This thesis presents distribution strategies for pursuit evasion games of networked multi-agent systems. The strategies are designed for both obstacle-free and obstacle-cluttered environments, leveraging potential maps as a method. The effectiveness of the proposed strategies was eval- uated through simulation and analysis, and the result is that combining a potential map and position extrapolation for obstacle avoidance was very successful at producing competent autonomous agents, and very com- patible when combined with specifically tailored pursuer algorithms for seeking and capture