In this article we propose a game design approach to build context adaptive games. This approach is based on a model of the game structure and a generic adaptation model. Our method consists of designing different game scenarios involving different gameplay for the game then, game engine selects and proposes the appropriate one according to the context. We have conducted a pilot experiment in order to evaluate the effect of this adaptation on game and player satisfaction. We stated as hypotheses that context adaptation does not have effects on the game and player satisfaction. We have developed two versions of a mobile game and a virtual city to test these hypotheses. The results show that context adaptation increase the number of game objectives which are succeeded and decreases the number which are failed. However, player satisfaction is not always superior for the version that includes context adaptation.