| In this talk I will propose a simple anlytically tractable model for cooperation in self-organized networks. In contrast to previous modelling approaches, this model allows agents to maintain different levels of cooperation with different self-chosen partners. Strating from a homogeneous intitial state, leaders, who hold positions of high centrality and extract high payoffs, can emerge spontaneously in this model. However, the model exhibits also a high degree of coordination between agents, which can be understood mathematically. To derive these results a novel approach to stabiltity analysis is used that is applicable to heterogeous, weighted, and directed networks and yields stability conditions on network motifs on the meso-scale. Finally, I consider the situation where agents are added to the self organizing network one-by-one. This leads to the formation of realistic social networks exhibiting broad degree distributions, strong clustering, and community structure. Interestingly, there are certain moments of oportunity in the network growth when agents enter that subsequently become very successful. |
|