Cohesive obstacle management for directed flocking in real-time strategy games

Scope and method of study. The purpose of this study was to introduce additional rules to directed flocking which allowed a flock to navigate obstacles while maintaining cohesion. These rules were chaining, memory, navigation markers, and dynamic leadership. Validation and performance tests were conducted to determine how effective the new rules were and how many units in a flock could be used in a real-time setting.; Findings and conclusions. The new rules were effective in allowing a flock to reach the destination while navigating obstacles. However, a few odd behaviors were observed during testing. The optimal flock size for performance considerations ranged from 25 to 50 units. Units allowed to catch up with the leader after falling behind could do so when bottleneck obstacles were absent, but catching up came at the cost of additional processing overhead. Performance deteriorated when the number of blocked units increased, which occurred when units were blocked by other units or by bottleneck obstacles.