A Data Analysis of Player in World of Warcraft Using Game Data Mining

Abstract

Online Games platforms have become a very prospective field of investigation to the game industry. Many business models are being created to understand the dynamics of player behavior to promote improvements in game design and to keep players in the virtual world so that they maintain and renew their respective subscription. From the business perspective it is crucial to predict how many players will join the game and how many will stay in the game, important factors for the company's revenue. This work aims to identify significant data from a popular Massively Multiplayer Online Game (MMOG) - World of Warcraft - using appropriate data mining methods for deduction of players profiles. Multiple linear regression was applied to check whether a player will leave or not the game in the near future. Also, a clustering technique, i.e. K-means, was used to extract clusters of players and identify their common characteristics. The regression model showed that "Level" and "Playing Density" contribute significantly to predict whether a player will or not renew the subscription, while the clustering revealed four forms of player profile: Beginner, Intermediate I, Intermediate II and Professional.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 12 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.