Network coevolution drives segregation and enhances Pareto optimal equilibrium selection in coordination games

Area of Science:

• Agent-based modeling
• Network science
• Game theory

Background:

• Coordination games involve agents selecting actions to maximize payoffs.
• Network structure can influence emergent behavior and equilibrium selection.
• Dynamical adaptation of interaction networks is crucial for understanding complex systems.

Purpose of the Study:

• To investigate how network structure dynamics affect global coordination and equilibrium selection in coordination games.
• To analyze a coevolution model where agents' actions and network connections adapt simultaneously.
• To explore the impact of 'network plasticity' on system-wide coordination outcomes.

Main Methods:

• Developed a coevolution model coupling agent action updates (Replicator Dynamics and Unconditional Imitation) with network rewiring.
• Introduced 'network plasticity' (rewiring probability p) to quantify the rate of network adaptation.
• Studied Pure Coordination Games (PCG) and General Coordination Games (GCG) with varying plasticity levels.

Main Results:

• A transition was observed from full coordination on a single component to fragmentation into two components with different coordinated actions as plasticity increased.
• The nature of this fragmentation transition varied depending on the action update rules (RD vs. UI).
• For General Coordination Games, intermediate plasticity levels facilitated full coordination on the payoff-dominant action, enhancing equilibrium selection.

Conclusions:

• Network plasticity plays a critical role in both system fragmentation and the selection of specific equilibria.
• Coevolutionary dynamics, particularly with intermediate network plasticity, can promote the selection of the payoff-dominant equilibrium in General Coordination Games.
• The interplay between agent strategy updates and network topology is essential for understanding emergent coordination patterns.