This paper proposes a model of learning about a game. Players initially have little knowledge about the game. Through playing the identical game repeatedly, each player not only learns which action to choose but also constructs his personal view on the game. The model is studied using a hybrid payoff matrix of the prisoner's dilemma and coordination games. Results of computer simulations show (1) when all the players are slow in learning the game, they have only a partial understanding of the game, but may enjoy higher payoffs than the cases with full or no understanding of the game; (2) when one of the players is quick in learning the game, he obtains a higher payoff than the others. However, all of them can receive lower payoffs than the case where all the players are slow learners.