Evaluating Interactive Entertainment Using Breakdown: Understanding Embodied Learning in Video Games

This study examines how players learn video games through embodied interaction, using the concept of "breakdown" as its central lens. Breakdowns occur when a player's everyday strategies for interacting with a game fail, forcing either an adaptation of behavior (breakdown in interaction) or a loss of immersion in the game world (breakdown in illusion). Drawing on talk-aloud protocols and interviews with 13 Indiana University participants, eight male and five female, who played unfamiliar games over sessions of roughly three hours, the authors identify breakdowns through observable signs such as surprise, erratic movement, and frustration.

The observational work yields a taxonomy of 17 breakdown patterns organized into four categories: perceiving the environment, developing strategy, taking action, and meaning making. Games examined include Myst, SSX 3, Sly 2, and Indigo Prophecy, spanning several genres. The authors find that many breakdowns stem from unclear environmental cues, poorly delivered or sequenced instructions, mismatched control mappings, and disconnects between tutorials and core gameplay. Notably, some breakdowns produce accidental discoveries rather than frustration, suggesting that failure can serve a constructive learning function.

From these patterns the authors derive 17 design guidelines, one corresponding to each pattern, grouped into areas such as cues and affordances, camera and navigation, and task instruction. The guidelines draw on HCI heuristics and are checked against game design practice. The central conclusion is that learnability can be evaluated heuristically by balancing immersion, which guards against breakdowns in illusion, with intentional opportunities for breakdown that enable players to generalize what they have learned.

The study's scope is limited to a small, single-institution sample and qualitative observation. It offers no quantitative data, and the guidelines have not been empirically tested. The authors explicitly leave implementation and validation of the framework to future research.