Mobile games have been around for quite a while now, but there are not that many games available that really take advantage of mobility. One of the main reasons is that creating games that are dependent on human behavior is hard. If we could predict how humans behave in certain situations it would be easy to create games for those situations. MIT has done research on human mobility and they have created methods for predicting human behavior. These methods could be utilized in mobile game design. Taking advantage of mobility could be made easy by integrating mobility into game patterns. In this paper we will describe human behavior models like serendipity and human landmarks, based on the results of MIT’s Reality Mining project. We will also study how these models could be integrated with social interaction game patterns. As a result we claim that creating games that truly take advantage of mobility would be possible, but further studies of human behavior models would still be needed.