A Review of Multi-Armed Bandit Algorithms in Player Modeling and Game Design

This paper explores the application of multi-armed bandit algorithms (MAB) in game design, focusing on player modeling and game optimization. The effectiveness of multi-armed bandit algorithms in modeling player characteristics such as skill level, play style, and social comparison orientation is in...

Full description

Saved in:
Bibliographic Details
Main Author: Zhan Xizhi
Format: Article
Language:English
Published: EDP Sciences 2025-01-01
Series:ITM Web of Conferences
Online Access:https://www.itm-conferences.org/articles/itmconf/pdf/2025/04/itmconf_iwadi2024_02016.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper explores the application of multi-armed bandit algorithms (MAB) in game design, focusing on player modeling and game optimization. The effectiveness of multi-armed bandit algorithms in modeling player characteristics such as skill level, play style, and social comparison orientation is investigated. The potential of MAB in optimizing game design elements like difficulty, rewards, and user interface is also explored. The paper presents empirical results from simulations and user studies and concludes by discussing the potential of MAB algorithms in game design and highlighting future research directions.
ISSN:2271-2097