Data mining on user behaviors is a practice aimed at supporting the design and development of systems. When the data analyzed comes from games and gameful environments, the term used is Game Analytics.
Game Analytics combines statistics, data mining, machine learning, as well as data visualization and communication to generate insights on player behaviors and preferences to inform game designers and developers. Gamification telemetry data can be used to extreact players’ preferences, identify players’ strategies, monitor behavioral evolution, and predict churn. The acquired information is also fundamental to inform the customization and adaptation of the user experience (Adaptive Gamification).
Game Analytics results are also very relevant for Games User Research (GUR), since they allow to better understand players, investigate what motivates players, how they interact with the gamified system, and how to improve the user experience.
Within the Game Industry, the user behavior is analyzed considering two main perspectives: the customer perspective (retention, churn prediction, purchasing/spending behavior, etc.) and the player perspective (user experience, engagement, etc.). When analysing the user behavior in gamified motivational systems, the customer (or business) perspective needs to focus on the system ulterior motive, namely the system goal in terms of user motivation and/or behavior change (e.g. promote more sustainable transport habits, raise the user awareness towards a certain problem or topic).
Research problems we investigate in MoDiS include:
  • Player profiling and behavioral analysis. Monitoring users interactions with the gamified system can help detect and fix game design faults (e.g., is the gamification ulterior motive being pursued?), moderate abnormal behaviors ensuring fairness (e.g., identify and block cheaters) and profile players to generate personalized experiences. In MoDiS we conduct research on player modeling techniques, information extraction from telemetry data, and players’ behavioral analysis to gain insights on players’ experience, behavior and preferences.
  • Churn prediction: When gamification is used as persuasive technology, long-term engagement leads to internalizing the behavior in the own routine. Therefore, being able to detect churners promptly could help in preventing such abandonment by actuating contingency strategies. In MoDiS, we investigate the usage of players’ telemetry data describing in-game activity to train machine learning algorithms for churn prediction in gamified applications.
  • Social influence in gamified systems: Social connections shape our behaviour. This phenomenon is amplified in online networks by particularly influential individuals: influencers. Although this concept originated in social media, recent research shows how influencers also exist in games and affect players’ long-term retention. Prolonged retention caused by influencers could benefit gameful systems, especially if the system’s goal is positive behavioural change. In MoDiS, we exploit social network analysis techniques to investigate influencers’ presence within motivational gamified system and to analyze their impact on other players’ behaviours.

Some relevant publications:

  • Loria, Enrica; Lennart, Nacke; Marconi, Annapaola, On Social Contagion in Gamification: The Power of Influencers in a Location-Based Gameful System, Proceedings of ACM CHI PLAY Conference, 2021
  • Loria, Enrica; Marconi, Annapaola, Exploiting limited players’ behavioral data to predict churn in gamification, in Electronic Commerce Research and Applications, v. 47, 2021
  • Loria, Enrica; Rivera, Jessica; Marconi, Annapaola, Do they Play as Intended? – Comparing Aggregated and Temporal Behavioral Analysis in a Persuasive Gamified System, 54th Hawaii International Conference on System Sciences, 2021
  • Loria, Enrica; Pirker, Johanna; Drachen, Anders; Marconi, Annapaola, Do Influencers Influence? – Analyzing Players’ Activity in an Online Multiplayer Game, Proceedings of IEEE Conference on Games (CoG), 2020, pp. 120-127
  • Loria, Enrica; Marconi, Annapaola, Reading Between the Lines – Towards an Algorithm Exploiting In-game Behaviors to Learn Preferences in Gameful Systems, Proceedings of FDG ’20: International Conference on the Foundations of Digital Games, 2020, pp. 1-12
  • Loria, Enrica; Marconi, Annapaola, A Data-Driven Approach to Deduce Players’ Preferences from In-Game Interactions in Gameful Systems, Game Analytics Workshop 2019, 2019
  • Loria, Enrica; Paissan, Francesco; Marconi, Annapaola, Exploiting General-Purpose In-Game Behaviours to Predict Players Churn in Gameful Systems, Game Analytics Workshop 2019, 2019
  • Loria, Enrica; Marconi, Annapaola, Player Types and Player Behaviors: Analyzing Correlations in an On-the-field Gamified System, Proceedings of the 2018 Annual Symposium on Computer-Human Interaction in Play, ACM, 2018, pp. 531-538