Abstract
By using a well-known chess program and a large data set of chess positions from real games we demonstrate empirically that with increasing search depth backed-up evaluations of won positions tend to increase, while backed-up evaluations of lost positions tend to decrease. We show three implications of this phenomenon in practice and in the theory of computer game playing. First, we show that heuristic evaluations obtained by searching to different search depths are not directly comparable. Second, we show that fewer decision changes with deeper search are a direct consequence of this property of heuristic evaluation functions. Third, we demonstrate that knowing this property may be used to develop a method for detecting fortresses in chess, which is an unsolved task in computer chess.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Of course, this is only an approximation: The terms “decisive advantage”, “large advantage”, and “small advantage” are not strictly defined in the literature.
- 2.
Stockfish 8 64-bit was used in the experiment. In the original study, the programs Rybka 3 and Houdini 1.5a x64 searched a subset of 12 positions up to 20 plies [23].
- 3.
The solutions to the problems given in Fig. 7 are 1.Ba4+!! Kxa4 2.b3+! Kb5 3.c4+! Kc6 4.d5+! Kd7 5.e6+! Kxd8 6.f5! (with a draw), and 1.Rxb7!! (1.Rxf7? g3!) 1...Rf8 (1...Rxb7 2. g3! Kg5 3.Ke2 Rb6 4.Kf1!) 2.g3! Kg6 3.Rb6+! Kg7 4.Rh6!! Kxh6 5.Ke2 Kg5 6.Kf1! Rh8 7.Kg2 a3 8.Kg1! Ra8 9.Kg2 Ra4 10.Kf1! (with a draw), respectively. The latter study was conceived by GM Miguel Illescas.
References
Abramson, B.: Control strategies for two-player games. ACM Comput. Surv. 21, 137–161 (1989)
Thompson, K.: Computer chess strength. In: Advances in Computer Chess, vol. 3, pp. 55–56. Pergamon Press (1982)
Luštrek, M., Gams, M., Bratko, I.: Is real-valued minimax pathological? Artif. Intell. 170, 620–642 (2006)
Pearl, J.: On the nature of pathology in game searching. Artif. Intell. 20, 427–453 (1983)
Scheucher, A., Kaindl, H.: Benefits of using multivalued functions for minimaxing. Artif. Intell. 99, 187–208 (1998)
Gomboc, D., Marsland, T.A., Buro, M.: Evaluation function tuning via ordinal correlation. In: Van Den Herik, H.J., Iida, H., Heinz, E.A. (eds.) Advances in Computer Games: Many Games, Many Challenges. ITIFIP, vol. 135, pp. 1–18. Springer, Boston (2004). https://doi.org/10.1007/978-0-387-35706-5_1
Donkers, H.H.L.M., van den Herik, H.J., Uiterwijk, J.W.H.M.: Selecting evaluation functions in opponent-model search. Theoret. Comput. Sci. 349, 245–267 (2005)
Nau, D.S., Luštrek, M., Parker, A., Bratko, I., Gams, M.: When is it better not to look ahead? Artif. Intell. 174, 1323–1338 (2010)
Heinz, E.: Dark Thought goes deep. ICCA J. 21, 228–244 (1998)
Heinz, E.: Modeling the “go deep” behaviour of CraFty and Dark Thought. In: Advances in Computer Chess, vol. 9, pp. 59–71, IKAT, Universiteit Maastricht (1999)
Heinz, E.: Self-play in computer chess revisited. In: Advances in Computer Chess, vol. 9, pp. 73–91. IKAT, Universiteit Maastricht (1999)
Heinz, E.: Self-play, deep search and diminishing returns. ICGA J. 24, 75–79 (2001)
Heinz, E.: Follow-up on self-play, deep search, and diminishing returns. ICGA J. 26, 75–80 (2003)
Hyatt, R., Newborn, M.: Crafty goes deep. ICGA J. 20, 79–86 (1997)
Junghanns, A., Schaeffer, J.: Search versus knowledge in game-playing programs revisited. In: Proceeings of the 15th International Joint Conference on Artificial Intelligence, vol. 1, pp. 692–697. Morgan Kaufmann (1999)
Steenhuisen, J.R.: New results in deep-search behaviour. ICGA J. 28, 203–213 (2005)
Guid, M., Bratko, I.: Factors affecting diminishing returns for searching deeper. ICGA J. 30, 65–73 (2007)
Müller, K., Pajeken, W.: How to Play Chess Endings. Gambit Publications, London (2008)
Guid, M., Bratko, I.: Computer analysis of world chess champions. ICGA J. 29, 3–14 (2006)
Guid, M., Perez, A., Bratko, I.: How trustworthy is Crafty’s analysis of world chess champions? ICGA J. 31, 131–144 (2008)
Guid, M., Bratko, I.: Using heuristic-search based engines for estimating human skill at chess. ICGA J. 34, 71–81 (2011)
Newborn, M.: A hypothesis concerning the strength of chess programs. ICCA J. 8, 209–215 (1985)
Guid, M., Bratko, I.: Detecting fortresses in chess. Elektrotehniski Vestnik 79, 35 (2012)
Dvoretsky, M.: Dvoretsky’s Endgame Manual, 2nd edn. Russell Enterprises, Inc., Bethlehem (2008)
Friedl, F.: On human and computer intelligence in chess (2017). http://en.chessbase.com/post/on-human-and-computer-intelligence-in-chess. Accessed 15 May 2017
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Guid, M., Bratko, I. (2017). Influence of Search Depth on Position Evaluation. In: Winands, M., van den Herik, H., Kosters, W. (eds) Advances in Computer Games. ACG 2017. Lecture Notes in Computer Science(), vol 10664. Springer, Cham. https://doi.org/10.1007/978-3-319-71649-7_10
Download citation
DOI: https://doi.org/10.1007/978-3-319-71649-7_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-71648-0
Online ISBN: 978-3-319-71649-7
eBook Packages: Computer ScienceComputer Science (R0)