Random Posts

Saturday, October 1, 2011

How Good are Grandmasters?

      An interesting and sometimes heated discussion arose on one forums about how good GM’s are compared to the rest of us.  There seems to be a whole generation of players raised on Internet chess and Blitz chess in particular who have never played in a tournament nor have they ever seen a GM in action up close and personal.  I'm also somewhat surprised how few of them are familiar with chess history or who have actually studied more than just a few GM games.The result is they seem to think the GM’s ability is not a big deal.  After all, they reason, a GM is only a little better than they are and thus, there is no reason to hold their ability in high esteem.
      I also remember one site where a player, who was verified by the site owners as being a real GM and as a result, was allowed to start at an inflated rating.  The brouhaha it created was incredible.  A whole host of players thought he should start at a 1200 rating and prove to them he really deserved his rating!
      GMs are unbelievably strong.  Take the area of tactics for example. A point a lot of people miss when they are doing tactical puzzles, even those guys who get very good at solving them on tactical servers and from books, is that when you see those positions, they represent only half the problem.  The GM probably had the line worked out in his head many moves before and then conceived of a way to reach the position from which the combination was unleashed.  Assuming, of course that his opponent did not make an outright gross blunder which GM’s rarely do.
      Many of these guys seem to think being good can be quick, painless, and easy. To achieve grandmaster status it requires a great deal of commitment, determination, and a certain affinity for the subject, not to mention enormous amounts of experience.  
      Most Grandmasters have average cognitive skills and average memories for matters outside of chess.  This suggests that expertise in chess has less to do with analytical skills - the ability to project and weigh the relative merits of hundreds of options - and more to do with long-term memory and pattern recognition.   They have experienced and stored thousands of game situations in their memory and have the ability to select the best answer from those stored memories. Of course it will also take uncountable hours and passion for the subject to accumulate enough knowledge to reach GM level.
      In the 1920s, a group of Russian scientists set out to quantify the intellectual advantages of eight of the world's best chess players by giving them a battery of basic cognitive and perceptual tests. To their surprise, the researchers found that the GM’s didn't perform significantly better than average on any of their tests.
      In the 1940s, a Dutch psychologist named Adriaan de Groot asked what seemed like a simple question: What separates an average player from world class GM’s? De Groot selected a few positions where there was one correct, but not obvious, move to be made. He then presented those positions to a group of masters and average players and asked them to think aloud while they selected their move.
      What De Groot uncovered was that for the most part the experts didn't look more moves ahead, at least not at first. They didn't even consider more possible moves. They tended to see the right moves, and they tended to see them almost right away.
       When De Groot listened to their verbal reports, he noticed that they described their thoughts in different language than less experienced players. They talked about configurations of pieces like pawn structures and immediately noticed things that were not quite right, like exposed pieces. They weren't seeing the board as thirty-two pieces. They were seeing it as chunks of pieces, systems of tension and lines of force.
      Grandmasters literally see a different board. Studies of their eye movements have found that they look at the edges of squares more than inexperienced players.  This, according to the experts, suggests they are absorbing information from multiple squares at once. Their eyes also search across greater distances and linger for less time at any one place. They focus on fewer different spots on the board and those spots are more likely to be relevant to figuring out the right move.
      But the most striking finding of all from these early studies was their memories. The experts could memorize the board after only a brief glance and they could reconstruct old games from memory.  Later studies confirmed that the ability to memorize board positions is one of the best overall indicators of how good a player somebody is GM’s can remember positions from games for weeks, even years, afterward.
       As impressive as the chess masters' memories were for chess, their memories for everything else were unimpressive. When strong players were shown random arrangements of pieces their memory for the board was only slightly better than those of average players.
      These chess experiments reveal something about memory and about expertise in general: we don't remember isolated facts; we remember things in context. A board of randomly arranged pieces has no context.  Such positions could be described as “white noise.” GM’s use the library of patterns that they've cached away in long-term memory to chunk the board.
       Contrary to the belief that chess is based on analysis, many of the GM’s decisions about which move to make happen almost as soon as he looks at the board.  Something like the time I was watching a television program where an expert was hunting rattlesnakes.  He had been doing it for years and could spot them instantly.  The show’s host never saw them. Or the way a major league baseball outfielder knows where to position himself to catch a fly ball almost as soon as it leaves the bat.
       Using a technique that measures magnetic fields given off by a thinking brain, researchers have found that high-rated players use a section of their brain that suggests they are recalling information from long-term memory. Lower-ranked players are more likely to use a different part of the brain where they are encoding new information. The experts are interpreting the board in terms of their knowledge of past ones. Lower-ranked players are seeing the board as something new.
       GM’s, while they may not be smarter than the rest of us, are way, way better at chess than we are. And, unfortunately, history seems to suggest that if you aren’t a GM by your early 20’s, you never will be.  In fact, if you are not a master by then, you probably never will be.  And that seems to be what upsets some people…who for some reason seem to be in the 20-30 age bracket…being told they have no hope of reaching the GM level.  They also seem to be people who have little or no knowledge of chess other than the way it’s played on the Internet. But then again, most of us think we are the exception that will beat the odds.

No comments:

Post a Comment