What: This project is an exploration into the ability of a game to change and adapt in response to the player as they are playing the game. This study looks into the use of preference classification and prediction algorithms (which are traditionally used by YouTube or Netflix) within the context of a digital game. We are exploring the possibility of a system like this allowing a game to procedurally generate content that is specifically and empirically tailored towards the player.

Why: Traditional digital games are part of a life cycle of player interest in that players disengage with the experience over time. Procedural Content Generation is a concept that automatically generates new and varied content for the player to directly counteract this life cycle, and to extend the engagement levels of the player. Another technique used to extend the life of a game is the increase in customisation or individualisation in the game, which allows for a wide range of different ways to play the same game. If this customisation was coupled with procedural content generation techniques, it could generate and personalise components of the game specifically towards you and how you play. This theoretical system is being investigated in attempts to extend player engagement, immersion and interest whilst also preserving the coherent nature of hand-crafted games. This would allow players to experience only aspects of the game that they enjoy, and not need to suffer through boring parts of the game.

How: To investigate this, the team will develop a prototype game using the Unity game engine that has the ability to turn the system described above on and off. This is important as we will validate this implementation with user testing, comparing the same game prototype with and without the system.

Team

• Padraic Heaton
• Dr William Raffe
• Dr Jaime Garcia