Development and Evaluation of a Multi-Agent Reinforcement Learning Model for Doppelkopf