Visualizing and understanding players' behavior in video games: discovering patterns and supporting aggregation and comparison

As video games become more popular, there is an urge for procedures that can support the analysis and understanding of players' behaviors within game environments. Such data would inform game and level designers of game design issues that should be fixed or improved upon. By logging user-initiated events in video games, analysts have exhaustive information regarding players' actions within games. However, visualizing such data is a challenging task due to the amount of data one has to deal with; the necessity of a deep understanding of the game and players' possible actions within the game plus a deep understanding of questions one wants to answer; the computation that has to be done on the data; and the limitations and/or complexities of current analysis tools. In this paper, we present a new visualization system that allows analysts to build visualization and interact with telemetry data, to identify patterns and identify game design issues efficiently. Besides the system itself, we propose a new approach to visualize players' behavior that has not been explored so far. For example, instead of using heat maps to visualize a single metric (e.g. deaths), our system allows analysts to superimpose and visualize a series of actions players take in the game. This is especially important when one should understand cause and effect within the game. We present examples of the visualizations using an RPG game, Dragon Age Origins (BioWare/EA, 2009). It should be noted that the system is currently under development and testing with analysts working at BioWare.