Difficulty Adjustment

Dynamic Difficulty Adjustment (DDA) aims to optimize player experience in games by adapting the challenge level to individual skill and engagement. Current research focuses on employing machine learning, particularly reinforcement learning and hybrid approaches combining techniques like Minimax and Monte Carlo Tree Search, to create AI opponents that dynamically adjust their playing strength based on player performance. This work is significant for enhancing user engagement in games and provides valuable insights into creating adaptive AI systems that personalize the difficulty of complex tasks, with implications for fields beyond gaming.

Papers

August 13, 2024

Personalized Dynamic Difficulty Adjustment -- Imitation Learning Meets Reinforcement Learning
Ronja Fuchs, Robin Gieseke, Alexander Dockhorn
Imitation Learning AI Agent Personalized Subject Difficulty Level Difficulty Adjustment

October 25, 2023

Hybrid Minimax-MCTS and Difficulty Adjustment for General Game Playing
Marco Antônio Athayde de Aguiar Vieira, Anderson Rocha Tavares, Renato Perez Ribas
Zero Sum Game Minimax McTs Game Play Board Game Difficulty Adjustment

August 24, 2023

Continuous Reinforcement Learning-based Dynamic Difficulty Adjustment in a Visual Working Memory Game
Masoud Rahimi, Hadi Moradi, Abdol-hossein Vahabie, Hamed Kebriaei
Reinforcement Learning Working Memory Difficulty Level Visual Reinforcement Difficulty Adjustment

August 4, 2022

Human Decision Makings on Curriculum Reinforcement Learning with Difficulty Adjustment
Yilei Zeng, Jiali Duan, Yang Li, Emilio Ferrara, Lerrel Pinto, C. -C. Jay Kuo, Stefanos Nikolaidis
Reinforcement Learning Human Decision Human Centered Curriculum Reinforcement Learning Large Scale Reinforcement Learning Reinforcement Learning Performance Difficulty Adjustment

November 11, 2021

AlphaDDA: Strategies for Adjusting the Playing Strength of a Fully Trained AlphaZero System to a Suitable Human Training Partner
Kazuhisa Fujita
Artificial Intelligence System General Strategy Game AI AlphaZero Based Artificial Intelligence Collaborative Training Single Shot Alpha MEPOL Difficulty Adjustment

Difficulty Adjustment

Papers

Personalized Dynamic Difficulty Adjustment -- Imitation Learning Meets Reinforcement Learning

Hybrid Minimax-MCTS and Difficulty Adjustment for General Game Playing

Continuous Reinforcement Learning-based Dynamic Difficulty Adjustment in a Visual Working Memory Game

Human Decision Makings on Curriculum Reinforcement Learning with Difficulty Adjustment

AlphaDDA: Strategies for Adjusting the Playing Strength of a Fully Trained AlphaZero System to a Suitable Human Training Partner