Building Placement Optimization in Real-Time Strategy Games

摘要: In this paper we propose using a Genetic Algorithm to optimize the placement of buildings in Real-Time Strategy games. Candidate solutions are evaluated by running base assault simulations. We present experimental results SparCraft — StarCraft combat simulator --- battle setups extracted from human and bot show that our system is able turn assaults losses for defenders into wins, as well reduce number surviving attackers. Performance heavily dependent on quality prediction attacker army composition used training, its similarity evaluation. These apply both