Imitation Learning to Inform the Design of Computer Games

This paper was accepted for publication at the IEEE Conference on Games 2023 in Boston, USA.

This paper was also accepted for publication and presentation at the Human-in-the-Loop Learning (HiLL) Workshop at NeurIPS 2022.

Authors: Alessandro Sestini (SEED), Joakim Bergdahl (SEED), Konrad Tollmar (SEED), Andrew D. Bagdanov (Universita degli Studi di Firenze), Linus Gisslén (SEED).

Towards Informed Design and Validation Assistance in Computer Games Using Imitation Learning

Download the full research paper. (8 MB PDF)

Design validation and testing of games is a huge challenge, and as systems grow in size, manual testing is becoming infeasible.

This paper proposes a new approach to automated game validation and testing. Our method leverages a data-driven imitation learning technique that requires little effort and time on part of developers, and no specialist knowledge of machine learning or programming.

We investigate the validity of our approach through a user study with industry experts. The survey results show that our method is indeed a valid approach to game validation and that data-driven programming would be a useful aid to increasing the quality of modern playtesting while reducing the effort involved.

The survey also highlights several open challenges. With the help of the most recent literature, we analyze the identified challenges and propose future research directions suitable for supporting and maximizing the utility of our approach.