User Experience of Visualizations in Motion: A Case Study and Design Considerations

TL;DR

This paper addresses how visualizations in motion embedded with moving referents affect viewer experience in dynamic tasks, using video games as a testbed. It combines a systematic review of $160$ in-motion visualizations from $50$ games, an empirical study in the RobotLife FPS prototype, and an initial set of design considerations for motion-embedded visualizations. Key findings reveal trade-offs among readability, aesthetics, task support, and immersion, with conventional, non-integrated designs often favored for glanceability, while fully integrated designs offer immersion at readability costs; a $66\%$ threshold is used to distinguish 'evil' from 'good' robots in the study. The work provides practical guidelines and a research agenda for designing visualizations in motion across contexts beyond games.

Abstract

We present a systematic review, an empirical study, and a first set of considerations for designing visualizations in motion, derived from a concrete scenario in which these visualizations were used to support a primary task. In practice, when viewers are confronted with embedded visualizations, they often have to focus on a primary task and can only quickly glance at a visualization showing rich, often dynamically updated, information. As such, the visualizations must be designed so as not to distract from the primary task, while at the same time being readable and useful for aiding the primary task. For example, in games, players who are engaged in a battle have to look at their enemies but also read the remaining health of their own game character from the health bar over their character's head. Many trade-offs are possible in the design of embedded visualizations in such dynamic scenarios, which we explore in-depth in this paper with a focus on user experience. We use video games as an example of an application context with a rich existing set of visualizations in motion. We begin our work with a systematic review of in-game visualizations in motion. Next, we conduct an empirical user study to investigate how different embedded visualizations in motion designs impact user experience. We conclude with a set of considerations and trade-offs for designing visualizations in motion more broadly as derived from what we learned about video games. All supplemental materials of this paper are available at https://osf.io/3v8wm/}.

Figures (7)

• Figure 1: An overview of the dimensions analyzed in our systematic review on in-game visualizations in motion.
• Figure 2: Examples of visualizations in motion embedded in different locations with respect to the data referents: (a) Stamina bars and names embedded under the feet of the players in NBA 2K21NBA_2K21; (b) Heatmap overlapping the game environment in Civilization VICivilization6; (c) Health counter with the character's suite design in Nintendo Landnintendo_land.
• Figure 3: Distributions of collected data visualized under motion per visual representation (left) and per game genre (right).
• Figure 4: A screenshot of RobotLife's UI. The player's health and ammunition count are shown at the bottom in red and white bar charts respectively. Primary and secondary objectives of the game are displayed on the left of the screen, together with a pause button.
• Figure 5: Left: Health visualizations considered using different embedding locations, colors, and types of encoding. Other design variations are easily possible for Rows 2 and 3 by using ideas from Row 1. These were not explicitly drawn out. Right: Three visualization designs illustrated in RobotLife. (a) Non-integrated design: a horizontal bar chart positioned outside of the robot; (b) Fully-integrated design: a vertical bar chart integrated into the texture of the robot; and (c) Partial-match design: a circular bar (donut) matching a part of the robot's shape.