Decoding Player Behavior: Analyzing Reasons for Player Quitting Using Log Data from Puzzle Game Baba Is You

In this paper, we study the reasons for unsuccessful level completion in Baba is You, a puzzle-based video game, using Epistemic Network Analysis (ENA). The study focuses on student cognition, which can be inferred through an in-depth examination of in-game actions and decisions recorded in log data from complex, contextualized game levels. To build epistemic networks around video game log data, chronological log records of player levels were set as codes. Epistemic networks of player actions paired with interpretive examinations of the context of each level offer insights into why students may quit levels in Baba is You. Findings suggest that (1) inadequate acquisition of knowledge from the previous level, (2) premature focus on winning at the earlier stage without engaging in distributed exploration and experimentation, and (3) over-reliance on undoing actions may all play a role in unsuccessful level completion. The goal of this work is to support the design of future game-based interventions that can address context-specific quitting and foster student engagement within the game.