Is there Value in Reinforcement Learning?