Understanding Player Behaviours in Third-Person Games

MSc (30HP) Topic (Individual or a Group)

Camera perspective in video games, teleoperation, and telepresence scenarios affects not only visual experience but also psychological framing, motor embodiment, and emotional identification. Third-person view allows encourages character-identification and allow player to build a bond with a character as an independent entity. In this setting, players normally control the character in a certain way to accomplish certain missions. The level, at which a character can act autonomously from the player can vary, depending on a multitude of factors.

Previous research demonstrated a possibility to identify the moment when the player is distracted (e.g. looking away) in a data-driven way, without knowing what the game task is. The current project will explore player’s behaviours beyond distractions and will investigate the third-person control situation from a perspective of autonomy adaptation between the player and the character.

Tasks and qualifications for the project:

1. Study relevant literature on adaptive autonomy, mixed initiative games, and other fields to define the experimental scenarios.
2. Develop a game environment in Unity game engine, including game levels, player mechanics and data collection scripts.
3. Develop a character behaviour mechanics with a small VLM, connected to the game engine.
4. Prepare and conduct preliminary experiment data collection, analyse the results.

The successful candidate(s) have to be fluent in Python or Unity, know basic principles of language models and/or game development.
Contact: Andrey Kiselev, andrey.kiselev@oru.se