The Research of Dynamic Difficulty Adjustment to Improve Sticky Factor of Novice Players in Fighting Games Based on Monte Carlo Tree Search

碩士 === 國立宜蘭大學 === 資訊工程學系碩士班 === 107 === With the development of the game industry, all kinds of games appear constantly. For novice players who play the new game, a problem which they encounter easily is having a prompt failure by strong opponents or computer’s artificial intelligences(AIs). Taking...

Full description

Bibliographic Details
Main Authors: CHEN, WEN-CHIEN, 陳文謙
Other Authors: WU, TIN-YU
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/78xys4
Description
Summary:碩士 === 國立宜蘭大學 === 資訊工程學系碩士班 === 107 === With the development of the game industry, all kinds of games appear constantly. For novice players who play the new game, a problem which they encounter easily is having a prompt failure by strong opponents or computer’s artificial intelligences(AIs). Taking the fighting game mentioned in this study as an example. In the past, most opponent’s AIs are designed to win the game. They didn’t think the player's psychological condition. In the study, we believe the flow theory is a good method to evaluate the attraction of games after we think about the player's psychological condition. The flow is a mental condition. It indicates a person immerses in favorite activity. In the study, we used the fighting game as a platform and proposed an AI which is designed by Monte Carlo Tree Search (MCTS) to control the character in order to make players immerse in the game by the action of AI, and then investigate player’s mental condition when they play the game with different opponents. Through researching, we discovered players have a better experience and sticky factor after they have an equal confrontation with the opponent in the game. On the contrary, players have a worse experience and sticky factor after they play the game with a powerful opponent.