We analyze the traditional board game the Game of the Goose. We are particularly interested in the probabilities of the different players winning, and we show that we can determine these probabilities exactly for up to six players and using simulation for any number of players. Our original motivation to investigate this game came from progress in stochastic process theories, which prompted the question of whether such methods are capable of dealing with well-known probabilistic games. As these games have large state spaces, this is not trivial. As a side effect we find that some common wisdom about the game is not true.