By the same authors

From the same journal

From the same journal

Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering

Research output: Contribution to journalArticle

Standard

Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering. / Cowling, P.L.; Ward, C.D.; Powley, E.J.

In: Computational Intelligence and AI in Games, IEEE Transactions on, Vol. PP, No. 99, 6218176, 2012.

Research output: Contribution to journalArticle

Harvard

Cowling, PL, Ward, CD & Powley, EJ 2012, 'Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering', Computational Intelligence and AI in Games, IEEE Transactions on, vol. PP, no. 99, 6218176. https://doi.org/10.1109/TCIAIG.2012.2204883

APA

Cowling, P. L., Ward, C. D., & Powley, E. J. (2012). Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering. Computational Intelligence and AI in Games, IEEE Transactions on, PP(99), [6218176]. https://doi.org/10.1109/TCIAIG.2012.2204883

Vancouver

Cowling PL, Ward CD, Powley EJ. Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering. Computational Intelligence and AI in Games, IEEE Transactions on. 2012;PP(99). 6218176. https://doi.org/10.1109/TCIAIG.2012.2204883

Author

Cowling, P.L. ; Ward, C.D. ; Powley, E.J. / Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering. In: Computational Intelligence and AI in Games, IEEE Transactions on. 2012 ; Vol. PP, No. 99.

Bibtex - Download

@article{7bca0c33e50142d4bd92d01ef450813c,
title = "Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering",
abstract = "In this paper, we examine the use of Monte CarloTree Search (MCTS) for a variant of one of the most popular andprofitable games in the world: the card game Magic: The Gathering(M:TG). The game tree for M:TG has a range of distinctivefeatures, which we discuss here, and has incomplete informationthrough the opponent’s hidden cards, and randomness throughcard drawing from a shuffled deck. We investigate a wide rangeof approaches that use determinization, where all hidden andrandom information is assumed known to all players, alongsideMonte Carlo Tree Search. We consider a number of variationsto the rollout strategy using a range of levels of sophisticationand expert knowledge, and decaying reward to encourage playurgency. We examine the effect of utilising various pruningstrategies in order to increase the information gained from eachdeterminization, alongside methods that increase the relevance ofrandom choices. Additionally we deconstruct the move generationprocedure into a binary yes/no decision tree and apply MCTS tothis finer grained decision process. We compare our modificationsto a basic MCTS approach for Magic: The Gathering using fixeddecks, and show that significant improvements in playing strengthcan be obtained.",
author = "P.L. Cowling and C.D. Ward and E.J. Powley",
year = "2012",
doi = "10.1109/TCIAIG.2012.2204883",
language = "English",
volume = "PP",
journal = "Computational Intelligence and AI in Games, IEEE Transactions on",
issn = "1943-068X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "99",

}

RIS (suitable for import to EndNote) - Download

TY - JOUR

T1 - Ensemble Determinization in Monte Carlo Tree Search for the Imperfect Information Card Game Magic: The Gathering

AU - Cowling, P.L.

AU - Ward, C.D.

AU - Powley, E.J.

PY - 2012

Y1 - 2012

N2 - In this paper, we examine the use of Monte CarloTree Search (MCTS) for a variant of one of the most popular andprofitable games in the world: the card game Magic: The Gathering(M:TG). The game tree for M:TG has a range of distinctivefeatures, which we discuss here, and has incomplete informationthrough the opponent’s hidden cards, and randomness throughcard drawing from a shuffled deck. We investigate a wide rangeof approaches that use determinization, where all hidden andrandom information is assumed known to all players, alongsideMonte Carlo Tree Search. We consider a number of variationsto the rollout strategy using a range of levels of sophisticationand expert knowledge, and decaying reward to encourage playurgency. We examine the effect of utilising various pruningstrategies in order to increase the information gained from eachdeterminization, alongside methods that increase the relevance ofrandom choices. Additionally we deconstruct the move generationprocedure into a binary yes/no decision tree and apply MCTS tothis finer grained decision process. We compare our modificationsto a basic MCTS approach for Magic: The Gathering using fixeddecks, and show that significant improvements in playing strengthcan be obtained.

AB - In this paper, we examine the use of Monte CarloTree Search (MCTS) for a variant of one of the most popular andprofitable games in the world: the card game Magic: The Gathering(M:TG). The game tree for M:TG has a range of distinctivefeatures, which we discuss here, and has incomplete informationthrough the opponent’s hidden cards, and randomness throughcard drawing from a shuffled deck. We investigate a wide rangeof approaches that use determinization, where all hidden andrandom information is assumed known to all players, alongsideMonte Carlo Tree Search. We consider a number of variationsto the rollout strategy using a range of levels of sophisticationand expert knowledge, and decaying reward to encourage playurgency. We examine the effect of utilising various pruningstrategies in order to increase the information gained from eachdeterminization, alongside methods that increase the relevance ofrandom choices. Additionally we deconstruct the move generationprocedure into a binary yes/no decision tree and apply MCTS tothis finer grained decision process. We compare our modificationsto a basic MCTS approach for Magic: The Gathering using fixeddecks, and show that significant improvements in playing strengthcan be obtained.

UR - http://www.scopus.com/inward/record.url?scp=84871392123&partnerID=8YFLogxK

U2 - 10.1109/TCIAIG.2012.2204883

DO - 10.1109/TCIAIG.2012.2204883

M3 - Article

VL - PP

JO - Computational Intelligence and AI in Games, IEEE Transactions on

JF - Computational Intelligence and AI in Games, IEEE Transactions on

SN - 1943-068X

IS - 99

M1 - 6218176

ER -