Churn Prediction in MMORPGs: A Social Influence Based Approach

Massively Multiplayer Online Role Playing Games(MMORPGs) are computer based games in which players interactwith one another in the virtual world. Worldwide revenuesfor MMORPGs have seen amazing growth in last few years and itis more than a 2 billion dollars industry as per current estimates.Huge amount of revenue potential has attracted several gamingcompanies to launch online role playing games. One of the majorproblems these companies suffer apart from fierce competitionis erosion of their customer base. Churn is a big problem for thegaming companies as churners impact negatively in the wordof-mouth reports for potential and existing customers leading tofurther erosion of user base.We study the problem of player churn in the popularMMORPG EverQuest II. The problem of churn predictionhas been studied extensively in the past in various domainsand social network analysis has recently been applied to theproblem to understand the effects of the strength of social tiesand the structure and dynamics of a social network in churn.In this paper, we propose a churn prediction model based onexamining social influence among players and their personalengagement in the game. We hypothesize that social influence is avector quantity, with components negative influence and positiveinfluence. We propose a modified diffusion model to propagatethe influence vector in the players network which represents thesocial influence on the player from his network. We measure aplayers personal engagement based on his activity patterns anduse it in the modified diffusion model and churn prediction. Ourmethod for churn prediction which combines social influenceand player engagement factors has shown to improve predictionaccuracy significantly for our dataset as compared to predictionusing the conventional diffusion model or the player engagementfactor, thus validating our hypothesis that combination of boththese factors could lead to a more accurate churn prediction.