Designerly Ways of Analysing Gameplay and Player Experiences

This paper argues that player experience is difficult for game designers to shape directly, since designers control the rules that produce fun rather than fun itself. Its central thesis is that existing methods, particularly delayed self-report and think-aloud protocols borrowed from digital-game research, distort the understanding of board games. The authors propose an event-contingent experience sampling protocol, grounded in Christensen's decision criteria, in which the researcher plays alongside participants and captures in-the-moment experiences tied to specific rules, thereby reducing interpretive leaps and recall bias.

The method was piloted with the abstract strategy game Pylos, combining gameplay video and audio, experience sampling forms with Likert scales, mood meters and open questions, post-game elicitation, and rule-modification exercises with think-aloud protocols. Three players of differing expertise, novice, board gamer, and Pylos-experienced, took part in roughly 35-minute sessions, with either player able to trigger sampling so that fun is captured interactionally. The dataset comprised 105 minutes of gameplay, 72 experience samples, 153 minutes of interviews, and 45 minutes of verbal protocols. Descriptive coding and thematic analysis produced themes such as "creating ambiguity" and "creating a difference in choices," linking formal rule structures to felt experiences of fun.

The authors conclude that treating players as "player-designers" and situating data collection close to the gameplay context improves the accuracy of experience reconstruction and the researcher's grasp of game structure. The approach is positioned as a design-oriented analytic tool rather than a player-experience modelling technique of the kind typical in traditional Game User Research. The evidence is qualitative and drawn from a small pilot, so the findings are illustrative and the study's claims about triangulation and reduced bias remain to be tested at scale.