Monte Carlo Tree Search Experiments in Hearthstone

André Santos, Pedro A. Santos, Francisco S. Melo

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

Abstract

In this paper, we introduce a Monte-Carlo tree search (MCTS) approach for the game “Hearthstone: Heroes of Warcraft”. We argue that, in light of the challenges posed by the game (such as uncertainty and hidden information), Monte Carlo tree search offers an appealing alternative to existing AI players. Additionally, by enriching MCTS with a properly constructed heuristic, it is possible to introduce significant gains in performance.We illustrate through extensive empirical validation the superior performance of our approach against vanilla MCTS and the current state-of-the art AI for Hearthstone.
Original languageEnglish
Title of host publication2017 IEEE Conference on Computational Intelligence and Games (CIG)
Place of PublicationNew York, NY
PublisherIEEE
Pages272-279
ISBN (Electronic)978-1-5386-3233-8
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event2017 Conference on Computational Intelligence in Games - New York University, New York, United States
Duration: 22 Aug 201725 Aug 2017
http://www.cig2017.com/

Conference

Conference2017 Conference on Computational Intelligence in Games
Abbreviated titleCiG2017
CountryUnited States
CityNew York
Period22/08/1725/08/17
Internet address

Fingerprint

Experiments
Uncertainty

Keywords

  • Monte Carlo Tree Search
  • Artificial intelligence for games
  • Hearthstone

Cite this

Santos, A., Santos, P. A., & Melo, F. S. (2017). Monte Carlo Tree Search Experiments in Hearthstone. In 2017 IEEE Conference on Computational Intelligence and Games (CIG) (pp. 272-279). New York, NY: IEEE. https://doi.org/10.1109/CIG.2017.8080446
Santos, André ; Santos, Pedro A. ; Melo, Francisco S. / Monte Carlo Tree Search Experiments in Hearthstone. 2017 IEEE Conference on Computational Intelligence and Games (CIG). New York, NY : IEEE, 2017. pp. 272-279
@inproceedings{e13a3c8af26e41b7b41f772b95e0a0b7,
title = "Monte Carlo Tree Search Experiments in Hearthstone",
abstract = "In this paper, we introduce a Monte-Carlo tree search (MCTS) approach for the game “Hearthstone: Heroes of Warcraft”. We argue that, in light of the challenges posed by the game (such as uncertainty and hidden information), Monte Carlo tree search offers an appealing alternative to existing AI players. Additionally, by enriching MCTS with a properly constructed heuristic, it is possible to introduce significant gains in performance.We illustrate through extensive empirical validation the superior performance of our approach against vanilla MCTS and the current state-of-the art AI for Hearthstone.",
keywords = "Monte Carlo Tree Search, Artificial intelligence for games, Hearthstone",
author = "Andr{\'e} Santos and Santos, {Pedro A.} and Melo, {Francisco S.}",
year = "2017",
doi = "10.1109/CIG.2017.8080446",
language = "English",
pages = "272--279",
booktitle = "2017 IEEE Conference on Computational Intelligence and Games (CIG)",
publisher = "IEEE",
address = "United States",

}

Santos, A, Santos, PA & Melo, FS 2017, Monte Carlo Tree Search Experiments in Hearthstone. in 2017 IEEE Conference on Computational Intelligence and Games (CIG). IEEE, New York, NY, pp. 272-279, 2017 Conference on Computational Intelligence in Games, New York, United States, 22/08/17. https://doi.org/10.1109/CIG.2017.8080446

Monte Carlo Tree Search Experiments in Hearthstone. / Santos, André; Santos, Pedro A.; Melo, Francisco S.

2017 IEEE Conference on Computational Intelligence and Games (CIG). New York, NY : IEEE, 2017. p. 272-279.

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

TY - GEN

T1 - Monte Carlo Tree Search Experiments in Hearthstone

AU - Santos, André

AU - Santos, Pedro A.

AU - Melo, Francisco S.

PY - 2017

Y1 - 2017

N2 - In this paper, we introduce a Monte-Carlo tree search (MCTS) approach for the game “Hearthstone: Heroes of Warcraft”. We argue that, in light of the challenges posed by the game (such as uncertainty and hidden information), Monte Carlo tree search offers an appealing alternative to existing AI players. Additionally, by enriching MCTS with a properly constructed heuristic, it is possible to introduce significant gains in performance.We illustrate through extensive empirical validation the superior performance of our approach against vanilla MCTS and the current state-of-the art AI for Hearthstone.

AB - In this paper, we introduce a Monte-Carlo tree search (MCTS) approach for the game “Hearthstone: Heroes of Warcraft”. We argue that, in light of the challenges posed by the game (such as uncertainty and hidden information), Monte Carlo tree search offers an appealing alternative to existing AI players. Additionally, by enriching MCTS with a properly constructed heuristic, it is possible to introduce significant gains in performance.We illustrate through extensive empirical validation the superior performance of our approach against vanilla MCTS and the current state-of-the art AI for Hearthstone.

KW - Monte Carlo Tree Search

KW - Artificial intelligence for games

KW - Hearthstone

U2 - 10.1109/CIG.2017.8080446

DO - 10.1109/CIG.2017.8080446

M3 - Conference article in proceeding

SP - 272

EP - 279

BT - 2017 IEEE Conference on Computational Intelligence and Games (CIG)

PB - IEEE

CY - New York, NY

ER -

Santos A, Santos PA, Melo FS. Monte Carlo Tree Search Experiments in Hearthstone. In 2017 IEEE Conference on Computational Intelligence and Games (CIG). New York, NY: IEEE. 2017. p. 272-279 https://doi.org/10.1109/CIG.2017.8080446