Limes Master Edition: Using Behavioural Data to Refine Game Mechanics in Research

Limes Master Edition, an educational game set in Roman history, serves as the focus of a methodological argument for using process-level behavioural data in experimental game-based learning research. The central thesis is that logging how learners interact with specific game features can reveal mechanisms that outcome measures alone cannot, and that such data should guide the iterative refinement of game design. The work situates Limes within an experimental research programme rather than presenting it as a finished commercial product.

The key empirical foundation lies in two prior experimental studies by Viccari and colleagues. Across both, engagement with optional Codex entries, measured through clicks and time spent, consistently predicted knowledge test performance even after controlling for prior knowledge and interest. Overall Codex use, by contrast, varied with how visible and integrated the feature was within the interface, indicating that design choices shape whether learners engage with supplementary content at all.

These findings motivated concrete changes in the Master Edition, including an animated Codex notification, an end-of-level prompt encouraging engagement, and entry-level logging that captures behaviour at a finer grain. The author argues that this redesign cycle demonstrates the value of pairing process-level behavioural traces with outcome measures in experimental designs. The approach is presented explicitly as a case for methodological integration, while acknowledging that evidence on educational games more broadly remains inconsistent.

The supporting literature spans learning analytics, educational data mining, and serious game design, including several studies on the Limes game addressing game format, instructional framing, and sound effects, as well as recent work on nudges in intelligent tutoring systems.