12-05-2012, 10:24 PM
well, I can't speak for gelfand, but I'd imagine there are a few other reasons he isn't blitzing out his moves even when he's presumably still in prep (although I doubt he is going to spend exactly the same amount of time on all his opening moves, and if he really doesn't have a scooby about a particular position then I'm sure he would spend even more time on it), although the main reason might well be so anand doen't know when he's out of book, although not all people agree with this - I'm sure kasparov said something in his book how life imitates chess about some people having a similar habit to what gelfand might be doing, but that he was so confident of his preparation that he didn't care whether his opponents knew whether he was in preparation or not (quote from kasparov after one of his world championship matches with anand - "I spent 2 minutes during the game over the first 20 moves - but 48 hours beforehand").(<!-- m --><a class="postlink" href="http://www.chessgames.com/perl/chessgame?gid=1018625">http://www.chessgames.com/perl/chessgame?gid=1018625</a><!-- m -->)
As has already been mentioned there's the possibility that he is warming up his brain (alan minnican (I think - might have been someone else) told me at the scottish about a game at an olympiad where rowson (when he was a lot younger) was playing some ex-soviet gm in the grunfeld, and the game followed this theoretical line for ages, and rowson knew his theory really well (and presumably so did his opponent) and was blitzing out his moves instantly, whereas the other guy was really taking his time and thinking deeply over every move from a relatively early stage and as soon as rowson was out of book then he started burning loads of time whereas the other guy was already in the zone and just kept up his fairly steady pace and won and the reason the GM had been taking so much time in the opening was because he was gradually getting his brain into higher and higher gears or something like that.)
Dvoretsky has some saying along the lines of "trust but verify" which he uses to say that you should never just blindly follow remembered (or more likely half-remembered!) analyses, particularly other people's. He also says that "yusupov usually verifies any preliminary analysis over the board" - secrets of chess tactics, so gelfand could be checking that there are no holes in his analysis or improvements, although perhaps all of that is irrelevant now that we have computers (so long as he trusts his memory).
Gelfand is certainly capable of finding strong novelties in the opening at the board(19.Rc7 in this game:http://www.jeremysilman.com/book_reviews_jw/jw_my_most_memorable_games.html is one example of a novelty he claims to have found over the board) so perhaps he is looking for improvements - I'm sure I got told by someone at my chess club of a game between leko and kramnik in the ruy lopez marshall where where kramnik did not make a single move that he had not prepared with whatever engine he used but when he followed his preparation to the letter (and leko was out of his preparation) then it turned out his preparation was losing and leko managed to drum up a mating attack which the computer had missed as it had been too long-term given the amount of time the computer had been given.
The moral of the story was that as there can be so many critical positions to analyse in preparation,particularly of such complex openings as the marshall, that the computer had only been given a small amount of time to evaluate that particular variation, so perhaps even nowadays the trust but verify approach still applies to a certain extent, even in world championship matches?
This is all speculation obviously, and it could just be that gelfand was out of his preparation and we are just surprised at this and reading too much into it!
As has already been mentioned there's the possibility that he is warming up his brain (alan minnican (I think - might have been someone else) told me at the scottish about a game at an olympiad where rowson (when he was a lot younger) was playing some ex-soviet gm in the grunfeld, and the game followed this theoretical line for ages, and rowson knew his theory really well (and presumably so did his opponent) and was blitzing out his moves instantly, whereas the other guy was really taking his time and thinking deeply over every move from a relatively early stage and as soon as rowson was out of book then he started burning loads of time whereas the other guy was already in the zone and just kept up his fairly steady pace and won and the reason the GM had been taking so much time in the opening was because he was gradually getting his brain into higher and higher gears or something like that.)
Dvoretsky has some saying along the lines of "trust but verify" which he uses to say that you should never just blindly follow remembered (or more likely half-remembered!) analyses, particularly other people's. He also says that "yusupov usually verifies any preliminary analysis over the board" - secrets of chess tactics, so gelfand could be checking that there are no holes in his analysis or improvements, although perhaps all of that is irrelevant now that we have computers (so long as he trusts his memory).
Gelfand is certainly capable of finding strong novelties in the opening at the board(19.Rc7 in this game:http://www.jeremysilman.com/book_reviews_jw/jw_my_most_memorable_games.html is one example of a novelty he claims to have found over the board) so perhaps he is looking for improvements - I'm sure I got told by someone at my chess club of a game between leko and kramnik in the ruy lopez marshall where where kramnik did not make a single move that he had not prepared with whatever engine he used but when he followed his preparation to the letter (and leko was out of his preparation) then it turned out his preparation was losing and leko managed to drum up a mating attack which the computer had missed as it had been too long-term given the amount of time the computer had been given.
The moral of the story was that as there can be so many critical positions to analyse in preparation,particularly of such complex openings as the marshall, that the computer had only been given a small amount of time to evaluate that particular variation, so perhaps even nowadays the trust but verify approach still applies to a certain extent, even in world championship matches?
This is all speculation obviously, and it could just be that gelfand was out of his preparation and we are just surprised at this and reading too much into it!