Final Adaptation Reinforcement Learning for N-Player Games