Facilitating Learning in a Real Time Strategy Computer Game

This research explores the development and implementation of a just-in-time (JIT) help system designed to improve user learning in real-time strategy (RTS) computer games. Traditional game tutorials and manual-based help systems often suffer from rigidity, lack of retention, and the potential to cause learning stagnation. The proposed system aims to overcome these limitations by monitoring player actions and delivering context-sensitive hints at optimal moments, thereby supporting the acquisition of motor, perceptual, declarative, and strategic skills.

The study involved the development of a custom RTS game, "Tubby Terror," which served as the research platform. The help system utilized a knowledge inference engine to estimate a player's mastery of specific game concepts and a scheduler to prioritize hint delivery based on game state and timing. Testing was conducted in two phases, each involving six first-year psychology students. Participants played the game for fifteen minutes while receiving aural hints, which they evaluated as either helpful or unhelpful.

Results indicated no statistically significant difference in the perceived helpfulness of the system between the two phases, with the ratio of helpful responses declining from 62% to 54%. The analysis suggests that the small sample size limited the ability to generalize findings or effectively refine the system's weighting parameters. However, the study identified emerging patterns when grouping hints into categories. Key conclusions emphasize that future iterations must account for varying player skill levels, distinguish between immediate and long-term player goals, and potentially organize hints into hierarchical structures rather than linear lists. The research concludes that while the current implementation requires further refinement, the JIT approach remains a viable method for enhancing player engagement and learning in complex gaming environments.