Beyond heatmaps: Spatio-temporal clustering using behavior-based partitioning of game levels

Abstract

Evaluating the spatial behavior of players allows for comparing design intent with emergent behavior. However, spatial analytics for game development is still in its infancy and current analysis mostly relies on aggregate visualizations such as heatmaps. In this paper, we propose the use of advanced spatial clustering techniques to evaluate player behavior. In particular, we consider the use of DEDICOM and DESICOM, two techniques that operate on asymmetric spatial similarity matrices and can simultaneously uncover preferred locations and likely transitions between them. Our results highlight the ability of asymmetric techniques to partition game maps into meaningful areas and to retain information about player movements between these areas

Inside the Group: Investigating Social Structures in Player Groups and Their Influence on Activity

“…Normoyle and Jensen [39] introduced Bayesian Clustering to behavioral profiling in games, drawing on data from Battlefield 3. Bauckhage et al [40] introduced spatiotemporal clustering and developed waypoint graphs that permitted behavioralbased partitioning of game maps. Drachen et al [41] developed behavioral profiles for Destiny, comparing four different cluster models.…”

The trails of Just Cause 2

“…With specific regard to analyzing spatio-temporal data, Drachen et al [12] clustered players according to spatio-temporal behavior and distance between players, while Rioult et al [26] used topological measures to predict esports match outcomes. Bauckhage et al [3] adopted the DEDICOM model [6,31,34] to cluster players of Quake: Arena and develop waypoint graphs for behaviorbased partitioning. However, other behavioral information was not integrated in the analysis (further details of DEDICOM is provided below).…”

“…Jointly, profiling helps build an understanding of the users. However, behavioral profiling in digital games is not a straightforward task due to the shifting requirements of a profiling exercise, common high-dimensionality in the data, volatility and the lack of clear guidelines for which types of behavioral features to incorporate into profiles [2,3,10,14].…”

“…Since OWGs allow players to move freely instead of limiting their trajectory to set paths, this increases the complexity of making data-driven calculations as each trajectory can be theoretically unique. The profiling technique presented here significantly expands the applications of the Decomposition into Directional Components (DEDICOM) framework adapted to games by Bauckhage et al [3], a spatial analysis technique that produces more detailed information than aggregate heatmaps. While Bauckhage et al [3] focused on player trails, this study integrates other dimensions of player behavior, such as characterizing player types by the actions performed during play.…”

“…The profiling technique presented here significantly expands the applications of the Decomposition into Directional Components (DEDICOM) framework adapted to games by Bauckhage et al [3], a spatial analysis technique that produces more detailed information than aggregate heatmaps. While Bauckhage et al [3] focused on player trails, this study integrates other dimensions of player behavior, such as characterizing player types by the actions performed during play. Furthermore, the DEDICOM technique has only been applied to game worlds with linear levels and shorter periods of play (a Quake III demo and an Unreal Tournament 2003 level), while the application is here focused on the unconstrained play of OWGs across the entire playing history of the players.…”