Secrets from the mind of an expert / Philip A. Ross

Studying mental processes in grandmaster chess players hints at how people achieve excellence in other areas as well

A person walks around a circle of chess tables, looks at each board for two to three seconds and then makes a move. Dozens of amateur players sit around the circle and consider the counter move they must make until the round is completed. The year is 1909, the man is the Cuban Jose Raul Capablanca and the result is his overwhelming victory: 28 victories in 28 games. This demonstration was held as part of a tournament campaign during which Capablanca won 168 games in a row. How did he manage to play so well at such speed? And how many moves forward was he able to calculate with such limitations? It is said that Kapbanka replied: "I only see one move in advance, but always the right move."

This is how Kapbanka summarized what would later yield a hundred years of psychological research: the main difference between a master chess player and a novice chess player lies in the first seconds of thought. This rapid knowledge-oriented perception, sometimes called aphasic, is also characteristic of experts in other fields. Just as a chess grandmaster will remember every move in the game he played, so, more than once, a talented musician will be able to reproduce in writing the notes of a sonata he has heard only once. And just as the rabbi finds the best step in a flash, a specialist doctor can sometimes arrive at an accurate diagnosis within a few moments of observing the patient.

But how do those who excel in their various fields acquire their extraordinary skills? What is the part of innate talent and what is the part of intensive training? Psychologists sought to find answers to this by studying the skills of grandmaster chess players. Summarizing the results of a hundred years of research led to the development of new theories concerning how information is organized in the brain and how it is retrieved when needed. Furthermore, it is possible that this research has important implications for the field of education. It may be possible to use the techniques with which chess players develop their skills to teach reading, writing and arithmetic in school.

The "fruit fly" of the cognitive sciences

The history of human expertise begins with hunting - a skill essential to the survival of our ancestors. The skilled hunter not only knows where the lion has been, but also infers where it will go. Repeated studies show that from childhood onwards the ability to probe is perfected and increases "in direct proportion to age, until the mid-30s, when the skill reaches its peak," according to anthropologist John Bock from the University of California at Fullerton. Training a brain surgeon takes less time.

The vast superiority over novices in their field is what creates true wizards. Without her, these would not be ordinary people who have impressive achievements, and there are, unfortunately, many of them. Careful research from the past 20 years shows that professional stock market wizards are no more successful in their stock market investments than amateurs, famous winemakers are no better at distinguishing between wines than commoners, and doppelgänger psychiatrists are no better at helping their patients than their lesser colleagues. And even if it is clear that expertise exists, such as in teaching or business management, it is often difficult to measure it and even more difficult to explain it.

Skill in chess, on the other hand, can be measured, broken down into components, tested through experiments in the laboratory, and it is easy to observe it in its natural environment - the competition halls. The chess thus served as a main ground for experiments in theories dealing with thinking, to such an extent that it was nicknamed "the fruit fly (Drosophila) of the cognitive sciences".

The measurement of chess skills has gone further than similar measurement attempts in games, sports, and other competitive activities. Statistical formulas give more weight to the results of a player's recent games than to old results, and also take into account the strength of the opponent in each game. The calculations give the players a score that allows them to predict the results of future games with a high degree of confidence. If Player A's score is 200 points higher than Player B's he will win 75% of their games on average. This prediction is valid whether it is a first class player or a completely ordinary player.

The Russian master Garry Kasparov has 2812 points and is therefore expected to win 75% of his matches against the 100th ranked master in the master list, the Dutch master Jan Tiemann, who has 2616 points. Similarly, a competitor in an American tournament who has 1200 points (a score close to the median) will win 75% of his matches against a player with a score of 1000 (close to the 40th percentile). The scoring allows psychologists to assess the degree of excellence based on performance rather than reputation, and to track changes in skill level throughout the career of an actor or actress.

Another reason why cognitive researchers choose chess as a model and not, for example, billiards or bridge, is the reputation they have for the game as a "touchstone for the intellect," according to the German poet Goethe. For many years, the feats of chess masters were attributed to the power of thought bordering on magic. This magic is especially evident in "blindfolded" games where the players do not see the board. The French psychologist Alfred Binet, who was among the inventors of the first intelligence test, asked chess players in 1894 to describe how they play without seeing. At first he assumed that they were picturing in their minds an exact image, almost like a photograph, of the board, but he soon came to the conclusion that the image that appeared before their eyes was much more abstract. The artist does not visualize the mane of the horseman (or his horse, one should say) or the pattern of the wood fibers from which it is made, but only needed an overall knowledge of the position of the tool in relation to the other tools in the playing set. This is unspoken knowledge, similar to what a subway passenger has about the stations on the line of travel.

The blindfolded artist adds to this knowledge details pertaining to the current game as well as salient memories from previous games. Suppose he forgot for some reason the exact location of one of the legs. He will be able to come back and "find" it based on the type of strategy at the beginning of the game, a phase that is frequent in his study and where there is a relatively limited number of possible arrays. He may also recall the logic behind one of his previous steps, for example through such a consideration: "I couldn't have taken the runner from him in the step before the previous one, that is, my legs must have been in my way..." He does not need to remember all the details all the time, because if he wants to he can reproduce Every detail by connecting to a well-organized link system.

If the mastery of such a complex and structured set of information would explain not only the success in blindfolded games, but also other skills of chess artists, such as calculating moves and planning, excellence in the game would not depend on natural talent but, of course, on training. The Dutch psychologist Adrian de Groot, a chess master himself, confirmed this idea in 1938 when he took advantage of a large international competition held in the Netherlands to compare average and strong players with the world's leading chess masters. One of the ways he did this was by asking them to describe their thoughts while looking at a set taken from one of the competition games. He discovered that despite the fact that players below the rank of master analyzed many more options than the very weak players, no significant differences were recorded in the analysis between them and between the artists and the multi-artists. The better players didn't consider more options but better options, just like Capablanca said.

Recent studies show that de Groot's findings also reflect the nature of the arrays he chose for testing. A set-up where a long and precise calculation will decide the balance allows the master to demonstrate his power, since in that case he will be looking further down the branching tree of possible moves than an amateur could even hope to see. Similarly, an experienced physicist may come up with more possibilities than a physics student. However, in both cases the expert does not rely so much on his natural analytical ability, which is supposedly better, but on an organized knowledge base. When a relatively weak player is required in a complicated situation, he may be considered for half an hour and will usually also see many moves ahead, but the correct move will be hidden from his eyes, while the rabbi sees the move immediately without analyzing at all.

De Groot also had his subjects observe play sets for a limited time and then try and reconstruct them from memory. Across the entire range of ranks, a parallel was found between the subject's ability to perform this task, and his strength as an actor. Novices were able to remember only a few details of a game formation, even if they were given 30 seconds to learn it, while master players usually performed a perfect reconstruction of the formation even after viewing it for only a few seconds. This difference is due to a certain type of memory that is special to the types of common game setups. This special memory is surely the result of training, since master artists do not do better than other people in general memory tests.

Similar findings were found among bridge players (who can remember which cards they played in many games), computer programmers (who can reproduce long pieces of software) and musicians (who remember long pieces of music). In fact, such a memory, dedicated to the subject of a certain field of occupation, is used as a standard test to identify excellence in the field.

The conclusion that experts rely more on structural knowledge than analysis is supported by a rare case study of a weak chess player who, within nine years, from 1978 to 1987, became one of Canada's leading chess masters. Neil Cherns, professor of psychology at Florida State University, showed that despite his improvement in the game, he did not include his ability to analyze game sets, but instead relied on vastly improved knowledge of game sets and appropriate strategies.

The grouping theory

In the 60s, Herbert A. Simon and William Chase, both of Carnegie Mellon University, tried to better understand expert memory by studying its limitations. They continued from the point where de Groot left off and asked players of different levels to reproduce artificial game sets, that is, sets in which the pieces were placed on the board at random, and not sets that were in artists' games [see box on the previous page]. The correlation between the player's level and his accuracy in reproducing artificial arrays was much lower than in the authentic arrays.

It was found, then, that the chess memory is even more specific than it initially seemed, it is not just aimed at the game of chess but rather at typical layouts of pieces. These experiments confirmed previous studies that convincingly showed that abilities in a certain domain do not usually transfer to other domains. The American psychologist Edward Thorndike first noticed the inability to transfer more than 100 years ago, when he showed that learning Latin, for example, does not improve the command of English, and that proving theorems in geometry does not help to think logically in everyday life.

Simon explained the relatively lesser ability of chess masters to reproduce artificial arrays using a model based on meaningful patterns called clusters. He needed this concept to explain how chess masters manage to handle huge amounts of stored information, a task that seems to tax the brain's working memory. Psychologist George Miller of Princeton University measured the scope of working memory, that record book of the brain formerly known as short-term memory, in his well-known article from 1956 called "The Magic Number Seven Plus Minus Two". Miller showed that people can only attend to five to nine items at a time. By attaching hierarchical structures of information to files, Simon argued, chess masters circumvent this limitation because this method allows them access to five to nine files, rather than access to a similar number of smaller items.

Take for example the sentence "Little Jonathan ran to the garden in the morning". The number of information files that this sentence contains depends on familiarity with the song and knowledge of the Hebrew language. For most Hebrew Hebrew speakers, this sentence is part of a larger group, namely the well-known children's song. For those who know Hebrew, but do not know the song, the entire sentence is a single cluster that stands alone. For a person who has memorized the words of the song but not their meaning, this sentence has six clusters, and for a person who knows the letters but not the words, the sentence has 21 clusters.

In the world of chess, the same differences distinguish between chess beginners and master chess players. A layout of 20 tools on the board presents a beginning player with many more than 20 files of information, since there are so many options for placing the tools on the board. A master, on the other hand, may see a part of the layout like this: "a small scream with a pianchetto runner" combined with "an Indian-style blocked leg chain on the kingside", thus compressing the entire layout into five or six clusters. By measuring the time it takes to commit a new block to memory and estimating the number of hours a player must spend studying chess until he reaches the rank of grandmaster, Simon estimated that the average grandmaster has access to 50 to 100 files of chess information. A great master is able to recall each of these blocks from his memory if he only looks at the set of the game, just as most Hebrew Hebrew speakers can recite "Little Jonathan" in its entirety after hearing only the first words of the song.

Nevertheless, the grouping theory was found to be somewhat problematic. It is unable to adequately explain some aspects of memory, for example the ability of experts to demonstrate their skills even when distracted (a common tactic in memory research). Chernes and Kay Anders Erickson of Florida State University argued that there must be another mechanism that allows high achievers to use their long-term memory as if it were also a "record book." "The very fact that very skilled players can play blindfolded almost at their normal level is almost impossible to explain with the tools of grouping theory, because the player must know the layout of the game and then turn it around in his memory," says Erickson. Such manipulation involves changing the stored files at least to some extent, a task that can perhaps be likened to reciting "Little Jonathan" from end to beginning. It can be done, but not easily, certainly not without many false starts and mistakes. However, games by master artists that are played quickly and blindfolded are usually of a surprisingly high standard.

Erickson also cites studies of doctors clearly storing information in their long-term memory and retrieving it in a way that allows them to diagnose patients. But perhaps the most everyday example that Erikson offers concerns reading. In a study he conducted with Walter Kinch of the University of Colorado in 1995, it was found that interrupting skilled readers did almost nothing to slow their return to deep reading, ultimately delaying them by only a few seconds. To explain this finding, the researchers turned to a structure they called "long-term working memory", a term that sounds almost like a word and its opposite, because long-term memory was attached to the only thing that was always defined as foreign to it: thinking. But brain scan studies conducted in 2001 at the University of Konstanz in Germany provided confirmation of the hypothesis after showing that outstanding chess players activate their long-term memory much more than novices [see image on opposite page].

Fernand Gobert of Brunel University in London advocates a competing theory he developed with Simon in the late 90s. In this theory, the grouping idea is expanded and it also includes large and highly characteristic patterns that involve about a dozen game tools. In such a "template", as they say, there will be several empty cells into which the artist will enter variable values ​​such as my feet or a runner. For example, the layout "queen-pawn position isolated from the Nimza-Indian defense" will be used as a template, and the chess master will be able to change the value of an empty cell and thus redefine it as "the same layout minus the black runner". If we return to the world of children's songs for example, this is similar to memorizing a variation of "Little Jonathan" by replacing words in certain "cells" with rhyming words, such as "Yohanan" instead of "Yonathan", "apples" instead of "chicks" and so on. Anyone who knows the original pattern will be able to memorize the variation in a moment.

The multiplicity of elevations

All the theorists of expertise agree on one thing, and that is that building these structures of thought requires a tremendous effort. Simon formulated his own psychological rule, the rule of the decade, according to which it takes about ten years of hard work to achieve excellence, whatever the field. Even child prodigies in their fields, such as Gauss in mathematics, Mozart in music and Bobby Fischer in chess, undoubtedly put in equal effort, and perhaps started earlier and worked harder than others.

According to this approach, the proliferation of chess prodigies in recent years only reflects the appearance of computerized training programs that allow today's children to learn much more than their predecessors games of artists and to play more frequently against master-level software. Fischer caused a sensation when he reached the rank of grandmaster in 1958 at the age of 15, currently the record holder is the Ukrainian Sergey Karjakin who did it at the age of 12 and seven months.

Erickson claims that what is important is not the experience in itself, but "effortful study" which means constantly dealing with challenges that are slightly beyond the student's ability. This is why passionate hobbyists sometimes spend tens of thousands of hours playing chess or golf or playing an instrument and never progress beyond the rank of amateur, while a properly trained student can achieve them in a relatively short time. It is interesting to see that the accumulated chess playing time, even in competitions, contributes less to the player's progress than hard study, the main contribution of such games is identifying weak points for future study.

Even novices put more effort at the beginning of their studies, so beginner golfers or driving students for example, usually improve very quickly at the beginning of their journey. But most people give up on the matter when they have reached a level that puts their minds at ease, a level that does not fall short of that of their friends on the golf course, or that will pass, say, their driving test. Then their performance becomes automatic and thus the door is closed to further improvement. Apprentice experts, on the other hand, keep an "open mind" so that they can examine, criticize and enrich what is inside, and therefore they approach and go to the level set by the leaders in the field.

In the meantime, the standards by which those who excel in their field are crowned are getting stricter and stricter. High school age runners are already running a mile in four minutes, conservatory students are playing pieces that once only virtuosos dared to attempt. But chess, again, is the domain that provides the most compelling comparison over time. John Nunn, a British mathematician and chess grandmaster, recently compared with the help of a computer all the mistakes made in two competitions, one held in 1911 and the other in 1993. The modern players are much more precise in their game. So Nan examined all the games of a player who reached "a good place in the middle" in the 1911 competition and calculated and found that the score he would have received today would not have exceeded 2100, a few hundred points below the minimum rating of a master, "and that too on a good day with Ruach Gavit. " While the best artists of the past were considerably stronger than this, they still ranked far below today's top players.

However, let's not forget that Capablanca and his contemporaries did not have computers and game databases at their disposal. They had to learn everything themselves, just as Bach, Mozart and Beethoven did. And if in technique they fall short of contemporary chess masters, in creativity they surpass them many times over. The same comparison can be made between Isaac Newton and a typical physicist who recently received a doctorate.

Here many skeptics will lose their patience. It is impossible, they will say, that all that is needed to win a gold medal is just training, training and more training. But surprisingly, the belief in innate talent, which the outstanding themselves and their coaches perhaps adhere to more than anyone else, is not supported by conclusive evidence. In 2002, Govet conducted a study that involved British chess players from amateurs to masters, and found no relationship between their strength in the game and their visual-spatial abilities, measured in shape memory tests. Other researchers have found no parallel between the predictability of horse racing wizards and their mathematical abilities.

Although no one has yet been able to predict who will excel greatly in any field, an interesting experiment has shown that it is possible to deliberately bring a person to a high level of excellence. László Polgar, a Hungarian educator, educated his three daughters at home, dedicating about six hours a day to chess training. The result: one international chess artist and two multi-artists - the strongest group in the history of chess for those parents. The youngest daughter of the Folger family, 30-year-old Yodit, is currently ranked 14th in the world.

Folger's experiment proved two things: that it is possible to raise master artists from childhood and that women can also reach the rank of master master. It is no coincidence that many chess prodigies were discovered after Laszlo Folger published his book on learning chess. The number of musically gifted children has also increased dramatically since the musical equivalent, from the home of Leopold, Mozart's father, 200 years earlier.

It seems, then, that motivation is a more important factor than natural ability in developing excellence. It is therefore no coincidence that in music, chess and sports, which are all fields where excellence is defined by competitive achievements and not by academic references, professionalism is becoming more common at younger and younger ages, with the support of parents who are dedicated to the matter and sometimes the extended family as well.

What's more, success is built from success, since every achievement may strengthen the child's spirit. A study conducted in 1999 examined professional soccer players from several countries and showed that many of them, compared to the general population, were born at a certain time of the year, and therefore their age at the time of their registration in the youth soccer leagues was higher than average. [See box on opposite page]. At the beginning of their journey as players, these children thus enjoyed a considerable advantage of size and strength over their teammates. And since kids who are bigger and faster than their peers get more opportunities to hold the ball, they scored more goals and their success in the game encouraged them to improve even more.

Teachers of sports, music and other subjects tend to believe that talent is important and that they know how to recognize talent by being told about it. But actually they probably confuse talent with early development. Usually there is no way to tell by listening to a recital alone, whether the young violinist's excellent playing is the result of innate talent or years of playing the Suzuki method. Capablanca, who is still considered the greatest "natural" chess player of all, boasted that he never learned the game. In practice, he failed his studies at Columbia University, among other things because he would spend long hours playing chess. His famous catch speed was a product of training and not a substitute for it.

Most psychological findings show that people become wizards and are not born as such. Furthermore, the proof that it is possible, in a relatively short period of time, to make children excel in chess, music and a host of other fields, poses a clear challenge to schools. Will educators be able to find a way to encourage students to engage in hard study that will improve their reading and math skills? Roland J. Fryer Jr., an economist at Harvard University, tried to offer financial rewards to motivate students from low-achieving schools in New York and Dallas. In one of the programs still operating in New York, the teachers test the students once every three weeks and reward those who do well on the test with a small amount, about $10 or $20. The initial results look promising. Instead of struggling with the question "Why can't Yossi read?" Educators should perhaps ask "Why should there be anything in the world that Yossi cannot learn?"

Philip A. Ross is an editor and writer at Scientific American, a chess player, and the father of Laura Ross, a chess master ranked 199 points above him.

The article was published with the permission of Scientific American Israel