Interesting article in Scientific American this month, looking primarily at Chess players, learning, and the development of expertise. Relevent to GURPS due to the time old controversy regarding learning per hour and IQ... This is from memory, BTW-- I didn't by the magazine-- just read it at the store.
The article has some holes in its conclusions, and isn't particularly rigorous with its interpolations compared with the data actually at hand, but it implies that Chess Grandmasters basically chunk information better, and are able to retrieve more information and more accurate information from long-term memory, but not short-term memory. In random memory tests, Grandmasters fared no better than novices. But in game-like situations, Grandmasters exhibited seemingly miraculous recall. Brain metabolic imaging showed significantly more long-term memory accessing as chess experience/expertise grew when trying to resolve chess problems. In fact, from novice to masters, the progression was fairly linear. From master to grandmaster, however, there was minimal improvement. The authors speculated that this last jump had more to do with problem-solving efficiency, but up until that point, much of chess expertise is the ability to draw from long-term memory (LTM) various scenarious and game situations and their solutions in order to assist with problem-solving. Chess is used as an example because of its rating system, and I think this is used effectively in their conclusions. Interesting tidbits-- more masters and grandmasters today than at any other time. Using recorded games from tournaments in the 1910s and 1920s, they were able to resolve that the best Grandmasters back then would only rate as Masters now-- meaning, today's players are not only more numerous in quality, but that the quality is better. Per this article, expertise is acquired via time and effort. Intense training/studying just a little beyond one's ability is recommended as the way to most quickly improve, but there was no specific rationale to this conclusion except that even the most gifted, precocious masters historically were documented to have gotten progressively better the more chess they studied. Research on current players suggested that tournaments were not particularly helpful for improvement, except as a means of pointing out weaknesses. Players got better by studying various scenarios and their solutions. Credit was given to the modern era's computers as a means of enhanced study. It cited Bobby Fischer, and even today's youngest Grandmaster as examples. Both imrpoved with time, and there was a clear correlation with hours of intensive study and that rise to prominence. They had no hard numbers, of course, but noted that many of today's youngest "Grandmasters" have the same number of hours of study as older chess experts of previous eras-- its just when the player got started. Where this article didn't quite snag me, was in its postulating that intense, hard work and study explained such icons as Mozart and Newton, as well. It used Tiger Woods as an example of someone that had studied from age -10 years old (sarcasm OK?). It noted that a disproportionate number of European professional soccer players were "old for their grade"-- in other words, they were first exposed to soccer in school at times when they were 6-8 months older than their classmates, and so tended to be 6-8 months more mature and developed. They received the early kudos, and therefore more likely went out and studied soccer intensively. I.e. it's not necessarily talent. Another study apparently looked at a fairly average family in which the father exposed his three children to chess at a very young age, and drilled them in it. All three are now considered prodigies. But the hours of study are in the same magnitude as other players of the same rating-- just the age is differe nt. Anyway, Sci Am does not publish their articles on the web, else I'd provide a link. Still, I thought this interesting enough to start a thread. I can provide more details if necessary. Best, -vk _______________________________________________ GurpsNet-L mailing list <[email protected]> http://mail.sjgames.com/mailman/listinfo/gurpsnet-l
