Random Posts

Monday, May 7, 2018

Chess Players' Thinking

     ...a cognitive psychological approach by Pertti Saariluoma. That's the title of a book I recently read; it was published in 1995. 
     The author investigated thinking, memory, problem solving and the differences between machine and human processes and presented a new view about experts' thinking and how it should be studied; a lot of technical psychological stuff. I was interested to see if the book had the secret to chess improvement. 
     Most lower rated players have read the chess is 99 percent tactics so the conclusion is that if they work a gazillion tactical puzzles a greatly improved rating is the logical result, but for many that's not the case.
     Simply learning rules and finding a solution by trial and error or by rules that are only loosely defined doesn't work very well. GMs don't rely on these heuristic rules when deciding on a move, but rather they use intuition and calculation. And, if you've ever watched a GM analyze, you have noticed that they just seem to “know” what the best moves might be and they seem to calculate at the speed of light. I've seen GMs analyzing and moving pieces around and I swear their hands were moving faster than my eyes could follow! 
     All players, even GMs, make tactical mistakes, but it's the frequency and severity that sets us apart from them. In the book, Saariluoma had Finland's GM Yrjo Rantanen analyze 400 games at four different levels in which the focus of attention was on on when an error was made that lost a piece and the number of plies between the erroneous move and when the piece was actually captured. There were four categories: 

I – the lost piece could be captured next move 
II- the lost piece could be captured within two moves by the opponent 
III- the lost piece could be captured within four moves by the opponent 
IV- the lost piece could be captured in five or more moves by the opponent 

     Even the strongest GMs made at east one tactical error per game while the lowest rated players made six or more per game. One of the charts was quite interesting. As you might expect, most tactical errors were made three or more moves out and that was the case with Masters and Grandmasters. But, here's a surprise...most players rated under 1900 committed piece-losing blunders that were only two moves away! For those rated under 1500, they were almost twice as likely to make a Category II blunder than any other! Why, I do not know. 
    In actual games when there isn't any hint as the what's hidden in the position lower rated players spend a lot of time looking at bad ideas and consequently good moves are never examined even though they might be sitting right there waiting to be played in one or two moves, making them well within calculating ability of even the lowest rated players. 
     When deciding on a move, we map out the future consequences of each move, sometimes going ahead 5, 6, or more moves.  The problem is that most of us are far more likely to convince ourselves that bad moves will work because we focus on moves by the opponent that benefit our strategy while ignoring moves that will refute it. In other words, in out analysis we have the opponent cooperating with us. 
     GMs tended put more time into thinking about their opponent's moves and could predict when the outcome of a move would weaken their position. That's what's known as falsification and it's what separates scientific and non-scientific thinking and is the best way to test your planned move. Easier said than done! 
     Research has shown that many people find falsification difficult. We seem to instinctively spend time searching for results that prove our ideas are right. Some think that the ability to falsify is somehow linked to the vast database of knowledge that GMs have and they are more likely to look for ways that things can go wrong. 
     Can the rest of learn to falsify? In a 2004 study Michelle Cowley, a cognitive scientist from Trinity College in Dublin, and her colleague Ruth Byrne studied the problem to see if falsification skill is transferable. Ireland's IM Mark Orr doesn't think it is; he thinks chess skill is specific.  Either you're good, or you're not.
     For me, not being a psychologist and not familiar with all of the psychological jargon, I can't be sure, but I think Saariluoma wasn't able to answer the question either.
     However, one thing I gleaned from the book is that if you're rated under 1500, if you can reduce blunders per game that lose a piece by half, from six to three, you can raise your rating to between 1700-1900. If you're already rated between 1700-1900, if you can halve your piece-losing blunders by half again, from three to one or two per game, you can become a master. Unfortunately, exactly how you do that is still a mystery.

No comments:

Post a Comment