When Contrast Is Not Enough: An Empirical Study of Dynamic Color Adaptation in FPS Games

We present a dynamic color assignment approach and an empirical study of context-adaptive color for fast-moving game visualizations. In first-person shooter (FPS) games, players often need to read visualizations that are attached to moving characters while making quick decisions. Health visualizations are particularly important because they help players identify targets and decide when to act. As characters move across changing backgrounds, maintaining sufficient contrast becomes a natural design problem. We implemented a real-time, background-adaptive color assignment approach in RobotLife, an open-source FPS game. Our approach samples the local background around a health visualization and dynamically adjusts its color to meet a WCAG-aligned contrast ratio. We evaluated our approach in a within-subjects study with 24 participants, comparing dynamic color assignment conditions and fixed color baselines across three health visualization designs. Our results show that, in RobotLife, dynamic-color assignment did not significantly improve readability or aesthetics for in-game visualizations under motion. Instead, participants prioritized glanceable readability and semantic consistency, revealing trade-offs among readability, aesthetics, and semantic stability between dynamic color assignment and the fixed-color baseline. All supplemental materials of our work are available at https://osf.io/csjrb/.