ICGA Journal - Volume 38, issue 2

Article Type: Other

DOI: 10.3233/ICG-2015-38201

Citation: ICGA Journal, vol. 38, no. 2, pp. 65-66, 2015

Authors: Teytaud, Fabien | Dehos, Julien

Article Type: Research Article

Abstract: Over the last few years, many new algorithms have been proposed to solve combinatorial problems. In this field, Monte-Carlo Tree Search (MCTS) is a generic method which performs really well on several applications; for instance, it has been used with notable results in the game of Go. To find the most promising decision, MCTS builds a search tree where the new nodes are selected by sampling the search space randomly (i.e., by Monte-Carlo simulations). However, efficient Monte-Carlo policies are generally difficult to learn. Even if an improved Monte-Carlo policy performs adequately in some games, it can become useless or harmful …in other games depending on how the algorithm takes into account the tactical and the strategic elements of the game. In this article, we address this problem by studying when and why a learned Monte-Carlo policy works. To this end, we use (1) two known Monte-Carlo policy improvements (PoolRave and Last-Good-Reply) and (2) two connection games (Hex and Havannah). We aim to understand how the benefit is related (a) to the number of random simulations and (b) to the various game rules (within them, tactical and strategic elements of the game). Our results indicate that improved Monte-Carlo policies, such as PoolRave or Last-Good-Reply, work better for games with a strong tactical element for small numbers of random simulations, whereas more general policies seem to be more suited for games with a strong strategic element for higher numbers of random simulations. Show more

DOI: 10.3233/ICG-2015-38202

Citation: ICGA Journal, vol. 38, no. 2, pp. 67-80, 2015

Authors: Hartisch, Michael

Article Type: Research Article

Abstract: We investigate the game “Karl’s Race” which has elements of chance and is closely related to “Ein-Stein Würfelt Nicht!”. Its state space is significantly smaller, without losing its fascination and complex nature. This game can be analysed completely by using retrograde analysis and thus creates an excellent testing ground for other heuristics dealing with games with chance nodes. Since analysing games with chance nodes requires saving expected values, and thus rational numbers, an enormous amount of memory is required during the calculation and when saving the exact data. Therefore, we examine the effect of intentional rounding during the retrograde …analysis. We find that certain types of rounding only slightly affect the playing strength of players using these data while reducing the required memory space to 25 percent. Show more

DOI: 10.3233/ICG-2015-38203

Citation: ICGA Journal, vol. 38, no. 2, pp. 81-93, 2015

Authors: Huntington, G. | Haworth, G.McC.

Article Type: Research Article

DOI: 10.3233/ICG-2015-38204

Citation: ICGA Journal, vol. 38, no. 2, pp. 94-99, 2015

Authors: Hartmann, Dap

Article Type: Book Review

DOI: 10.3233/ICG-2015-38205

Citation: ICGA Journal, vol. 38, no. 2, pp. 100-101, 2015

Article Type: Other

DOI: 10.3233/ICG-2015-38206

Citation: ICGA Journal, vol. 38, no. 2, pp. 102-102, 2015

Authors: Krabbenbos, Jan | van den Herik, Jaap | Haworth, Guy

Article Type: Other

DOI: 10.3233/ICG-2015-38207

Citation: ICGA Journal, vol. 38, no. 2, pp. 103-114, 2015

Authors: Krabbenbos, Jan | van den Herik, Jaap | Haworth, Guy

Article Type: Other

DOI: 10.3233/ICG-2015-38208

Citation: ICGA Journal, vol. 38, no. 2, pp. 115-123, 2015

Authors: Krabbenbos, Jan | van den Herik, H.J.

Article Type: Other

DOI: 10.3233/ICG-2015-38209

Citation: ICGA Journal, vol. 38, no. 2, pp. 123-123, 2015

Authors: Levy, David

Article Type: Other

DOI: 10.3233/ICG-2015-38210

Citation: ICGA Journal, vol. 38, no. 2, pp. 124-124, 2015