Portal de Periódicos da CAPES

A BDI Game Master Agent for Computer Role-Playing Games

2017; Association for Computing Machinery; Volume: 15; Issue: 1 Linguagem: Inglês

10.1145/2629700

1544-3574

Bao Vo Luong, John Thangarajah, Fabio Zambetta,

Reinforcement Learning in Robotics

In this paper we describe an approach for developing an intelligent game master (GM) for computer role-playing games. The role of the GM is to set up the game environment, manage the narrative ow and enforce the game rules whilst keeping the players engaged. Our approach is to use the popular Belief-Desire-Intention (BDI) model of agents to developing a GM. We describe the process for creating such a GM and how we implemented a prototype of it for a scenario in the Neverwinter Nights (NWN) game. We describe the evaluation of our prototype with human participants who played the chosen NWN scenario both with and without the BDI GM. The comparison survey completed by the participants shows that the system with the BDI GM was the clear winner with respect to game replayability, flexibility, objective setting and overall interest; thus, validating our hypothesis that a BDI GM will provide game players with a better gaming experience.