A pilot study on estimating players dispositional profiles from game traces analysis

Abstract

This work is a part of a project concerned with providing immersive and realistic experiences in video and serious games. Among other questions, the project relates to finding the best approach to represent and analyze game logs (traces) in order to improve the player experience through, for example, adaptive game flow. Understanding the player’s profile is one key point towards adaptive, personalized and immersive experiences. To this end, this paper aims to present a non-intrusive approach to calculate the player’s profile by analyzing his/her game traces (indirect profiling). This calculation is based on in-game information like the current situation and the player interactions with non-player characters. The representation of the player’s profile is based on a social-psychological model called the dispositional approach using 4 dimensions: time, control, motivation and relations. Pilot experiments were run with registered game traces of 15 participants playing a serious game on change management in professional environments. Calculated players’ profiles from traces analysis are then compared with the profiles issued from a dispositional profiling questionnaire and a post-game interview with a social expert. This first experiment shows that applying a trace-based approach for player profiling can be efficient. It also highlights that when game instructions and objectives are specific and precise, they can cause a deviation of the player behavior from his/her real-life behavior and personality.