Mathematical Biology Seminar: Human strategy updating in a spatial game

  • Date: 06/10/2010
Arne Traulsen (Max-Planck Institute for Evolutionary Biology)

University of British Columbia


Probably the most thoroughly studied mechanism that can explain the evolution and maintenance of costly cooperation among selfish individual is population structure. In the past years, hundreds of papers have mathematically modeled how cooperation can emerge under various dynamical rules and in more and more complex population structures [1,2]. However, so far there is a significant lack of experimental data in this field. Milinski et al. have conducted an experimental test to address how humans are playing a particularly simple spatial game on a regular lattice [2]. The data shows that the way humans choose strategies is different from the usual assumptions of theoretical models. Most importantly, spontaneous strategy changes corresponding to mutations or exploration behavior is more frequent than assumed in many models. This can strongly affect evolutionary dynamics [4] and decrease the influence of some spatial structures.


This experimental approach to measure properties of the update mechanisms used in theoretical models may be useful for mathematical models of evolutionary games in structured populations.

[1] Ohtsuki, Hauert, Lieberman, and Nowak, Nature (2006)
[1] Szabo and Fath, Evolutionary games on graphs, Physics Reports (2007)
[3] Traulsen, Semmann, Sommerfeld, Krambeck, and Milinski, PNAS (2010)
v[4] Traulsen, Hauert, De Silva, Nowak, and Sigmund, PNAS (2009)


2:00pm - 3:00pm, WMAX 110