- Our replication has been implemented using RepastJ 3.1. The source code is released under the GNU public license.
- The data archive contains all datasets generated for the analysis presented in the article. For each candidate model there is a unique file containing the focal measures named numcooperators and numdefectors, respectively, (and a lot of other measures) sampled at t=500 along with the parameters used for that candidate model and the values of the random seeds (column rngseed) for each of the 200 runs per model. The data files are named according to the model-number (as assigned in the result tables in the appendix) and the setting number (1 and 2, respectively).