ICGA Journal - Volume 31, issue 3

Authors: van den Herik, Jaap

Article Type: Other

DOI: 10.3233/ICG-2008-31301

Citation: ICGA Journal, vol. 31, no. 3, pp. 129-130, 2008

Authors: Guid, Matej | Perez, Aritz | Bratko, Ivan

Article Type: Research Article

Abstract: In 2006, Guid and Bratko carried out a computer analysis of games played by World Chess Champions in an attempt to assess as objective as possible one aspect of the playing strength of chess players of different times. The chess program CRAFTY was used in the analysis. Given that CRAFTY ’s official chess rating is lower than the rating of many of the players analysed, the question arises to what degree that analysis could be trusted. In this paper, we investigate this question and other aspects of the trustworthiness of those results. Our study shows that, at least for …pairs of the players whose scores differ significantly, it is not very likely that their relative rankings would change if (1) a stronger chess program was used, or (2) if the program would search more deeply, or (3) larger sets of positions were available for the analysis. Experimental results and theoretical explanations are provided to show that, in order to obtain a sensible ranking of the players according to the criterion considered, it is not necessary to use a computer that is stronger than the players themselves. Show more

DOI: 10.3233/ICG-2008-31302

Citation: ICGA Journal, vol. 31, no. 3, pp. 131-144, 2008

Authors: Chaslot, Guillaume M.J-B. | Winands, Mark H.M. | Szita, István | van den Herik, H. Jaap

Article Type: Research Article

Abstract: Recently, Monte-Carlo Tree Search (MCTS) has become a popular approach for intelligent play in games. Amongst others, it is successfully used in most state-of-the-art Go programs. To improve the playing strength of these Go programs any further, many parameters dealing with MCTS should be fine-tuned. In this paper, we propose to apply the Cross-Entropy Method (CEM) for this task. The method is comparable to Estimation-of-Distribution Algorithms (EDAs), a new area of evolutionary computation. We tested CEM by tuning various types of parameters in our Go program MANGO . The experiments were performed in matches against the open-source program GNU …Go. They revealed that a program with the CEM-tuned parameters played better than without. Moreover, MANGO plus CEM outperformed the regular MANGO for various time settings and board sizes. From the results we may conclude that parameter tuning by CEM genuinely improved the playing strength of MANGO , for various time settings. This result may be generalized to other game engines using MCTS. Show more

DOI: 10.3233/ICG-2008-31303

Citation: ICGA Journal, vol. 31, no. 3, pp. 145-156, 2008

Authors: Vučković, Vladan

Article Type: Research Article

Abstract: This note addresses theoretical and practical aspects of an important issue in computer-chess programming - the problem of chessboard representation. There are two main approaches used by the majority of computer-chess programs: arrays and bitboards. However, after many years of research and experiments with the chess engine AXON and its parallel version ACHILLES , we would like to introduce an alternative chessboard representation CCR (Compact Chessboard Representation ), which performs very well, both on 32-bit and 64-bit hardware platforms. We will try to portray critical advantages and disadvantages in the programming and utilization of both data structures.

DOI: 10.3233/ICG-2008-31304

Citation: ICGA Journal, vol. 31, no. 3, pp. 157-164, 2008

Article Type: Other

DOI: 10.3233/ICG-2008-31305

Citation: ICGA Journal, vol. 31, no. 3, pp. 165-165, 2008

Authors: David-Tabibi, Omid

Article Type: Other

DOI: 10.3233/ICG-2008-31306

Citation: ICGA Journal, vol. 31, no. 3, pp. 166-170, 2008

Authors: David-Tabibi, Omid

Article Type: Other

DOI: 10.3233/ICG-2008-31307

Citation: ICGA Journal, vol. 31, no. 3, pp. 171-171, 2008

Authors: Levy, Alvin

Article Type: Other

DOI: 10.3233/ICG-2008-31308

Citation: ICGA Journal, vol. 31, no. 3, pp. 172-178, 2008

Authors: Coulom, Rémi

Article Type: Other

DOI: 10.3233/ICG-2008-31309

Citation: ICGA Journal, vol. 31, no. 3, pp. 178-178, 2008

Authors: Chaslot, Guillaume | Hoock, Jean-Baptiste | Rimmel, Arpad | Teytaud, Olivier | Lee, Chang-Shing | Wang, Mei-Hui | Tsai, Shang-Rong | Hsu, Shun-Chin

Article Type: Other

Abstract: During the last three years, the level of Go programs has improved tremendously due to a new technique: Monte-Carlo Tree Search. In 2008, several Human-Computer Go events have been organized to assess the level of the new programs. MO GO (in some contests called MO GO TITAN ) and CRAZY STONE , the two pioneer Monte-Carlo Go Tree Search programs, managed to defeat professional players for the first time in history. MO GO won on the 9x9 board, and MO GO TITAN on the 19x19 board with a 9-stones handicap. A short time later, CRAZY …STONE set a new record by winning a game on a 19x19 board with an 8-stones handicap. In this report, we present the results of the most important human-computer matches that took place in 2008, in the USA, Japan, and Taiwan. Show more

DOI: 10.3233/ICG-2008-31310

Citation: ICGA Journal, vol. 31, no. 3, pp. 179-185, 2008