Cognitive Psychology and Chess Skill
John Rummel
experimental cognitive psychology that have investigated the factors
underlying chess skill. I did this as a project last year (this was written
in APril of 1993) while in the process of writing a paper for publication
in an educational research journal. The paper is not complete yet. <sigh>
There are many more excellent studies available in the psychological
literature. These represent that ones that I was able to read and
(in some cases) annotate.
John Rummel
Avni, A. (1990). Assessment and calculation in top chess players'
decision-making during competition: A theoretical model. Psychological
Reports, 67, 899-906.
Important conclusion regarding his review of the literature: p. 905
"It is commonly believed that qualified chess playing requires the
player to be intelligent, to have an outstanding memory, and in addition,
that chess has educational benefits and develops a systematic, rational
thinking. These views have no solid empirical support. Similarly, the
belief that chess is a game of ongoing calculations may be more myth than
reality."
Subjective Conclusion:
Importance of Memory stressed:
1. Computers search quickly and examine as many positions as time and
processor speed permit, then make a move decision. Human chessmasters seem
to have developed a long-term chess memory, which takes the form of
subconsciously remembered patterns stored as chess "vocabularies," and
begin with patterns such as "have I seen this position before? What worked
before?" etc.
"The crucial variable in explaining the advantage of chess masters and
grandmasters over lower class chess players seems to lie not in the
cognitive processes related to the decision-making and problem solving but
rather in the much richer "vocabulary" or patterns at their disposal,
developed through long years of formal and informal chess education, and
especially through a long an intensive variegated practical experience in
chess.
Avni, A., Kipper, D., & Fox, S. (1986). Personality and leisure activities:
An illustration with chess players. Personality and Individual Differences,
8, 715-719.
specific study investigating the personality characteristics of three
groups of people; Highly competitive chess players, Moderately competitive
chess players, and non-players. All three groups were tested using some
scales of the MMPI (to assess Infrequency [ususual thinking], Paranoia,
Neuroticism, Hostility), the Rosenzweig Frustration-Aggression Test (to
measure Aggression), and the Tavistock Self-Assessment Inventory (to
measure Compulsiveness). The authors found that chess players did differ
from the nonplayers in terms of unconventional thinking and orderliness,
and highly competitive chess players differed from moderately competitive
chess players in terms of being significantly more suspicious.
Empirical Findings:
1. There is evidence to suggest that high rated chess players differ
significantly in their personality makeup from non-chess players in terms
of unconventional thinking (possibly rebellious), orderliness (a tendency
to pay attention to to small details, and an ability to sustain
undistracted concentration in the face of prolonged, tense situations) and
suspiciousness (a tendency to have a greater degree of distrust,
guardedness, and a continuous state of alertness).
2. Moderately rated chess players shared two of these personality traits
(i.e., orderliness and unconventional thinking).
3. "The results failed to show that chess players were more aggressive than
the non-players although there was a slight tendency for the highly
competitive group to score higher on the measure of aggression." p. 719
Subjective Findings: 1. "It appears, therefore, that chess playing contains
two components; playing to win, and playing to participate in a shared
activity." p 719
Holding, D. (1989). Counting backwards during chess move choice. Bulletin
of the Psychonomic Society, 27 (5), 421-424.
Emperical Findings:
1. Backward counting, as a method of interfering with working memory, has a
strong effect on the search processes involved in choosing a move,
universially causing all subjects to consider fewer moves, poorer moves,
and to a smaller depth.
Subjective Findings: 1. The search process is a controlled process, not an
automatic process. Since more highly skilled players tended to experience
more distruption than the weaker players, it appears that this process does
not become more automatic with increase in chess skill. This appears to
argue against the pattern recognition hypothesis in chess skill since
pattern recognition should be relatively automatic.
Therefore, this experiment supports the conclusion that move choice is
based on processes other than pattern recognition.
(paraphrased from the discussion section, p. 424).
Holding, D. & Pfau, D. (1985). Thinking ahead in chess. American Journal of
Psychology, 98 (2), 271-282.
No findings of an emperical nature were discussed.
Horgan, D. & Morgan, D. (1990). Chess expertise in children. Applied
Cognitive Psychology, 4, 129-140.
Emperical Findings:
1. Practice per se is not sufficient for expertise, experience is
important, especially appropriate training and feedback. In children, there
is a positive relationship between level of club activity and improvement
in national rating.
2. For children, spatial and logical abilities such as those measured by
the Raven's Progressive Matrices and the Knight's Tour may be useful for
identifying chess talent.
3. For children (as in adults), skill level in chess is correlated with the
number of pieces correctly remembered from a chess scene.
4. In non-random memory for board experiments, when context is given (white
just captured one of black's pieces, and black will probably capture back),
child subject's performed significantly better. When context is introduced
at the right time in the right circumstance, it can greatly enhance
performance. The same information at the wrong time, however, can reduce
performance.
5. Increase in experience results in judgements based more on abstraction
than concrete principles. This trait is age independent and skill
dependent.
Pfau, H. & Murphy, M. (1988). Role of verbal knowledge in chess skill.
American Journal of Psychology, 101 (1), 73-86.
Excellent study looking at the role of verbal knowledge of chess
information.
Emperical findings:
1. Verbal knowledge differentiates players below the expert level, but not
at the master or grandmaster level.
2. At the master and above level, memory (reconstructing actual game
positions) is a much better discriminator.
Pfau and Murphy's little review of the literature is excellent (pp. 73-75)
Reynolds, R. (1982). Search heuristics of chess players of different
calibers. American Journal of Psychology, 95 (3), 383-392.
Nice introductory statement on the value of chess for investigative
cognitive psychology.
p. 384. The only differences discovered between strong and weak players was
that the former have a quicker grasp of the position. [quoting deGroot
(1965) p. 324] "Within the very first five to ten seconds, the master
subject is apt to have more relevant information about the position
available to him than the lesser player can accumulate in, say, a quarter
of an hour of analysis."
Reynolds, R. (1991). The application of a search heuristic by skilled
problem solvers. Bulletin of the Psychonomic Society, 29 (1), 55-56.
Emperical Findings:
1. There is evidence of a differential heuristic used by master/grandmaster
players as opposed to expert/class players. The strategy, called the homing
heuristic by Reynolds, is characterized by expanding the search
immediately following a negative of a position, and narrowing the search
immediately following a positive evaluation. This strategy was found in the
thinking of the masters and grandmasters, but not in the group of lower
rated players (experts and class players).
Saariluoma, P. (1985). Chess players' intake of task-relevant cues. Memory
and Cognition, 13 (5), 385-391.
In introduction, good discussion and review of the memory vs. perception
models in problem solving. Early research, particurlarly deGroot, has
strongly favored the memory as most important in chess.
Emperical Findings:
1. There was a clear skill difference in speed of perceptual classification
(determining whether or not the king is in check). The higher rated players
were faster than the lower rated players.
2. All players found the random positions more difficult than the game
positions (no skill differences). All players were slower in random
positions.
3. The difference between the game and random positions was additive (the
random positions were proportionately as difficult for the skilled players
as for the less skilled). This is in direct opposition to deGroot's
findings. There was no interaction between the level of skill and the type
of position.
Subjective Findings:
1. The most important implication of the qualitative difference between
intake of task-relevant cues and recall is that memory-driven models of
human problem solving are not sufficient. The intake of task-relevant cues
is necessary for reasonable problem solving in chess. If a player misses an
important task-relevant cue, he is no longer able to solve a problem
properly. As the intake of task-relevant cues is necessary for chess
players' problem solving, it is also necessary to take into account the
qualitative difference between perceptual and memory tasks when considering
chess players' problem solving.
Saariluoma, P. (1991). Aspects of skilled imagery in blindfold chess. Acta
Psychologica, 77, 65-89.
Big study, 7 experiements in all.
Emperical Findings:
1. Some type of chunking is the basis of image formation in blindfold
chess. Players utilize imagery stored in the long-term memory but accessed
temporarily from the working memory. Cooperation and coordination of
long-term and working memory operations is central in the skilled imagery
of blindfold chess players. Their huge storage of chess-specific patterns
of chunks makes it possible for the m to constuct skilled images.
2. Master level (and above) chess players used some very deep coding
system. After each sequence of moves (being read to them) they stopped for
a time and thought about the position. They reported that they had tried in
each position to find a good move, so as to be able to recall it later and
not to confuse the games.
3. The limits of skilled memory are wider than has been thought previously.
4. Skilled players report unaminously that the 'meaning' or 'sense' of the
position is very central in their attempts to store a position. They try,
for example, to find the best move in the position, which they try to
store.
Saariluoma, P. (1992). Visuospatial and articulatory interference in chess
players' information intake. Applied Cognitive Psychology, 6 (1), 77-89.
Emperical Findings:
1. Skilled chess players are normally faster than non-skilled players in
visual information intake.
2. Chess players rely strongly on modules of cognition which are
responsible for visuospatial processing, and not at all on articulatory
systems.
Simon, H. & Chase, W. (1973). Skill in chess. American Scientist, 61,
394-403
Excellent general, non-emperical discussion of chess players and heuristics
used by grandmasters.
p.403 Discussion of the importance of pattern recognition in chess mastery.
Yoskowitz, J. (1991). Chess versus quasi-chess: The role of knowledge of
legal rules. American Journal of Psychology, 104 (3), 355-366.
Emperical findings suggest that the memory model may not adequately explain
chess skill.
(this paper was an undergraduate honors thesis)
Unannotated:
Calderwood, R., Klein, G. & Crandall, B. (1988). Time pressure, skill, and
move quality in chess. American Journal of Psychology, 101 (4), 481-493.
Chase, W., & Simon, H. The mind's eye in chess. In W. G. Chase (Ed.),
Visual information processing. New York: Academic Press, 1973.
Holding, D. (1989). Evaluation factors in human tree search. American
Journal of Psychology, 102 (1), 103-108.
Holding, D., & Reynolds, R. (1982). Recall or evaluation of chess positions
as determinants of chess skill. Memory and Cognition, 10 (3), 237-242.
Klein, G. & Peio, K. (1989). Use of a prediction paradigm to evaluate
proficient decision making. American Journal of Psychology, 102 (3),
321-331.
<end of file>
--
John Rummel
Compuserve Chess Forum
71333...@compuserve.com