Chess and Psychological Experiments

Chess has been a popular choice for psychological research, as it provides psychologists with insights into how humans think about problems.

Studying the way masters think should be able to guide one in their quest for improvement. After all it’s possible to obtain a lot of chess knowledge and still not be a particularly good player. What do masters have that average players are missing?

Fortunately a great memory isn’t a prerequisite even though many GM’s have phenomenal memories. Very often it’s specific to chess. Of course GM’s can calculate further ahead and more accurately than average players, but practically speaking, it often surprises players how little they actually calculate in many positions. In De Groot’s study the findings showed that on average, masters calculated no deeper than weaker players, and often examined fewer variations but they almost always selected superior moves.

The opinion of many psychologists is that the greatest difference between masters and amateurs is in the realm of pattern recognition. Often they only need to take a brief look at a position to assess it accurately. They can instantly see positional themes like pawn chains, weak squares, and open lines, as well as tactical possibilities like Knight forks. Patterns of pieces such as weakened King positions and Rook batteries are recognized and evaluated very quickly. I once witnessed GM Jim Tarjan glance at a position for a few seconds while on his way to the coffee machine and refute the analysis that a local master had been working on for several minutes. How did Tarjan do it? What enabled him to see in seconds something the master had missed in several minutes of examining the position?

Lack of pattern recognition is why new players are often the victim of back rank mates, etc. because they simply failed to notice the danger. Contrast this to an experienced player who automatically recognizes the threat of a back rank mate. You can search this Blog for some interesting comments by others on pattern recognition.

Can the average player use the knowledge gained from these psychological studies to improve his skill? Knowing something is not the same as knowing how to implement what you know. There follows a brief synopsis of some studies that have been done in this area and I will leave it to the reader to draw their own conclusions on how this information may be of help in the area of chess improvement.

The main question in psychological studies of chess is what are the factors that make a strong player? Unfortunately there appears to be no one advantage that always leads to better chess-playing.

The first serious psychological study was conducted by Alfred Binet in 1894. Binet observed blindfold chess players as part of his investigation into memory. His experiment was taken by players of all skill level, from novice to master, and he came to the conclusion that blindfold chess players need knowledge, experience, imagination, and memory. (Isn’t that needed when you can see the board, too?) The masters gave accounts that had some similarities but, at the same time, several differences concerning their blindfold play. They were generally able to remember all the moves played. One master was able to quickly recall all 336 moves that he made in 10 blindfold games played simultaneously. Binet concluded that verbal memory was part of blindfold play. Finally the subjects reported the need to be aware of a general plan for each game the same as in regular, sighted play.

Most masters only an abstract representation of the board, but one player, Joseph H. Blackburne, claimed to visualize an actual chessboard with pieces on it, "just as if before the eyes." As a result of his experiments Binet realized that his original hypothesis of a strong visual memory being essential for blindfold play was wrong. Dr. Reuben Fine claaimed any master should be able to play at least one game of blindfold chess.

The first real psychological enquiry of how chess players think was conducted by Dutch master and psychologist Aadrian De Groot. He interviewed Alekhine, Euwe, Keres, Tartakower, Flohr, and Fine as well as several masters, experts, and "class" players.

De Groot gave them a position set up on a board and their task was to determine the best move to make, and to attempt to verbalize all of their thoughts. One problem was that only the conscious thought could be recorded. This presents a problem because things are sometimes thought out on a subconscious level. Some of the positions DeGroot used were tactical positions and some were more positional in nature.

Four stages in choosing the next move were noted.

1- Orientation. The subject assessed the situation and determined a general idea of what to do next.
2- Exploration. The subjects looked at candidate moves.
3- Investigation. In this stage the subject actually chose what he believed to be the best move.
4- Proof. In this stage the subject convinced himself that the results were correct.

De Groot divided the four phases of problem-solving into two broader progressions: integration to elaboration. This constitutes a transition to a deeper understanding of the problem brings the solver ever closer to the goal. By the time the fourth stage is reached the players had convinced themselves they had chosen the correct move.

De Groot exposed subjects to a position taken from a game, for about 3 to 4 seconds. He found that the top players (grandmasters and masters) were able to recall 93% of the pieces, while the experts remembered 72% and the class players only 51%. De Groot interpreted this to mean the better results of the stronger players was because they had better pattern recognition skills. This is borne out by a study done by Chase and Simon, in 1973, in which they tested players’ memories of positions of games versus random positions. In all legal positions, performance on this test declined as the player's ELO rating declined. But when given random positions, all levels of players did approximately the same. This experiment confirmed that higher-ranked players used some form of chunking, or pattern-matching, that allows them to rapidly encode features of the positions. For instance, even an average player would remember the six pieces comprising a castled king and rook, fianchettoed bishop and three pawns in a group or “chunk.” Beginners would have to remember this formation as separate pieces. Players also tended to remember other small “subsets” of the position such as a bishop was pinning a knight to a queen would be remembered in terms of the pin relationship (or chunk), rather than by recalling the bishop to be at g5, the knight at f6 and the queen at d8.

Chase and Simon defined a chunk as remaining intact through its encoding into long-term memory if at least two-thirds of its pieces remained together upon recall. Using this, they found that 96% of the class A player's chunks remained constant between trials. However, the results of masters were very surprising. Their chunks remained the same only 65% of the time, which is poor in comparison to the class A players, and similar to weaker class players. At first this seemed to dispel the theory of chunking, but a new hypothesis was created; that masters are able to make imaginative insights that involve the restructuring of pieces. What that means is simply that masters had the ability to take a known formation and morph it into another formation that applied to the specific situation…they were more adaptable I guess you would say.

The concept of chunks, or structural units was shown by the "pennies-guessing" task. Subjects are shown a chessboard with pennies representing each piece, taken from a real game position. Their task is to recreate the position by guessing which piece belongs on each square. In addition, the player is told the number of moves that have been made in the game, and whose turn it is to move. Masters are virtually perfect in placing the correct pieces on the board, while players rated 1800-1900 still average over 90% correct. These results show that large libraries of likely piece configurations are known to skilled players. In fact, given a position that occurs fairly early on in a game (up to move 25-30), masters are generally able to reconstruct all of the moves that led up to it.

Chess players need to be able to perceive threats in order to determine their next moves. One experiment suggested that grandmasters are much quicker than novices in certain lower-level perceptual processes. In the first of these experiments, a king of one color was placed on the board, along with a piece of the other color. The subject had to state whether the king was in check or not. The average amount of time to determine if the king was in check was as follows: novices: 1550 ms, class players: 1250 ms, experts: 900 ms, grandmasters: 650 ms. The stronger the player, the faster they could make the determination.

The same test was conducted with a position containing 20 pieces. It was determined that adding 19 pieces to check did not cost a proportional amount of time.

Tikhomirov and Poznyanskaya performed an experiment in which subjects' eye movements were studied to determine the location of the board that they were focusing on. They noted that the eyes followed the paths of the candidate moves and over primary attack or defense. It was also noted that players tended to focus more often on vacated squares suggesting that spatial relationships between pieces are very important. I have read some authors who talked about such things as relationship of the pieces and “aura of the pieces” and how players often think in terms of “lines of force” when visualizing future positions.

Chess requires an immense amount of knowledge to be played at its highest levels. It has been estimated that grandmasters have learned between 50000 and 100000 patterns and moves and it takes years of intense study to achieve a high level of play.

Most amateur players find it difficult to "see" more than five moves ahead in a position. This limit on searching forces inaccuracies into play because such players are simply unaware of the long-term consequences of their moves. Although they can make long-term plans, it is the long-term tactical results that suffer. In other words, it is extremely difficult for average players to determine whether their combinations are truly sound.

Several experiments have proven that stronger players are able to search deeper than weaker ones. One experiment found that a player with a rating of 2600 would be able to search to an average depth of 13.5 plies (6-7 moves). This is consistent with statements from highly-ranked players. In De Groot’s experiments I saw many cases of lower rated players trying to calculate 6-7 moves ahead and most were quite unsuccessful.

One piece of experimental evidence that seems to conflict with the previous findings comes from studies of players of different ages. One study compared players' skill levels, ages and search protocols. The average 20-year-old player rated 1569 considered 22 total moves, while the average 20-year-old rated 2000 considered 49 total moves. 50-year-olds with a rating of 2000 averaged 36 total moves, which are numbers much closer to the lower-ranked 20-year-old rather than the higher-ranked one. Clearly, age makes the search process more efficient, since the players are achieving the same results with different amounts of effort. In terms of base moves, the average 20-year-old searched a mean of 4.1 base moves (starting moves from that position) while the average 50-year-old started his search from only 2.8 base moves. This data suggests that brute force search is not the only way to become a better chess player. Presumably, the experience of the older players allows them to choose better candidate moves, and to search more efficiently. There may be a higher-level process which screens out certain moves based on a database of chess knowledge.

One important factor in chess skill is being able to make an accurate evaluation of a position. In other words, to be able to judge which side stands better. In one experiment in which players rated from about 1000 to 1900 were asked to evaluate a position the number of errors dropped steadily as the skill level increased. For example, players rated ~1000 averaged 7.6 while 1900’s averaged 1.6.

In positions where it is unclear to which side the advantage lies it was interesting to note that players of differing abilities tend to judge middle-game positions as a win, draw or loss equally well! However, in judging end game positions, weaker players tend to make poorer judgments, presumably because their knowledge of positional play is weaker.

In an experiment in 1982 showed that higher-ranked players draw on a lot of experience from real games, rather than just on piece relationships to make their judgments.

Computer program will evaluate king safety after each move but humans will not even consider the safety of their king, knowing that danger is virtually impossible in typical situations. However shortcuts can also result in mistakes. In one experiment strong players were given a position that resembled in a standard smothered mate. These players followed the standard sequence which took 5 moves. However the position had been altered so that the mate was possible in 4 moves. None of the strong players found it. This experiment showed that players use their existing knowledge at the expense of finding the most efficient solution.

There appears to be certain abilities possessed by chess players that tend to improve skill, but the dominating aspect is simply that better chess players have more chess knowledge and experience. This allows them to form patterns more easily, to apply their knowledge of types of positions, and to simply recall useful elements from a vast quantity of memorized information which is directly applicable to the position.