Studies in Chess Expertise

Principal Investigator: Fernand Gobet

This project of studying (chess) expertise under its various aspects has been on since 1990, and has been carried out with many collaborators, including the late Herbert Simon (Carnegie Mellon University, Pittsburgh) and Adriaan de Groot (University of Groningen). The goal is to understand the acquisition of (chess) expertise, with a special focus on learning, memory, perception and attention processes in skilled individuals. It has included experimental investigations, computer simulations, and theoretical work.

An updated list of references can be found at my publication page. A good summary of this research, as well as other research on chess and board games in general, can be found in the book Moves in mind--The psychology of board games.

• Main sources on expertise in general

The computer model developed in this project (CHREST) and the experiments that were carried out to test it are inspired by Chase and Simon's chunking theory and by Simon's earlier models of chess expertise, such as MAPP and PERCEIVER. The strength of CHREST is to embody these programs into a single model. A major addition to the model is also the idea of template, which offers the high-level structures that were absent in the earlier models. The CHREST framework has motivated most of the experiments reported below.

References:

• Simon, H. A., & Barenfeld, M. (1969). Information processing analysis of perceptual processes in problem solving. Psychological Review, 76, 473-483.
• Simon, H. A., & Gilmartin, K. J. (1973). A simulation of memory for chess positions. Cognitive Psychology, 5, 29-46.
• Chase, W. G., & Simon, H. A. (1973). The mind's eye in chess. In W. G. Chase (Ed.), Visual information processing. New York: Academic Press.

CHREST

• CHREST (Chunk Hierarchy and REtrieval STructure) is a computer model (programmed in LISP) that simulates human behavior in perception and memory chess experiments, such as those described later. The model, which is based on the notion of chunking, accounts for most of the empirical evidence.
• Main sources on CHREST

CHUMP

• This program, based on CHREST, plays chess by pure pattern recognition and without search. It is planned to add a search module to it.
• Main sources on CHUMP

Eye movement studies

• Study of eye movements as an indicator of chessplayers' cognitive processes and heuristics (in collaboration with Prof. Adriaan de Groot, University of Groningen).
• Main sources on Eye movements

Example of eye movements (presentation time = 5 seconds)

Master's eye movements

Novice's eye movements

Recall of random and distorted positions

• Use of random and distorted positions (e.g. by mirror image) help understand the organization of knowledge in expert chessplayers.
• Chess information seems to be locally encoded, as mirror-image reflections decrease recall performance.
• Contrary to a widely held opinion, chess masters do better with random positions than weaker players.
• Main sources on random and distorted positions

Recall of multiple boards

• In these experiments, subject have to recall a sequence of rapidly presented chess positions.
• These experimentes give us invaluable information about the interaction of short-term memory and long-term memory in chess experts
• We are still analysing a unique experiment where a single subject was trained to recall as many boards as possible
• Main sources on recall of multiple boards

Identification of chunks

• These experiments replicate and extend the classical study of Chase and Simon (1973)
• One key conclusion is that chunk size was underestimated by Chase and Simon (1973)
• The observed chunk size, large for experts, is what is predicted by the template theory.
• Main sources on identification of chunks

Research on chess problem solving

Two directions of research in this domain:

• Replication of the classical study of De Groot (1946). The main difference found is that stronger players search deeper than weaker players, while De Groot didn't find any skill difference in search depth.
• Analysis of the role of pattern recognition and search in simultaneous games. We analyse the results of PCA world-champion Kasparov against national teams.
• Main sources on problem solving

• Use of the CHREST model in the study of the role of aging in expert behavior. In collaboration with Neil Charness.
• Development of CHUMP into a search program and validation of the program using empirical data.

Psychology of chess

http://freud.psy.fsu.edu/~charness/

Computer chess

http://www.xs4all.nl/~verhelst/chess/programming.html

http://www.dcs.qmw.ac.uk/~icca/

Cognitive science

http://mitpress.mit.edu/MITECS

http://www.psy.cmu.edu/psy/faculty/hsimon/hsimon.html

GENERAL EXPERTISE (see also: papers on expertise) 

• Gobet, F. (1998). Chess, Psychology of. MIT encyclopedia of the cognitive sciences. Cambridge, MA: The MIT Press.
• Gobet, F. (in press). Memory in chess players: Comparison of four theories. Cognition.
• Gobet, F. (1996): Expertise und Gedaechtnis (Expertise and Memory). In H. Gruber & A. Ziegler, (Eds.). Expertiseforschung. Theoretische und methodische Grundlagen. Opladen: Westdeutscher Verlag.
• Richman, H. B., Gobet, F., Staszewski, J. J., & Simon, H. A. (1996). Perceptual and memory processes in the acquisition of expert performance: The EPAM model. In K. A. Ericsson (Ed.), The Road to Excellence. Mahwah, NJ: Lawrence Erlbaum.

CHREST (see also: papers on CHREST)

• Gobet, F. (1993). Les memoires d'un joueur d'echecs. Fribourg (Switzerland): Editions Universitaires. 
• de Groot, A. & Gobet, F. (1996). Perception and memory in chess. Heuristics of the professional eye. Assen: Van Gorcum. 

CHUMP

• Gobet F. & Jansen, P. (1994). Towards a chess program based on a model of human memory. In H. J. van den Herik, I. S. Herschberg, & J. W.Uiterwijk (Eds.), Advances in Computer Chess 7. Maastricht: University of Limburg Press.

EYE MOVEMENTS

• de Groot, A. & Gobet, F. (1996). Perception and memory in chess. Heuristics of the professional eye. Assen: Van Gorcum. 

RANDOM AND DISTORTED POSITIONS (see also: papers on expert learning and memory)

• Gobet, F. (1996). EPAM-like simulations of the recall of random chess positions. In U. Schmid, J. Krembs, & F. Wysotzki, (Eds.): Proceedings of the First European Workshop on Cognitive Modeling (Rep. No. 96-39). Berlin: Technische Universität Berlin.
• Gobet, F. & Simon, H. A. (1996). Recall of Random and Distorted Positions: Implications for the theory of expertise. Memory & Cognition, 24, 493-503. 
• Gobet, F. & Simon, H. A. (1996). Recall of rapidly presented random chess positions is a function of skill.Psychonomic Bulletin & Review, 3, 159-163. Abstract
• Gobet, F. (1993). Les memoires d'un joueur d'echecs. Fribourg (Switzerland): Editions Universitaires. Content s (in French)

MULTIPLE BOARDS

• Gobet, F. & Simon, H. A. (1996). Templates in Chess Memory: A Mechanism for Recalling Several Boards. Cognitive Psychology, 31, 1-40. 
• Gobet, F. (1993). Les memoires d'un joueur d'echecs. Fribourg (Switzerland): Editions Universitaires. 

IDENTIFICATION OF CHUNKS

• Gobet, F. & Simon, H. (in press). Expert Chess Memory: Revisiting the Chunking Hypothesis. Memory.

PROBLEM SOLVING (see also: papers on problem solving)

• Gobet, F., & Simon, H. A. (in press). Pattern recognition makes search possible: Comments on Holding (1992). Psychological Research.
• Gobet, F. (in press). Chess players' thinking revisited. Swiss Journal of Psychology.
• Gobet, F. (1997). Roles of pattern recognition and search in expert problem solving. Thinking and Reasoning, 3, 291-313.
• Gobet, F. & Simon, H. A. (1996). The Roles of Recognition Processes and Look-Ahead Search in Time-Constrained Expert Problem Solving: Evidence from Grandmaster Level Chess. Psychological Science, 7, 52-55.

www.brunel.ac.uk/~hsstffg/frg-research/chess_expertise.html

Fernand Gobet

Brunel University

This is a PERSONAL page and does NOT necessarily reflect the position of Brunel University

Last Modified: 15/02/2009