In the past, I filtered my chess databases to contain only players of 2400+ ELO. Indubitably, this tells more about the true strength of certain opening variations, but I'm discovering now that it does not give more insight into how _I_ would actually perform in each variation.
While preparing for an upcoming tourney, I decided to limit Mega Database 2007 to an ELO range of 1500-1800 and games from 1990 or later. The results shocked me. Some variations that are most popular at the GM level are virtually nonexistent at this level and the scores for some openings are almost reversed. For example, the mainline King's Indian scores 70% for Black while at the GM level, this scores about 60% for White.
The Nimzo-Indian on the other hand, scores slightly better than 50% for White. In essence, the Nimzo is far worse in terms of Black's winning chances. However, at the GM level, it is the Nimzo that scores better than the King's Indian.
It's really a different world of chess.
It's very clear that with each class level, there are certain openings that seem to be most suitable; or at least score far better than anything else. This is a strong argument against the approach of learning one system for White and Black that you play for the rest of your life. The opening repertoire has to change as your skills do. While I think this may improve tournament results, I think it would also improve one's understanding of chess, accelerating progression. I know that I learned many different lessons from 1. e4, 1. d4, and lately 1. c4. It takes more time, but it should give better results and a larger window for progress. However, this should stop at about 2000 when specialization is key.
It's also true that there are many good players out there who don't study openings. One should focus first on endgames, middlegame themes, and tactical combinations before trying to tackle too much opening theory.
The statistics are flawed also in the sense that while the Nimzo may score higher at GM level games because GMs are using them. At lower level or 1700 club level play it is being used at exactly that calibre.
Originally posted by DissidentICNThere is some debate about the role of openings. It's true that they shouldn't be studied exclusively, however, if studied well, they should give you a middlegame and endgame plan. The phases of a chess game are natural progressions and can't really be separated. Really understanding an opening can go a long way. It is not about memorization.
It's also true that there are many good players out there who don't study openings. One should focus first on endgames, middlegame themes, and tactical combinations before trying to tackle too much opening theory.
The statistics are flawed also in the sense that while the Nimzo may score higher at GM level games because GMs are using them. At lower level or 1700 club level play it is being used at exactly that calibre.
I also don't see how the statistics are flawed based on what you say. The statistics were adjusted specifically for 1500-1800 ELO and the players who use the King's Indian are not significantly better. Maybe you can elaborate...
Originally posted by exigentskyDo you think any GM could score as well as I do using the KG playing it at GM level? No way, obviously they're different worlds.
In the past, I filtered my chess databases to contain only players of 2400+ ELO. Indubitably, this tells more about the true strength of certain opening variations, but I'm discovering now that it does not give more insight into how _I_ would actually perform in each variation.
While preparing for an upcoming tourney, I decided to limit Mega Database ...[text shortened]... window for progress. However, this should stop at about 2000 when specialization is key.
Originally posted by 93confirmedI used to have my record in my profile, but I got tired of keeping track. I stopped when I had a 22-2-2 record, but there are a lot of KG players on here that know it much better than me.
You play the KG that well? I'd like to get a game some time...
It seems Chessbase messed up the statistics. I saw the great King's Indian statistics after I set it to give me an opening report on 1. c4 Nf6 2. d4 g6 3. Nc3 ... I assumed that it included all games in the database with this variation. However, it only included the games that transposed to this from 1. c4. I think this is really stupid. But in any case, while the King's Indian still scores better for Black, it only scores better than the Nimzo, the difference isn't as great when you take into account games starting 1. d4.
Originally posted by exigentskyVery interesting. What does it say about the Kings Gambit and the Scotch at those levels (I am moving back to 1. e4 and trying to decide between the two).
In the past, I filtered my chess databases to contain only players of 2400+ ELO. Indubitably, this tells more about the true strength of certain opening variations, but I'm discovering now that it does not give more insight into how _I_ would actually perform in each variation.
While preparing for an upcoming tourney, I decided to limit Mega Database ...[text shortened]... window for progress. However, this should stop at about 2000 when specialization is key.
Originally posted by zebanoWho cares if everybody plays it? It's solid, the Scotch just isn't dynamic enough for me, the KG and Ruy both are.
Because everybody else plays the Roy Lopez. I could, I could also play the Evan's but I actually have moderate experience with both the Kings Bishop Gambit and the Scotch, so it would be more of a refresher than a new repertoire.
