Modeling players' behaviors in games has gained increased momentum in the past few years. This area of research has wide applications, including modeling learners and understanding player strategies, to mention a few. In this paper, we present a new methodology, called Interactive Behavior Analytics (IBA), comprised of two visualization systems, a labeling mechanism, and abstraction algorithms that use Dynamic Time Warping and clustering algorithms. The methodology is packaged in a seamless interface to facilitate knowledge discovery from game data. We demonstrate the use of this methodology with data from two multiplayer team-based games: Boom-Town, a game developed by Gallup, and DotA 2. The results of this work show the effectiveness of this method in modeling, and developing human-interpretable models of team and individual behavior.

Currently, there is no formal taxonomy for the activities that users engage in when interacting with and making meaning from spatio-temporal game data visualizations. As data visualization, especially spatio-temporal visualization, becomes more popular for game data analytics, it becomes increasingly crucial that we develop a formal understanding of how users, especially players, interact with and extract meaning from game data using these systems. However, existing taxonomies developed for InfoVis are not directly applicable due to domain differences and a lack of consensus within the literature. This paper presents the beginnings of a taxonomy for user interaction with spatio-temporal data specific to the domain of games, developed from the results of a qualitative user study (n=7) in which experienced players were tasked with using a spatio-temporal visualization system to explore and understand telemetry data from Defense of the Ancients 2 (DotA 2). The taxonomy includes seven activities organized into three categories: Data Interaction, Sense Making, and Validation. We discuss the implications of these activities on design and future research.

No abstract

No abstract

Many successful games rely heavily on data analytics to understand players and inform design. Popular methodologies focus on machine learning and statistical analysis of aggregated data. While effective in extracting information regarding player action, much of the context regarding when and how those actions occurred is lost. Qualitative methods allow researchers to examine context and derive meaningful explanations about the goals and motivations behind player behavior, but are difficult to scale. In this paper, we build on previous work by combining two existing methodologies: Interactive Behavior Analytics (IBA) [2] and sequence analysis (SA), in order to create a novel, mixed methods, human-in-the-loop data analysis methodology that uses behavioral labels and visualizations to allow analysts to examine player behavior in a way that is context sensitive, scalable, and generalizable. We present the methodology along with a case study demonstrating how it can be used to analyze behavioral patterns of teamwork in the popular multiplayer game Defense of the Ancients 2 (DotA 2). CCS CONCEPTS• Human-centered computing → Visualization.