Bishop pairs actually one of the few things which can be (sort of) quantified and does actually sort of hold up. Watson did that in Secrets of modern chess strategy with some database stats - pure endings go heavily (roughly 2:1) in favour of the bishops vs N&N or N&B. 
(Lots of very interesting discussion/examples in that book and its successor too of course.)

A bit less noted once you add more pieces in but present, probably partially because of the effects of potential endings!

Of course plenty of openings/positions where you concede them early enough to do something about it/get something to compensate for them But they're unarguably a good thing in general.

I suppose the two bishops are generally a temporary advantage, that can't usually be maintained for a long time but can be used to gain concessions from the opponent who wants to get rid of them.  In certain positions the knight pair can be lethal as well.

I think one thing classical training did leave me with was a reluctance to play f4/...f5 when necessary.

chk wrote on 10/05/11 at 13:06:27:


It's not really a misconception, it's a useful rule-of-thumb.

To me it seems it hasnt much to do with unlearning, but more that you come to notice that certain rules are special cases of broader rules. Same way that Newrtonian mechanics is a special case (same way that castling quickly is a special case of keep your king safe). 

However most rules are still applicable in 90% of the cases, it is just that in a certain position they may conflict or less important than other elements. Plus rules shouldnt prevent you from thinking, otherwise one should play checkers or apply for a job in government.

Great posts, thanks.  You are right about the GM games, some of them appear crazy compared with the classical views.  Too sharp for me generally, I like to be active of course but prefer to attack from a solid position.

The Silman book sounds interesting as well.  I've just taken up 1. c4 and am working through Tony Kosten's 'Dynamic English' and am really enjoying the positions I'm getting.  

It's ironic, I've also been working through James Vigus's "Play the Slav", and from what I've seen, the English, a less mainstream opening than the Slav, has far more 'natural' moves than the Slav book, where I'm constantly asking 'why on earth this?' or 'why not that?'.  It would be interesting to figure out how much of this tendency is due to the style of the authors' play and how much is due to the style and theory of the opening itself.

Probably anyone on any level has misconceptions that should be unlearned to progress further.

I've thought of one obvious way to do this that should also be pretty time-efficient: Play through lots of high-level games quickly while trying briefly to guess the next move or list the obvious candidate moves before looking. Whenever a move looks completely surprising/unnatural to you, you mark that position and store the game in a database (and maybe check the position with a strong engine to make sure the surprising move isn't simply bad). 

After a while you will have a personalized database of good moves that don't fit your current chess concepts! Then you can look for patterns, maybe organizing the moves by type just like you say (ridiculous lengths to preserve a bishop, exchanging a good bishop to double pawns, etc.). Simply repeating these "new" concepts occasionally should help to be more aware of them during games. Or you can go further and look for good training material on those concepts in advanced books.

In recent GM practice I've seen some concepts I could usefully unlearn: Lots of games where people play with king positions that look very shaky to me, either staying in the centre or advancing pawns in front of the castled king. Probably as long as they keep the initiative and are sure of their calculation skills, modern GMs get away with this more often than in earlier times. (Study material for this concept: Agaard; Practical Chess Defence.) I also often see GMs successfully defend positions with a bad bishop against a knight or a good bishop, even with no other minor pieces on the board, that look absolutely dreadful to me. So I should get better at recognizing when a bad bishop isn't the end of the world, and maybe in general which positions to evaluate as "worse but defensible".

Btw, I don't agree that unlearning should be easier than learning! We chess players, like everyone else, grow very attached to and comfortable with the things we "know", and it can take a long time to change well set intuitions.

Quote: