Towards Efficacy-Centered Game Design Patterns For Brain Injury Rehabilitation: A Data-Driven Approach
Jinghui Cheng, Cynthia Putnam, Doris C. Rusch · 2015 · ASSETS '15: Proceedings of the 17th International ACM SIGACCESS Conference on Computers & Accessibility · doi:10.1145/2700648.2809856
Summary
This paper presents a data-driven methodology for generating game design patterns specific to brain injury (BI) rehabilitation, addressing a communication gap between game designers and therapists. Brain injury affects approximately 6.4 million US children and adults, causing impairments in gross and fine motor control, problem solving, memory, learning, and speech. Rehabilitation typically involves repetitive therapeutic activities, and games are used to motivate patient engagement. However, commercial off-the-shelf (COTS) games often prove too fast-paced, provide inappropriate negative feedback, or lack the adjustability needed for diverse patient abilities. Purpose-built rehabilitation games exist but frequently fail to be engaging, partly due to limited collaboration between game designers and clinical experts. The authors propose design patterns—originating from Christopher Alexander's architectural work—as a common language to bridge this divide. Unlike previous rehabilitation game research that relied on subjective speculation, this study grounds pattern generation in empirical data from actual therapy sessions. Researchers conducted paper-based diary studies with 16 therapists at Schwab and Marianjoy Rehabilitation Hospitals over two-week periods, then expanded data collection through digital diary forms during beta testing of a game recommendation tool with 29 additional therapists. This process yielded 566 game therapy cases involving 89 patients (ages 19-95) and 112 different games. Statistical analysis revealed correlations among therapeutic goals, identifying four goal groups that tend to be addressed together: (1) dynamic balance, endurance, and standing; (2) dynamic balance, standing, and weight shifting; (3) turn taking and socialization; (4) turn taking and verbal expression. The most frequently targeted goals were dynamic balance (350 cases), attention/concentration (319), standing (291), and endurance (242). Notably, game use was highly concentrated: the top ten games appeared in 49% of all cases.
Key findings
The analysis produced 14 efficacy-centered game design patterns organized into four categories. Physical mechanics patterns (6) include Change Hands for bilateral hand use, Integrated Standing Duration, Moving Different Body Parts to reveal insight into deficits, Self-paced Weight Shifting for dynamic balance, Weight Shifting to the Extremes, and Fine Control. Game rules patterns (3) include Minimalist Task for attention/concentration, Optimal/Adjustable Pace for processing speed, and Randomized Events for hand-eye coordination. Perception patterns (2) include Step by Step for sequencing and command following, and Focus and Distraction for standing endurance. Social patterns (2) include Collocated Multiplayer for socialization and Turn-based Multiplayer for turn taking. Each pattern contains a name, category, associated therapeutic goals, problem statement, solution, example games, and related patterns. Critical findings emerged about why certain games succeed or fail. Self-paced Weight Shifting explains why Wii Fit Soccer Heading was rated ineffective for dynamic balance: it requires weight shifting at a fixed pace, but BI patients have limited reaction times and need to shift weight at their own speed. Wii Fit Table Tilt and Penguin Slide, which allow self-paced play, were rated highly effective. Randomized Events addresses hand-eye coordination by introducing unpredictability, but requires adjustable AI or human opponents to match patient abilities. Focus and Distraction solves the problem that standing exercises require considerable effort from BI patients—games must be engaging enough to "distract" players from the physical challenge. The patterns also exhibit two relationship types: synergistic (patterns combine for multiple goals) and facilitative (one pattern promotes another). For example, Collocated Multiplayer facilitates Focus and Distraction by creating engaging social contexts, while Randomized Events and Optimal/Adjustable Pace work together to ensure unpredictable gameplay matches patient reaction capabilities.
Relevance
This work provides a replicable methodology for generating domain-specific game design patterns from empirical therapy data, applicable beyond brain injury to other rehabilitation contexts. The finding that games requiring fixed-pace interactions are often therapeutically ineffective has direct implications for accessible game design generally—adaptive pacing benefits not only BI patients but anyone with motor or cognitive differences affecting reaction time. The pattern library offers concrete guidance for game designers entering rehabilitation game development, reducing the communication barrier with clinical experts. For accessibility practitioners, the emphasis on adjustability emerges as a consistent theme: effective rehabilitation games allow self-paced play, adjustable difficulty, and multiple input modalities. The observation that social patterns (Collocated Multiplayer) create "safe socialization spaces" for patients struggling with disability-related social challenges suggests games can address psychosocial rehabilitation alongside physical goals. The authors explicitly note these patterns should be treated as a creative toolkit rather than rigid guidelines—rehabilitation games must also be engaging experiences, requiring attention to player motivation beyond therapeutic efficacy.
Tags: brain injury · rehabilitation · game design · serious games · design patterns · physical therapy · cognitive rehabilitation · game-based therapy