Date of Award
Daniel Felix Ritchie School of Engineering and Computer Science, Computer Science
Chris J. GauthierDickey
Constraint satisfaction, Dynamic difficulty adjustment (DDA), Game development, Machine learning, Reinforcement learning, Video games
Video games are an incredibly popular pastime enjoyed by people of all ages world wide. Many different kinds of games exist, but most games feature some elements of the player overcoming some challenge, usually through gameplay. These challenges are insurmountable for some people and may turn them off to video games as a pastime. Games can be made more accessible to players of little skill and/or experience through the use of Dynamic Difficulty Adjustment (DDA) systems that adjust the difficulty of the game in response to the player’s performance. This research seeks to establish the effectiveness of machine learning techniques for use in developing DDA systems in real-time games for the purpose of making games more accessible and therefore more popular.
A first-person platforming game where the player dodges attacks from AI-controlled turrets was created to gauge player’s perceptions of difficulty and measure how much their perceptions could be changed through the use of machine-learning backed DDA systems. Twenty players played through four similar game environments, some with DDA systems present and some without, and answered survey questions about their perception of the games’ difficulty between each game.
While the players had varying levels of skill and experience with video games, all players showed a decrease in perception of difficulty and an increase in performance in the game environments where DDA systems were present. Additionally, cost benefit is demonstrated. The workflows used in developing the experimental game and the DDA systems within were found to be feasible in terms of the added development burden required, as compared to the quality of the results produced.
These findings support the suitability of machine learning techniques, specifically reinforcement-learning-trained neural networks (RLNN), in the development of effective DDA systems. Consequently, this research serves as a baseline which justifies the use and development of machine-learning-controlled DDA systems in video games and videogame- adjacent applications.
Copyright Statement / License for Reuse
All Rights Reserved.
Copyright is held by the author. User is responsible for all copyright compliance.
Ryan Adare Dunagan
Received from ProQuest
Dunagan, Ryan Adare, "An Investigation into Machine Learning Techniques for Designing Dynamic Difficulty Agents in Real-Time Games" (2023). Electronic Theses and Dissertations. 2275.