Using Player Type Models for Personalized Game Design – An Empirical Investigation

TLDR

Personalized games should provide a better player experience than one‑size‑fits‑all games, and player type models have been proposed as a method for personalization, but their relationship to players’ experience of specific game mechanics has never been empirically investigated. The study investigates whether player type scores can reliably predict player experience and explores explanations and design strategies for personalized games based on player type models. We examined the predictive power of two player types (Mastermind and Seeker) from the BrainHex model in a field study with a mobile game prototype. The field study with 51 participants showed that player type scores did not significantly predict player experience of the corresponding game mechanics, indicating that player type models still need improvement.

Abstract

Personalized games should provide a better player experience than one-size-fits-all games. As a method for personalization, player type models have been discussed recently. Player type models would be useful tools in the personalization of games, if they have a relationship to the players’ experience of specific game mechanics. However, this relationship has never been empirically investigated. To close this gap, we examine whether player types—as a specific appearance of personality traits—can significantly and reliably predict player experience. We investigate the predictive power of two player types (Mastermind, Seeker) of the BrainHex player type model. Results of a field study (n = 51) with a mobile game prototype tailored to the two player types Mastermind and Seeker suggest that player type models still need improvement: Player type scores do not significantly predict player experience of according game mechanics. We discuss possible explanations and a way to design personalized games that adapt to users gaming preferences with player type models.