Better Thinking for Better Chess

 By

Mike Serovey - Author of Better Thinking for Better Chess

Mike Serovey – Author of Better Thinking for Better Chess

Michael Robert Serovey, MA, MISM

Copyright © 2012 by Michael Robert Serovey & Mike Serovey Enterprises


 

Forward

 

This book is based upon a directed study I did back in 1988 as a part of my Bachelor of Arts degree in psychology. Although I have made a few corrections in spelling and grammar and reformatted the tables, this is basically the same report that I turned in for a grade back in 1988. At that time the only person who was going to read it was the psychology professor who was grading it. In spite of my errors in that original version I received a grade of “A” This is my second attempt to make this report available to a wider audience. Based upon the limited feedback that I have received so far, chess players who have no background in psychology or statistics will not understand or care for a major portion of this book. On the other hand, people who understand statistics and psychology but do not like chess will not get much out of this book either.

This book is not a course in statistics and leaving out the chi square tables will compromise a huge portion of this work. In order to deliver on my promise of information on how to think better, I have found it necessary to break this information into a series of books that will have two or more volumes. This book, Volume 1, is the background information. Volume 2 will be more for the “average Joe”.

In Volume 2 I plan to expand upon and clarify topics that were not adequately covered in Volume 1. Again, feedback on what you would like me to cover better is always welcome.

More information about me and my chess career can be found at http://MikeSeroveyOnChess.com. My Facebook page is at http://www.facebook.com/mserovey. My Twitter profile is at https://twitter.com/mserovey. My Google + chess page is at https://plus.google.com/b/115567312592622744297/115567312592622744297/posts and my Amazon author’s page is at https://www.amazon.com/author/michaelserovey. My Hubpage on Chess is at http://mserovey.hubpages.com/hub/How-to-Win-at-Chess.

 

 


 

AN INQUIRY INTO THE COGNITIVE PROCESSES OF CHESS PLAYERS UNDER TOURNAMENT OR MATCH CONDITIONS

By

 Michael Robert Serovey

AN ABSTRACT

Of a directed study done in partial fulfillment of the requirements for the degree of Bachelor of Arts in the Department of Psychology in the University of South Florida

JULY 1988

Professor: Joseph Blount, Ph.D.


Introduction

 

 

The present study assessed the awareness of tournament chess players mental processes, the processes used during serious play, and other tournament strategies. Thirty one male, rated tournament players were surveyed during 3 separate chess tournaments. Ten of these then participated in a board experiment.

 

The results indicate that overall, the stronger players are more accurate and thorough in their assessments of positions that are actually in front of them and in mentally tracking moves from that position.

 

However, when analyzing from an internal representation of a position ahead of this, without using sight of the board, the stronger players are only slightly more accurate than the intermediate level players who are clearly better than the novices. This is contrary to what was expected.

 

The 2 main differences between novices and intermediates are that the novices never or rarely attempt blindfold play while the intermediates seldom do and the novices usually don’t plan ahead while the intermediates do.

 

The differences between the top and middle players were more of quality than quantity. Both groups did and said pretty much the same things, only the masters were better at them.

 

Most psychologists that study chess players are interested in what makes the masters better than the average players. The early studies by Simon and Chase (1973) and deGroot (1965) concluded that pattern recognition and memory were the bases for expertise. The implication of more recent studies, for example Charness (1981), is that chess masters use better rules for selecting their moves and memory for stored patterns is greater in masters than non-masters. That would make experience and study all that is necessary to become a master. Yet, a visit to any large chess tournament would show that this is not the case. One would see class D players that have 40 years’ experience and class A players who have less than 1.

 

Also, these notions have come under attack by some of the chess players themselves as well as some psychologists. For example, in a letter to the editor of Chess Life, Dr. Lawrence Crawford said that he would, “merely like to inform your readers that there are those within psychology who would dispute the usefulness of an information-processing approach to chess.” He also doubted that masters such as used in deGroot’s study would show the memory differences that Simon and Chase’s subjects did (December 1961).

 

An article in the September 1961 issue of Chess Life by Dr. Dennis Holding summarized the work of Simon and Chase and then presented some data of his own. He trained 50 tournament players to give numerical estimates of the differing advantages for White and Black. He found that the stronger players (rated 1500-2000) actually did rate the problem positions more accurately than those rated less than 1500. He also found that Cat I players (formerly class A) tended to pick the correct side as winning. The Cat Ill (class C) players would call it a draw and the Cat Y (Class E) would call the wrong side as winning. In a 1969 study done with Bob Reynolds, he found that stronger players would center their searches on and around the point where the functions of the pieces interact, not around the spatial center of the board. Examples of these concepts can be found in many grandmaster games.

 

An article by Dr. Don Ifill appeared in the April 1966 issue of Chess Life. He did a survey of 83 players ranging from under 1400 to over 2400. He wanted to know to what the different levels of players attributed their respective wins and losses. He found that 56 out of 83 players said they win because of skill and lose because of blunders! The “eight” most likely reasons for miscues were then rank ordered. A ninth reason was added in by four players. The reasons were as follows: 1) Poor chess judgment; 2) Faulty analysis of a complicated position; 3) Falling into a “trap” set by the opponent; 4) Under time pressure; 5) Fatigued after a long playing session; 6) Tried an unsound sacrifice; 7) Overconfident when winning; 8) Recognized an error before the opponent replied; and 9) Moved too rapidly (the added one). He found that those who win because of skill lose because they move too quickly, or faulty analysis of complicated positions. Those who win because of luck, lose because they have poor chess judgment.

 

In his book, Psychology in Chess, Nikolai Krogius outlines several different errors and the reasons for them. He also gives suggestions on how to correct them. One error was the retained image. “This is the transference of an assessment of a past position, or of the action of the separate pieces, in an unaltered form to a new situation that has arisen on the board.” (20) (This report does not explain how to overcome these errors. A future report will cover these topics.) Tactical and strategical ideas can also be retained from earlier positions. In some situations, positions from earlier games can interfere with evaluations of the present game. The retained image does not always hinder creative thinking, though. It can help regulate thinking and promote active self-control. It can help maintain continuity in thinking during the game, and can help in keeping a tentative plan in mind until it can be executed.

 

Next is the inert image. “Inert images are characterized by the fact that the assessment of an existing position is held to be the final assessment of the entire game.”(29) These tend to be generalized statements about who is ahead that are often hastily done and may be incorrect. Even so, they often are held long enough to interfere with an objective evaluation of the current position. Inert images are always a negative factor. To counter their negative effects, Krogius suggests that a player withdraw his thoughts from his own plans and try to think of the game from his opponent’s point of view. After all, he gets to move too!

 

Third is the forward image. “The forward image arises when considering possible future changes in the situation. The role of future events in the game is over-estimated to such an extent that they appear to the player as if they already exist in the present.”(34) This can lead a player to exaggerate the significance of his opponent’s threats, which are often non-existent anyway. These threats can become an obsession. As someone once said, “The threat is greater than the execution.” The other problem is that one can attach too much importance to the action of one’s own pieces. One can forget that his opponent can move and interrupt the plan. Sometimes the forward image can lead the player into believing that his opponent must move a certain way or capture a piece when he does not have to. This can lead to many madcap schemes that don’t work. Sometimes the forward image is so strong that a player gets caught up in the future and overlooks the most obvious dangers. He can make the grossest one-move oversights.

 

Krogius quotes Blumenthal who said,” No matter how strong the visual imagination, it is quite clear that the mental picture is paler than the visual impression. Therefore, after the opponent has made a move, even if it was anticipated, one should never make one’s reply without thought (except of course, during extreme time pressure). After all, the move has been prepared whilst the given position existed in the imagination alone.”(41)

 

Krogius then devotes a whole chapter to attention. He says that many blunders are inexplicable if we were to consider the nature of the errors and the players who made them. For example, how does one explain the master who overlooks that his Queen is en prise? The master could claim fatigue or time pressure for some of these errors, but not all. The more likely explanation, according to Krogius, is to examine the individual peculiarities of a player’s attention. These personal and quite typical defects in attention can appear at the worst time and with regularity.

 

Krogius lists four reasons that attention can fall off during a game. The first reason is unfamiliar conditions at a competition. He cites as an example a time when he “bravely” sacrificed a pawn without much thought in order to relieve tension caused by playing at an unfamiliar site. He then speculated that this may be the reason that some foreign players perform well below their respective strengths at international tournaments.

 

The second reason is one’s position in the tournament and the significance of the result of the game. Some players are in a must win situation and will take chances that they might not otherwise take. Also, this can cause an over-excitement that could cloud judgment.

 

The third reason is an “unpleasant” opponent. Once a player loses to a particular player, the first may experience a feeling of doom when paired with that player again. This feeling may cause one player to consistently lose to the same player even though he can beat other players of the same strength. This may also fall into the category of personal grudges to some degree.

 

The last reason is the opponent’s behavior. By watching one’s opponent, one can gain insight into his personality. We can also assess his emotional state at any given time. Don’t let your opponent’s good mood shake your confidence in your own position. But, if you know how to legally rattle your opponent, then by all means do so. One example of this is when a player sits back in his or her seat, stretches his or her arms and seems quite relaxed. This may be a sign of overconfidence.

 

Hypotheses

 

There are several predictions which can be made based on the literature on chess expertise and chess literature in general:

 

  1. Masters   and experts have more accurate rules for assessing positions than lower rated   players. This is based on training, study and experience. The masters should   assess any given position better than a non-master.

 

  1. Masters   have more accurate, elaborate and/or detailed internal representations of the   chess board than non-masters. These internal representations are probably   visual in the stronger players. Weaker players may not have any representation   at all or they may just talk to themselves in an effort to keep track of what   is going on in the game.

 

  1. Class C   players and above should be able to mentally track a short series of moves   while looking at a position on the board with one hundred percent accuracy.   This is what allows them to follow a combination of moves in their heads prior   to playing the first move in the series on the board.

 

  1. Masters   will be able to make more accurate analysis from their internal images than   the lower rated players. This allows them to find better moves than the lower   rated players during a game.

 

  1. Memory   may not be as important after the opening as it is during the opening since it   is nearly impossible to memorize mid game-moves. Usually a player is on his   own after the first ten moves or so since most opening books don’t go much   beyond that point. Also, most players get out of “the book” by that point.   Although a grandmaster may prepare a line against a particular opponent that   can go twenty moves deep, this is rare in anything other than match play   between grandmasters.

 


 

Method

 

Design

 

This study employed a survey and a board experiment as means to gain insights into the cognitive processes used by chess players during a chess tournament or match. The subjects were divided into three groups by rating and their answers were compared between groups using chi square tables (see glossary of terms for definition of chi square tables). Also, those participating in the board experiment were divided into groups by rating classes and their responses are compared between classes. The subject’s performance on the board experiment is then compared to the subject’s responses on the survey to see if there are any differences in what the subject says he does and what he actually does.

 

Instruments

 

Survey

 

The survey consisted of 28 questions, with some questions having more than one part. The questions asked what the players do, or believe that they do, in many different situations. They are designed to reveal what thinking processes and strategies are used by players of differing strengths, how often the strategies are used, and how aware the players are of their own thinking processes.

 

Board experiment

 

Two positions were selected from two different chess books. The first (Position 1) is from a book on irregular openings. This was selected to reduce the possibility that the stronger players could rely on memory of the opening rather than actually tracking the moves and analyzing from a mental image of that position. The second position was selected from a book of complete games. The position may be familiar, but it is unlikely that the actual game being followed would be. As it turned out none of the players recognized either position. Some of the stronger players did recognize the opening used in the second position, but not the game.

 

A stop watch was used to keep track of the time used by each subject to track the moves in each position. These times were to the nearest one hundredth second.

 

A cardboard box was used to cover the board during the time that the subjects were not allowed to look at it. The board itself was smaller than that normally used in a tournament because it was more convenient to the experimenter. This drew some complaints from some of the subjects who would have preferred a larger board.

 

Subjects

 

Thirty-one males were surveyed. The ages ranged from 12 to 73 with the mean age being 32.065. The ratings ranged from 906 to 2549, with the mean rating being 1765 and the standard deviation being 421.462. The years of experience ranged from 1 to 30 with the mean being 10.339 and the SD being 9.027.

 

The subjects were broken down into three groups by rating. The top group is the candidate masters and above. This group was those rated over 1999. The ages ranged from 22 to 44 with the mean being 30.6. The mean rating was 2236.1 with a SD of 146.072. The years of experience ranged from 3 to 30 with a mean of 15.25 and a SD of 8.671. N=10.

 

The middle group is the class A thru class C players. The ratings ranged from 1400 to 1999 with the mean being 1718.9 and SD=155.67. Age ranged from 16 to 45 with mean=33. Years of experience ranged 1 to 26, mean=6.4 and SD=7.97. N= 14.

 

The bottom group is classes D & E. These are rated under 1400, N=7. Ages ranged from 12 to 73, mean=32. Average rating is 1161, SD=167.624. Years of experience ranged from 1 to 28 with mean=5.85, and SD=9.822.

 

Ten of these subjects participated in the board experiment. There were two masters, one expert (candidate master), one class A, two class B’s, two class D’s, and two class E’s. Their years of experience ranged from 1 to 30 years, mean=13.3, and SD=12.43. The ages ranged from 12 to 73, mean=29.4 and the SD= 18.93.

 

The subjects for this study were selected from the players of three different chess tournaments who agreed to do the survey. The subjects who were selected were not busy at the moments they were approached or could be interrupted. Those who did the survey were asked to participate in the experiment until at least three in each main group had been obtained. Since the experimenter could not afford to compensate the subjects, he had to rely on volunteers.

 

Procedure

 

Surveys were given to whomever would take one at the time, provided that he had a rating. The purpose of the survey was explained, which helped to gain the co-operation of a couple subjects. After the surveys were collected, they were placed in a notebook until they were no longer needed. Those who agreed to do the experiment were then tested individually. Two of these subjects were tested first and then did the survey.

 

The experiment went as follows. The first position was set up on the board from a computer generated diagram of the start position. The subjects were told that they had 15 minutes to mentally track the moves that they were to be given. It was emphasized that they could not move the pieces but they were to mentally track the moves instead. They were told that they would be given the moves for both sides on demand. All subjects faced the board as if they were playing White. When the subject said that he was ready, the stop watch was started and the subject was given each move for each side when requested. The subject was given the choice of following the moves in English Descriptive or Algebraic notation. If the subject requested it, the sequence of moves was started over again. There was no limit on the number of times the move sequence would be repeated as long as he did not exceed 15 minutes. At the end of the tracking, the subject was informed of how much time he had left and that he could use it to study the position if he so chose to. When the subject indicated that he was finished, the stop watch was stopped, the time noted and the board was covered with the cardboard box. While the box was covering the position, the subjects were asked to perform four tasks. They were told at the start that they would be asked some questions about the end position while the board was covered, so they were expecting to answer questions, they just did not know what the questions would be for the first position. The same procedure was followed for the second position. After answering the four questions, they were asked to reconstruct from the start position the end position any way they wanted to. The manner in which the subject went about this task was noted as well as how each of the questions was answered.

 

The five tasks included answering four questions as best as they could. The first question was, “What is White doing in this game?” The second was, “What is Black doing in this game?” The third was, “Who is ahead and why?” The fourth task was, “Develop a plan of continuation for White.” These tasks were intended to get the subject to think about the game much as he would if actually playing it.

 

After doing the five tasks in position one, the end position was checked for accuracy and the number of correctly placed pieces noted. The subject was then asked if he had changed his assessment of the game and any changes were noted. Then the same procedure was followed for position two.

 

Results and Discussion

 

Board Experiment

 

Table 1 contains the times used by the subjects to track position one. The first time listed in each rating is for the subject in that class labeled “one” and the second is for “two”. Contrary to what was expected, the expert was the fastest at tracking and spent no real time analyzing the position. Next fastest were the class D players. They spent some time analyzing, but not to the extent that the masters did. The masters were quick at tracking, but wanted to think about the moves while they were being played. The same things apply to Table 2.

 

Table 3 is the means of the three main rating groups, with times (in seconds) for both positions combined.

 

 

 

Table 1

Position 1

 

Tracking Times

Masters: 2:22.31 & 2:50.52

Expert: 0:47.52

A: 3:16.34

B’s: 5:10.19 & 2:56.18

D’s: 2:51.97 & 2:31.34

E’s: 5:03.62 & 4:50.26

Table 1

Table 1

 

 

Table 2

Position 2

Tracking Times

 

Masters: 4:51.52 & 2:50.74

Expert: 0:42.46

A; 2:03.56

B’s: 0:52.42 & 5:38.51

D’s: 3:51.03 & 2:47.47

E’s: 6:11.13 & 2:51.16

 

Table 2

Table 1

 

Table 3

 

Combined Times

 

Top group: 144.3

Middle group: 199.6

Bottom: 231.9

 

Table 3

Table 1

When combing the times for both positions, the stronger players were faster at tracking, on the overage, than the lower rated players.

 


 

Five Tasks

 

Unfortunately, I could not find the diagrams that I used for these two positions and I also could not remember what books that I got them from. This part of the experiment cannot be replicated exactly. Although I can no longer remember exactly how many moves for each study position the subjects were being asked to track, it was about four moves for each side in each position.

 

The first position was unclear; the book did not say who was ahead. One continuation had White winning, and the other has Black winning. The subjects did not agree on who was ahead or why. Only two subjects said White castled to the wrong side. B2 said that the White Q-side was the K-side because the King was castled there.

 

The second position was clearly a win for White, but the end position for that game (the one that the subjects would analyze) was selected because the winning moves were sacrifices that were hard to find. The best move was a rook sacrifice that was hard to find because White already had a knight and a bishop en prise. M1 said that White was losing because he failed to see that White could win with the rook sacrifice or a less preferred knight sacrifice. M2 did consider the rook sac, but it was last choice. B2 found the knight sac, but not the rook sacrifice. The others could not come up with any specific moves even though they believed that White was winning. E2 reversed the colors when visualizing the position. (See Appendix 2.)

 

It appears that the internal representation of the board is not as accurate as the visual impression, which agrees with what Blumenthal said about visual images as quoted above. It also appears that the master’s visual images are not any more accurate than the Class A players, and the higher rated players are only a little better than the novices.

 

As expected, the players above 1500 had little or no trouble tracking the moves, with one exception. With one exception, those rated below 1500 had some difficulty in tracking the moves and most lost at least one move in both series. (See Appendix 2.)

 

The higher rated players were better able to come up with specific moves for both positions, and their plans tended to be more specific. But, they were not necessarily more accurate when done from internal images. Only when seeing the board did the Class A player clearly outperform the lower rated players. (See Appendix 2.)

Related Photos:

This is the non-standared size board and set that I used in the original 1988 study. One subject did complain about the size of the board and set, but I needed something that I could easily carry around and cover with a cardboard box when I wanted to keep it out of view of the test subjects. This box is similar to the one that I used back in 1988.

Original Board and Set from 1988 Experiment

Original Board and Set from 1988 Experiment

Original Board and Set from 1988 Experiment

The Original Board and Set from the 1988 Experiment

The Original Board and Set from the 1988 Experiment

The Original Board and Set from the 1988 Experiment

In order to read the rest of Better Thinking for Better Chess, you need to buy it from one of the following web sites:

  1. Amazon author page for Mike Serovey
  2. Lulu spotlight for Mike Serovey
  3. Create Space
  4. Barnes and Nobel

People who are interested in Better Thinking for Better Chess may want to check out my recommended reading list.

  1. Think Like a Grandmaster
  2. Play Like A Grandmaster
  3. Train Like a Grandmaster
  4. What It Takes to Become a Chess Master
  5. Invisible Chess Moves: Discover Your Blind Spots and Stop Overlooking Simple Wins

Mike Serovey, MA, MISM, Author of Better Thinking for Better Chess