 643-646). All reported results can be explored through a virtual lab (Java applet). The fascinating spatio-temporal patterns are both entertaining as well as of scientific value in order to improve the intuition and understanding of the spatial dynamics.

Spatial 2×2 games

Tutorial on 2×2 games in populations with rigid spatial structures. The games are played on regular lattices with different geometries and the players interact only with their nearest neighbors. The dynamics of this system can be explored through a virtual lab implemented as a Java applet. This allows for hands-on experiences of the fascinating dynamical world of spatio-temporal patterns.

Volunteering in public goods games

Tutorial on voluntary participation in public goods games. Most theoretical and experimental work on social dilemmas such as the free-rider problem, the tragedy of the commons or the prisoner's dilemma has tacitly built on the fact that the interacting individuals are actually caught in the dilemma. In most real-life examples, however, individuals do have considerable freedom to choose their partners. To explore the dynamics of this system in both, well-mixed populations as well as populations with rigid spatial structures, we provide further virtual labs. They demonstrate the persistent rock-sissors-paper-type cycles of cooperators, defectors and so called loners in mixed populations. The same mechanism acts as the driving force for travelling waves and other fascinating spatio-temporal patterns in populations arranged on a lattice.

Reward, Punishment and Reputation in public goods games

Tutorial on effects of reward, punishment and reputation in public goods games. Various experimental studies on social dilemmas have shown that punishment is very efficient in creating incentives for cooperative behavior. Reward, however, is considerably less efficient. The underlying mechanisms are illustrated with a simple game theoretical model.


Revision history

January 2005:
Version 2.3 released - addition of the virtual lab on evolutionary dynamics on graphs complementing a research article in Nature.
October 2004:
Version 2.2 released - addition of the virtual lab on the origin of cooperators and defectors complementing a research article in Science.
April 2004:
Version 2.1 released - addition of the virtual lab on effects of spatial structure in Snowdrift games complementing a research article in Nature.
March 2003:
Version 2.0 released - new labs added as well as a complete rewrite and update of all existing labs. Upon retirement of version 1.0 of the virtuallabs, they have attracted more than 3100 visitors.
April 2002:
Version 1.0 of Virtuallab on Public Goods Games released.
January 2002:
Version 1.0 of Virtuallabs on Spatial 2x2 Games released.

Last updated on Thursday, January 20, 2005 by Christoph Hauert.