Characterization of user behavior in a multi-player online game

Abstract

In recent years, multi-player online games (MOGs) have gained enormous popularity and become a major trend in the entertainment industry. Much research has been focusing on improving the performance and scalability of MOG systems. However, relatively little attention has been paid to the study of user behavior. As with other complex interactive applications, a good understanding of user behavior is important to the design of MOG systems. In this paper, we discuss the advantages of a user behavior workload model to the software industry and the research community, and describe a method to develop such a model. For illustrative purposes, we study user behavior by using the measurement data collected from an existing MOG system. This includes the characterization of the interarrival time of logon users, the transition probability of avatars within the virtual environment, the residence time at a room, and the session length. Our results not only provide an insight into user behavior in MOGs, but they are also useful in the development of workload models in performance studies of MOG systems.