Speed Chess Study Proves Practice Makes Perfect

Speed Chess Study Proves Practice Makes Perfect

Sep 5, 2007, 4:55 PM |

Watching a chess match may seem about as exciting as watching paint dry, but new research shows that a champion's brain works at lightning fast speed, making decisions in seconds that would elude a lesser player with hours to ponder each move.

And that's not just because they are especially gifted at the game. Grand masters reach that level of skill because of the hurdle that keeps most of us mediocre. They practice, practice and practice some more.

And what works for chess probably works in many fields where highly developed skills are essential.

In baseball, a major league slugger couldn't possibly figure out what the pitcher is sending his way in the brief time it takes for the ball to travel from the pitcher's mound to the batter's box, according to various studies. So Barry Bonds knows what he's going to do long before he begins to swing the bat. Subtle clues tell him what to expect, and all those years spent trying to perfect his art make it possible for him to connect far more often than would be possible without all that practice.

So what's the answer to that old debate? Are experts made, or are they born? Psychologist Bruce Burns of Michigan State University in East Lansing thinks it's most likely a little of each. Burns studies how humans acquire skills, and how we deploy those skills under different circumstances. To help him find out, he turned to an old passion. Chess.

Testing the Masters

Burns had competed in chess tournaments in his native Australia, but stopped short of reaching the level of master. He says he gave up because he didn't have the time to develop his skills to the level he needed.

"I played quite a lot, but I really should have practiced more," he says.

For his research, Burns turned to something called "blitz chess." It's like any other chess game, but with one huge exception. The game only lasts 10 minutes, so each player has only five minutes to make all his moves. That boils down to an average of about seven seconds per move in a typical game of 40 moves. That includes the time it takes to ponder the board, move the piece, and punch the time clock.

Obviously, that doesn't leave enough time for a player to sort through all his memory banks and come up with likely scenarios for the next 100 moves, as top players sometimes do. Instead, they have to play almost instinctively.

Burns, whose research will appear in the July issue of Psychological Science, a journal of the American Psychological Society, studied the records from "blitz" chess tournaments in Holland, the United States, and Australia. He would have preferred to stage the tournaments himself, but top chess players are not known for their easy-going, cooperative spirit. So Burns did the best he could with what he had to work with.

Professional chess players have specific ratings, based on their past performances, which are similar to ratings in various sports. Burns wanted to know if top players lived up to their ratings in blitz chess, where they didn't have much time to consider each move.

"There's an advantage to having more time," he says. "You can search [your memory] more, you can consider the possibilities more. If you look even five moves ahead, you have to consider lots and lots of possibilities. What if I do this, and he does that?"

So it would seem likely that performance would drop far below their ratings if pros were denied the time to do all that searching, but Burns found that was not the case. If they had only five percent of the normal time available for a match, the champs still performed at 81 percent of their rated skill level. That's why a grand master can play dozens of games simultaneously, and still beat the pawns off a bunch of top amateurs. He doesn't need all that time to search for the right answers, because most of his skill is directed by something called "pattern recognition."

Top players recognize patterns on the chess board instantly, because they've seen those same patterns so often before.

And in most cases, their first hunch is the right decision. Bonds probably does the same thing when he sees the pitcher beginning his wind-up.

Human Skill

That's a very different mental function than searching through the memory, and it's why computer geeks had so much trouble coming up with a machine that could beat a grand master at chess.

"Pattern recognition is an extremely difficult problem for computers," Burns says. "Training chips to search is actually much easier from a computational point of view. A computer is fantastic at searching."

The problem gets a lot harder when the computer is asked to determine if one pattern is different from another. So today's chess computers may be as good as the top players in the world, Burns says, but they do it the wrong way. They do it by searching, not by pattern recognition, and that's a lot easier for a computer to do.

That doesn't mean searching is unimportant. Burns says even the top players in the world could lose if they had only five minutes to complete all their moves while their competitors had two hours. Searching all the possibilities gives a slight edge, but it's not what distinguishes the grandest of grand masters.

That's the ability to recognize various patterns almost instantaneously. And they learn that by doing it over and over again. Even in chess, Burns says, there are no shortcuts. Child prodigies may astonish us with extraordinary skills, but even that gift won't get them to the top.

"The way you get to be a grand master," Burns says, "is to practice a lot."

There's a neat expression for it in wood working. It's called "time on the tool."