While the ball in the digital soccer is rolling, where the non-player characters should go in a defensive situation?

摘要

The non-player characters (NPCs), i.e. the artificial characters that are not under direct control by the user, are essential part of many digital games. Achieving the realistic behavior by NPCs in digital sports games such as the simulated soccer is challenging. Here we limit our scope to the defensive situation, i.e. when the ball is controlled by the opponents, and propose a systematic method for optimal NPC positioning. So far the methods for automatically finding defensive positions by the intelligent robotic soccer players have been investigated by some scholars within RoboCup, an international research and educational initiative in Artificial Intelligence and robotics. Although simulated soccer teams using these methods have proved to be reasonably good, the collaboration issue in defensive situations has been overlooked. In this paper we propose a systematic approach based on solving a multi-criteria assignment problem. This allows gracefully balancing the costs and rewards involved in defensive positioning to achieve better results.