At Ubicomp2005, there was a workshop about pervasive games (the website seems to be down). An interesting analysis of the pervasive game Mogi Mogi has been presented by Benjamin Joffe (from the french company Newt Games). The results of the field study are quite pertinent. From his presentation, I like two things. The first one is that as he says "Ubiquity does not necessarily mean “the same service on all devices”, which is so true, the richness of ubiquitous application might emerge from the complementarity between services (be it games or other applications).

The second one is that though the game was mobile and location-based, most of the game was played at home. Of course it refers to different part of the gameplay (trading and sending messages) but that's an important component.