Extracting game-play metric data from audio/video processing: A practical solution for game studies research

Abstract

Today videogame classification works to a set of guidelines initially designed for, and more conducive to linear mediums (film, television and literature). As a result, digital games receive an age restriction rating based on both their depiction of harmful content and its prospective impact on players. Not accounting for the medium’s interactive qualities means that envisaging the player’s experience and the nature and impact of interactions between players and game texts remains a largely inferential practice and an exercise in caution. Given the medium’s interactive nature, we argue that classification processes would be supported by research that provides empirical accounts of the interactive experience of games. In order to take into account the unique experiential properties of games, a mixed methodological approach located at an intersection between humanities, social sciences and computer science is being employed in a large-scale study of player experiences. This paper presents one of the earliest challenges for this study: the capacity to gather game-play metric data from game texts that are selected by participants as an ongoing process throughout the course of the study. With no real advance knowledge of game choices, access to source code or game developers, a pragmatic solution was required to capture player interactions with game texts. This paper presents the beginnings of a method of gathering game-play metrics through utilizing audio and video processing and its required synchronization with other forms of data output.