Assessing Empathy Across Game Fidelity Levels: A Case Study of 3D and Text-Based Versions of Brukel

A 2025 study presented at DiGRA examines whether a low-fidelity, Twine-based text adaptation of the narrative game Brukel can elicit empathic responses comparable to its high-fidelity 3D original. Motivated by the cost advantages of text-based development and Twine's widespread use in prototyping narrative games, the authors randomly assigned 42 participants to one of the two versions and measured empathy using Shen's State Empathy Scale. The conceptual framework distinguishes cognitive, affective, and associative empathy, allowing the researchers to assess each dimension separately rather than treating empathy as a single construct.

Frequentist analyses found no significant differences between the versions on any empathy dimension, but equivalence tests (TOST) were also non-significant, so equivalence could not be confirmed under this approach. Bayesian analysis offered a more differentiated picture, providing moderate evidence of equivalence for cognitive empathy and only anecdotal evidence for affective and associative empathy. This pattern suggests that the text-based format may be particularly effective at fostering understanding of characters' perspectives, while its capacity to evoke emotional and associative responses remains less clear.

The authors conclude that visual fidelity is not essential for cognitive empathy in narrative games, supporting Twine as a low-cost prototyping tool and a viable final format where cognitive empathy is a central design goal. The study's scope is limited to a single game, a modest sample, brief exposure, and exclusively quantitative measures, which constrain generalizability. The authors identify open questions about how adaptations affect players' sense of a game's identity and call for research on player perceptions of different versions and on the longitudinal effects of empathic engagement.