Identification of Groups in Online Environments: The Twist and Turns of Grouping Groups

This study describes an approach to track groups over time as they participate in larger networks of groups. Groups were tracked in a Massive Multi-Player Online Role Playing Game (MMORPG), a context typical of many online venues and similar to some situations in the real world. Based on social scientific theories the study identified possible avenues to detect groups, explored game logs to identify relevant variables for group detection, and developed an algorithm to detect groups from individual level data. This study demonstrates that it is possible to identify groups in MMORPG environments using available individual level data. The study has also captured some group dynamics that corresponds to existing real world examples, including changes in groups over time and development of larger groups through unification of smaller groups.