Policy gradient approach for learning of soccer player agents: Pass selection of midfielders

This research develops a learning method for the pass selection problem of midfielders in RoboCup Soccer Simulation games. A policy gradient method is applied as a learning method to solve this problem because it can easily represent the various heuristics of pass selection in a policy function. We implement the learning function in the midfielders' programs of a well-known team, UvA Trilearn Base 2003. Experimental results show that our method effectively achieves clever pass selection by midfielders in full games. Moreover, in this method's framework, dribbling is learned as a pass technique, in essence to and from the passer itself. It is also shown that the improvement in pass selection by our learning helps to make a team much stronger.