Practice blindfold chess? Research says no! (Openings)

by mindcramp, Tuesday, July 24, 2007, 13:38 (6479 days ago)

I just came across a 2005 research paper by Gobet and Jansen titled "Training in Chess: A Scientific Approach." (You can download a PDF from the web.) It contains a lot of guidelines for improving one's play via training, and backs up their recommendations with findings from psychology, neurophysiology, etc. Nothing really surprising in their conclusions, except this: they specifically recommend *against* training to acquire blindfold chess ability. They say that blindfold chess ability comes almost automatically once the player has acquired enough "chunks" and "templates" (=pattern recognition and deeper understanding); unless the player has reached that point, it is not efficient to practice it.

Actually I disagree with one other point. The authors recommend sticking to a very narrow opening repertoire, and practicing not only the opening lines but also the typical middlegames and endgames that result from those opening choices. This, I believe, may be good for short term results, but could be detrimental to one's long term development as a chessplayer. At some point, an aspiring player ought to learn how to play with and against an Isolated Queen Pawn, conduct a minority attack, launch an h-pawn against a kingside fianchetto, maneuver against a hedgehog, etc. Which is to say, acquire a rich collection of chunks and templates, including the most important ones.