Global
average gain as a function of . The results
for the partially altruistic case
(solid curve) are always better than those for the naΓ―ve (random) case (dotted) or the selfish case (circled). The latter shows a maximum at some and is better than the naΓ―ve case only when is large enough. (The
results are the average of 30 samples of random initial strategies. For each
sample, we ran time steps and
collected the data of the last time steps for the average.)