This paper addresses the question of whether neural networks, a realistic cognitive model of the human information processing, can learn to backward induce in a two stage game with a unique subgame-perfect Nash Equilibrium. The result that the neural networks only learn a heuristic that...