BPaulsen wrote on 07/05/09 at 11:20:04:

I apologise up front if I seem to ignore certain points you make. I dont have much time to type in deep response at the moment, though my intention is to answer each point.
First Karpov made extensive use of the QGD in teh early days, usually when his opponent allowed a nimzo, though he played the nimzo as well. Furthermore it is a bit misleading to focus on those lines as he faced 1.e4 in about 60% of his games as black and there he played various Ruy lines (even a Schlieman), some Frenchies and some others (also Petrov). And that was all in the 60s.

Second you seem to think that you need deep knowledge of the openings you play in order to beat better players. I entirely disagree. The first time I beat a 2200+ player was with a sicilian line I never play as white and it wasnt a c3-sicilian, but a real main line Schevy/Najdorf. The trick for me was that instead of learning the whole Be2/Be3 complex I studied a game between Ljubo and Ulfje. Same goes for other lines. If you play a main line most moves that seem good to a 1800 player are certainly not bad. The advantage for yourself being that you came up with the moves yourself and see why another move may have been better, iso "Oh damn 11..g6 was the book move".

Third you seem to base your argument on memorisation/familiasation: If you play the same lines over and over you will be better acquainted with the possibilities and thus play them better. There are imo 3 problems with this:
1. You miss out on different types of game, narrowing you potential in the long run. You wont get really better if you dont improve on different type of positions.
2. I personally think the line of reasoning is wrong. If you play the same lines you wont get deeper understanding, you will make the exact same mistakes only later in the game.
3. People get bored (you alluded to that in an earlier post). Therefore a change is often needed in the lines.

Fourth you seem to suggest that picking a repertoire that fits you and helps you in the long run is easy. I think most developing players even with Kasparov as their coach will have real trouble with this.

Fifth (just a slight tease). I dont see people running away from the KID when they face Radjabov. I saw Kasparov running away from the KID when he faced Krammers (as he had done before with the Tarrasch and Grunfeld against Karpov). What I do see is the top players who dont vary early enough get stuck despite their talents (Van Wely being a prime example)

I am enjoying the debate in this thread. I  do tend to lean towards the argument in favour of a more specialized approach to openings, although I am not convinced that this necessarily means just one single defense to 1. e4 and 1. d4.  Two might work just as well. While speicializing in the Semi-Slav and the Nimzo might be prohibitive; playing two related  defenses does seem to make some sense. For instance many Semi-Slav players also play the Mainline Slav. You don't have to know independent lines against the exchange Slav or Slow Slav, so the work is reduced. Similarly a Nimzo/QID player might also play the QGD Tartokower. These classical defences do share a lot in common, and he might only enter the QGD via the moveorder 1.d4 Nf6 2. c4 e6 3. 3. Nf3 d5 reducing his workload.

On the other hand. I do recall reading somewhere that narrow specialization worked a lot better in pre database world where it was a lot harder to prepare against opponents. There is a school of thought that having too narrow an approach in today's database age could leave you a lot more vulnerable to preparation. Yes, one might say, if you believe in your defence strongly and know it deeply, you should not fear any preparation, but it does make your opponent's task a bit easier if he knows exactly what you are going to play. 

To add some further food for thought, Bo Hansen in his "How Chess Games are Won and Lost" suggested that your approach could be decided based on the type of tournaments you play in. By that he meant that if you play mainly in tournaments where there is at most one round a day and hence where your opponent has more time to prepare - as is typical in Europe - then a more diverse opening repertoire might be better. However, if you play largely in large American type swiss tournaments where there is typically more than one round a day and thus there is little time for preparation, then one might want to stick to a narrow repertoire where you can really reap the benefits of your specialzation. 

He also suggested that narrow specialized repertoires might be  suitable for innovative types who like to go their own way and use their own ideas. And wide repertoires may be suitable for players who primarily copy - gulp....I think that is the majority of us.  

I don't share this timeless view of theory, perhaps, I suspect, because I've been around long enough to see so many new ideas rise and old ones fall.  I admit there are certain broad constants that more or less flow from the rules of the game and the initial placement of the pieces, which include the serpentine grip of the Spanish, the power of the Queen's Gambit, and the dynamism of the Najdorf.  But even within these regal systems, there has been a tumult of change.  

At one time, 9.Qe2 was supposed to be a very significant challenge to the Open Defense, while 9.Nbd2 or 9.Be3 were footnotes; now it's reversed.  At one time, the Zaitsev Defense did not exist and Smyslov's 9...h6 was widely played.   The Spanish Exchange was once considered a fearsome weapon, now it's not.

Nobody used to play the Queen's Gambit with Bf4, now it's a big deal.  People used to sneer at the Tarrasch, now it's considered borderline acceptable.  The Exchange Variation was once considered very strong for White with Nf3 and Rab1; now that's not considered a very big challenge for Black, and most people like the Exchange with Nge2.  The Noteboom was once considered highly dubious; now people go out of their way to avoid it.

At one time the Poisoned Pawn was considered a bit dicey and 7...Be7 was usually played against 6.Bg5 e6 7.f4, while Polugaevsky's 7...b5 was also a frequent choice.  Then for a long time, the Poisoned Pawn was thought to be cast-iron for Black.  Now that system is once again under challenge.  Polugaevsky's is in the dust.  Black almost always played 6...e5 in reply to 5.Be2; now he almost always plays 6...e6.

And it's been the same right across all the openings.

When Fischer as going strong the Modern Benoni was also, and nobody understood that Taimanov's system was a major threat.  Fischer relied exclusively on Bc4 against the Sicilian, while nowadays it's considered so-so, particularly in the Steinitzian way that Fischer played it.  

The Winawer at one time was thought to be God's gift to dynamic chess, while 3...Nf6 was considered a quaint antique; then for a long time the Winawer was thought to be borderline unsound and 3...Nf6 rose in prominence.

I can remember when Nf3 in the Exchange Gruenfeld was considered to be an elementary mistake; when Alekhine's was the choice of many GMs; when the Slav was considered unassuming and dreary; when Reynold's Attack was seen by many as a positional refutation of the Meran and the Latvian Bayonet was unknown; when the Lowenthal Sicilian was the hot new thing; when 5.Qc2 was never played against the Nimzo; when 5.Bh4 was considered doubtful against the Moscow; when what is today called the Sveshnikov was considered positionally unsound; when 3.d4 was the unanimous choice of the elite against the Petroff.  I could go on and on.

My aim is not to demonstrate my personal aquaintance with these things, but merely that they have happened, in light of which, any claim of timelessness in association with given approaches to bringing out the chess pieces rings quite hollow.

The other, unrelated thing that I would observe is that the elite of the elite pay scant attention to the choices of the mere elite.  They more typically blaze trails and take significant risks in doing so.  So I'm not sure if following along behind a set of chess grubbers who desperately need an extra point or two to pay for their trip to Tounament X is the best way to go about building a repertoire.  More than those who slavishly follow Khalifman's series or the systems favored by the grubbers, I admire those, whether amateur or professional, who develop their own theory of given systems and play them out of love, creativity and faith.

Just a few observations on this debate.

You might want to consider that it is easier with computer databases to get the games necessary to review an opening - certainly in comparison to the 1970s.

Secondly, I do feel that Karpov playing 1 d4 had an incredibly narrow but successful repertoire- think of his 4g3 v the queens indian in particular, the samisch v the kings indian (acknowledged he later in the 90s played fiancetto, occasional main line nd2/b4 and be3 system), the queens gambit main lines, 

Thirdly with 1c4, it's incredible how many seem to transpose to either a QGD, KID, Slav, Semi Slav or Dutch. If with 1c4 you aimed to transpose the amount of work to convert to 1d4 would mean losing a lot of previous knowledge and experience.

Fourthly is the old conundrum about how much time do you have to play critical lines where preparation is important eg You can't really play the poisoned pawn and yugoslav dragon without the necessary studying. A related issue is it better to play a respectable line that you know well over the latest most topical line and also how easy is going to be to get your line. For example, until recently I would agree with Danny King saying that Bg5 v Najdorf was a rare opening to face ! (Radjabov has changed my view on that).

Finally, occasionally if you play sharp lines a bazooka hits you and you have to change eg in the Zaistev, Kasparov broke Karpov's ba, a5 and ra6 system totally. You do need a line in these occasions that isn't going to be refuted. Perhaps the Chigorin Ruy is a good example as advocated by Marin.

MNb wrote on 07/04/09 at 20:56:18:

Even Bobby varied. The thinking that the greats didnt vary their openings or their play in general is imo nonsense. I think a lot of it is due to perception. We all know Bobby played the KID and Grunfeld and the Najdorf as black, dont we? Yet he played Benonis, QGD, Nimzos, Alekhines as well. It is true that he started with a narrow repertoire, but the moment he got real opposition he started varying as he got into trouble when he didnt (just check what happened at the candidates tournament of '59).

BPaulsen wrote on 07/04/09 at 14:33:40:

Not really. His basis was the QGD and Ruy as black, but he definately experimented with other lines when he wanted (eg he has a surprising amount of Taimanov sicilians before he became wch). He also changed his repertoire for matches, where he would prepare specific lines aimed at his opponent's weaknesses (eg the CK against Spassky).
Quote:

I think it all depends on the way you work with openings.
Quote:

I know (as in saw them at tournaments) quite some former youth players who varied their openings a lot (or rather they didnt have a repertoire at all, they played openings based on their interest at that moment). Eg a chap like Stellwagen didnt only play the French, he played quite a few sicilians as well.
Quote:

There's quite a difference between having a very diverse repertoire and not having a fixed repertoire at all. 
Quote:

I am of the opinion that having a fixed and consistent repertoire while developing does more harm than good in most cases. Most dont know which lines are good for them and have no idea on what basis to choose. 
You dont let 12-year-olds decide what they want to do in uni, they first must get a proper preparation and self-knowledge. Selecting openings before having a deep "chess education" is imo similarly wrong.
Quote:

It really depends. If you use your games and thus your openings as a means to get better (by really analysing your own games), I think playing very diverse openings gets you different middle games and endgames, giving you far more ammo to improve.
Ie if you play 10 games in the Ruy as black with the Zaitsev in a relative short period or you play 10 different openings I think you will benefit more from a deep analysis of the latter and learn more.

Lastly I think most players below 2200 level dont know their openings very well and it wont matter what you play against them, before move 10 both will be out of book.

BPaulsen wrote on 07/03/09 at 13:49:46:


Now I come to think of it, even Fischer is not a convincing example. There is no post-war WCh who played less serious games. Even Botvinnik played more. So Fischer's repertoire just did not get the opportunity to wear out.

This is an interesting repertoire : 

As black you play as GM Evgeny Gleizerov

Against 1.e4  you play  1...e6    The french defence.

Against  1.d4  you play 1...e6   Go for the Stonewall dutch or the french.


Against   1. Nf3 you play 1...d5 
Against   1.c4    You play 1...e5 (pick a line that restricts white as much as possible....se GM Gleizerov)

Against  1.f4  you play  1..f5!? (better than it looks and maybe only 2% of all players facing the Bird play this move)

As white you play  1.f4 !? (Polar Bear system by GM Henrik Danielsen) Just to restrict black as much as possible.

This repertoire will not take all your study time for just theory. You have time to study the beautiful games of the worldchampions. Study tactics,endgames and so on.

Good luck!