B are bargaining with neighbors, wi

B are bargaining with neighbors, with imitate-the-best-neighbor and imitate-
with-probability-proportional-to-success dynamics respectively. The results are
barely distinguishable. C and D are the random neighborhood models corre-
sponding to A and B respectively. These results are much closer to those given
for the replicator dynamics in table 1. The dramatic dierence in convergence
to fair division between our two models is due to the structure of interaction
with neighbors.
7