Overview of the Chapter

One of the most effective methods for exploratory analysis is visualization. Game data analysis and communication of results of such analysis aren’t easy processes. The sequential and contextual nature of gameplay actions makes it hard to understand user behavior and infer the user’s goals, intent, plans, and preferences. In this chapter, we’ll discuss how visualization can be used to allow analysts to make such inferences. However, to be able to do this, the visualization system itself needs to be capable of handling multidimensional spaces. Therefore, when discussing how visualization can be used as part of the analysis process, we’ll also discuss visualization systems and tools that have been proposed and are currently used for this purpose.

Library used for visualization in R

Throughout this course, we used different types of visualization tools to communicate analysis or compare results. Previously, we used bar charts and line graphs to show data distributions. In later chapters, we introduced the TraMineR package, which enabled us to see sequential and temporal data through charts and colored sequences or actions. In this chapter, we’ll discuss other tools and systems and their uses. In particular, it should be noted that we’ll focus our discussion here on visualization tools and systems that are used for the analysis of player behaviors. We will not focus on communicating aggregate results or distributions of data since these were covered in other chapters.

With this focus, we'll discuss the following:

• Heatmaps: Heatmaps are aggregate (where variables are aggregated over time, that is, total kills over time) visualizations on top of game maps that use color-coding to show high-density regions in the map. The visualized variables can be kills, deaths, gold collection, or other variables of interest to the designer/analyst.

• Spatio-temporal visualization systems: Spatio-temporal visualization systems are systems that show game actions over time overlaid on the game map. They usually allow designers and analysts to zoom into a specific time window or play a specific segment within the game to better understand sequences of behaviors within a game for multiple players across multiple playthroughs.

• State-action transition visualization systems: State-action transitions are basic elements of problem-solving behaviors, showing a game state and an action that a player or group of players took. Using such visualizations, analysts and designers can more deeply understand the problem-solving strategies that their players took or where players failed to deduce design problems in terms of puzzles or mechanics.