Caro-Kann Advance Variation

  • IM Silman
  • | Jun 13, 2010

James D asked:

I seem to have come to the conclusion that the advance variation against the Caro-Kann is the most aggressive, but what then? I’ve tried various ways to play the Advance for White, but the latest is 1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5 which basically provokes black to play …Qb6, forking my pawns.

I managed to play this against someone and win:

1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5 Qb6 6.Bd3 Bxd3 7.Qxd3 Qxb2 8.e6 Qxa1 9.Qb3 Qxd4 10.Qxb7 fxe6 11.Be3 Qe4 12.Nf3 Kd8 13.O-O g5 14.Nc3 Qc4 15.Qxa8 Qxc3 16.Rb1, 1-0.

However, I’m under the impression that this kind of crazy opening is not one that I should use as my primary weapon against the Caro-Kann. Should I keep playing it?

Dear James D:

Though 5.c4 is white’s most popular move (992 games in my database), 5.Bg5 has scored insanely well for White in recent years – my database has 18 games featuring 5.Bg5, with White winning 16 and drawing two. Even more telling is that your game against “someone” followed M.Perunovic (2580) - B.Lalith (2480), 11th Dubai Open 2009 up to black’s 11th move: 1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5 Qb6 6.Bd3 Bxd3 7.Qxd3 Qxb2 8.e6 Qxa1 9.Qb3 Qxd4 10.Qxb7 fxe6 11.Be3 Qa4 12.Qxa8 Qa5+ 13.Nd2 Qd8 14.Ngf3 Nf6 15.0-0 g6 16.Rb1 Nbd7 17.Qxc6 Bg7 18.Ng5, 1-0.

Clearly you should quit playing this line for White immediately! After all, if you win every game too easily, chess will become boring.

Of course, not everyone will go into this stuff as Black. After 1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 Black has interesting alternatives to 4…h5 like 4…Qb6!? and the popular 4…h6!? when 5.g4 Bd7 (5…Be4!? and 5…Bh7!? also get air time now and then), played by both Leko and Ivanchuk, has proved to be a solid choice.

Naturally, the Advance Variation isn’t joined at the hip with 4.h4. If White gets tired of pushing his h-pawn, other moves are plentiful and successful: 4.Nf3 leads to both positional and sharp lines, 4.Nc3 is a prelude to one of the most tactical systems in chess, and less common moves like 4.Bd3, 4.g4, 4.Ne2, 4.Be3, 4.c4 4.c3, 4.Nbd2, 4.f4, and 4.Be2 also have their grandmaster devotees.

To answer your question, you should ride the 5.Bg5 wave until someone cuts it down to size, or until you find you don’t like what you’re getting vs. 4…h6. Then you can give all those other 4th move White alternatives a try! Simply put, the Advance Variation is one of white’s best ways to deal with the Caro-Kann, so why not keep milking it?


Jemptymethod said (commenting on an article I wrote about the Budapest Gambit):

As a long-time Budapest/3...Ng4 player, I started flirting with the Farajowicz (3…Ne4) two or three years ago, and I can tell you with certainty: it’s not any good. I’ve used chess engines to contradict the analyses of both Moskalenko and Gutman in many lines, not just 4.Nf3. 

Dear Empty:

Most strong players don’t trust 3…Ne4, but you state that it’s no good because you’ve “used chess engines to contradict the analyses of both Moskalenko and Gutman.”

I know you meant well by this, and that you were trying to give good advice to other readers, but please take a moment to think about what you said. Tick, tick, tick, tick, tick. Done with that moment of Zen? Okay, now we can analyze this whole situation.

First off, it’s you vs. two strong grandmasters. Clearly, you’re outnumbered and vastly outgunned, but … you make it clear that your use of a computer gives you an edge. However, ponder this: don’t you think that both of these grandmasters also use computers? And, when you mix their use of computers with their vastly superior strength, their overwhelming superiority in basic chess understanding, and their ability to lead a computer in the right direction (something lesser players aren’t able to do), how can you even entertain the notion that you have a pristine vision of the board that this pair of titled giants lack?

I should add that Gutman is still writing articles to this day defending the Farajowicz. Are you going to say he’s wrong because your engine is bigger than his? That might impress the non-chess playing ladies, but those in the know will just widen their eyes, snicker, and walk away.

Now, I’m not arguing with your opinion that 3…Ne4 sucks – it probably does. But note that I said, “probably.” How can I insist that Moskalenko and Gutman are just wrong when both are stronger players than I am, both have devoted infinitely more time to this opening than me, and both undoubtedly use every chess engine money can buy to support their analytical views? At the very least, I can voice my opinion in a polite way and not bash my chess superiors.

The advent of computers, and the illusion they give to the masses who think their computer’s analysis is their analysis, is very destructive. It illogically strips grandmasters of the little respect they have, creates a false view of chess knowhow in the amateur’s mind, and ultimately (if engines aren’t used properly) stunts the chess student’s growth as a player by making him believe he understands things that he clearly doesn’t.

When going over a game with a student, he will often say something like, “I’m better here.” I’ll ask why, and he’ll say, “My computer says I’m 0.48 ahead.” I’ll insist: “Yes, but why are you better?” And, in many cases the student won’t have a clue. Of course, the computer might be right, and at other times it might be wrong – that doesn’t concern me. What does concern me is the student increasing his understanding of the game, and mindlessly parroting a computer’s numeric is a surefire road to chess ruin.

In future, if you intend to make a sweeping statement or challenge the analysis of a professional player who has devoted his life to the game (and I welcome people to do so since analytical errors are extremely common), please offer an analysis mixed with a polite query. That way a meaningful dialogue can begin.

Kasabiian said (commenting on the same article):

Gutman’s analysis is p00p.

Dear Mr. Kasabiian:

Are you a stronger player than grandmaster Gutman? Or are you parroting what your computer said, never imagining that Mr. Gutman is also making use of modern technology? Instead of thinking things through in a logical manner, you just slander the guy’s work (offering us nothing but childish rudeness) and expect us to bow down in awe and wonder.

For those that think chess professionals don’t own computers, don’t have databases, and don’t have chess engines, you somehow figured us out! I admit it, you win, our dirty secret has finally come to light! We communicate via smoke signals and drums, and write by etching letters onto stone tablets. Your magical boxes with their strange “chess engines” are a mystery to us – supernatural, scary things you make use of to refute everything we say.


  • 5 years ago


  • 6 years ago


    OMGOSH!  My ears are burning!! Sealed

  • 6 years ago


    There is also a difference between 'busting' an opening with a computer and losing to rybka 3 twenty moves after the opening position.

  • 6 years ago


    I am getting interested with this line but what if 9... b5 instead of 9... Qxd4 defending the b pawn to avoid conters from white. I don't seeing the queen being trapped because of 10... a5 11... a4.

    Sorry Rybka is not around for this.

  • 6 years ago


    Gee. Looks like some people need to play a few OTB games against titled players, reassess the situation, and then start posting on internet forums.

  • 6 years ago


    @ nerv - it's not... that's just the engine from the game.  my chess programs aren't that out of date

  • 6 years ago


    Quoted from  "A snob is someone who adopts the worldview of snobbery — that some people are inherently inferior to him or her for any one of a variety of reasons, including real or supposed intellect, wealth, education, ancestry, taste, beauty, et cetera."

    RaisetheCurve, or IhavesomeNurve, or whatever people call you.. I have a feeling Mr. Silman does not share your same affliction.

    I think my friendly advice of the day would be "A sense of humor goes a long way in this world."

  • 6 years ago


    Getum Silman!

  • 6 years ago


    Dear Mr. Silman,

    As an intelligent, aspiring chess player I grow increasingly weary of articles (not just by yourself, granted) which give voice to the lowest rung on the tree of knowledge. So you put an immature 10 year old, or a mindless middle aged hobbyist in his/her place. What have you accomplished with your writing other than to give recognition to someone who deserves it not?

    As a professional I'm sure your time writing is better spent educating those who wish to learn from an esteemed individual, rather than mocking the fish who waste their time, and wish to bring the rest of us down with them.

  • 6 years ago


    to baconisdelicious: Since when Fritz 4 is a program of GM strenght?

  • 6 years ago


    @FirebrandX - I have not used Deep Rybka 3, and while it is good to see that engines are getting better at this sort of position, there are still many positions that computers are not good at playing - especially closed positions and against positional sacrifices.  I would like to see how Rybka follows up against 9. ... Nb4!, but I suspect that it still doesn't play perfectly because it is still relying on calculation and is incapable of really learning strategy.

  • 6 years ago


    Good article indeed.

  • 6 years ago


    quote from baconisdelicious:

    "For example, after 1. d4 d5 2. c4 dxc4 3. e4 c5 4. d5 Nf6 5. Nc3 b5 6. Bf4 Qa5 7. e5 Ne4 8. Nge2 Na6 9. f3 Nb4!, every computer engine I've used puts white more than a pawn ahead, and yet I always win in this line as black, even against GM-strength computers. "

    That's funny, when I put the same line into Deep Rybka 3, it evaluates the position as advantage to black at -0.40ish. I guess you must mean "even against GM-strength computers... except Rybka".

  • 6 years ago


    I like going through Botvinnik's games- esp against Tal, which featured the Caro-Kann as black, and he would confront the advance variation every once in a while.  Amazing that he used that make World Champ in the 1950s.

  • 6 years ago


    jemptymethod - You put too much faith in computers.

    For example, after 1. d4 d5 2. c4 dxc4 3. e4 c5 4. d5 Nf6 5. Nc3 b5 6. Bf4 Qa5 7. e5 Ne4 8. Nge2 Na6 9. f3 Nb4!, every computer engine I've used puts white more than a pawn ahead, and yet I always win in this line as black, even against GM-strength computers.  Fritz lost a game as black from this position.

    I don't want to diss computers - they play chess very well in most positions - but some positions require an understanding that computers are (currently) incapable of having, and in those positions, they cannot be relied upon.

    That said, the Farajowicz seems dubious (as does the entire Budapest Gambit), but you should still have respect for the opinions of the experts.  I am sure that Moskalenko and Gutman could both beat you using the Farajowicz, even if it is bad.  You should try to learn from their analysis to try to figure out why they like 3. ... Ne4.  You can't really complain that an opening is bad if you do not understand why somebody is playing it.

    As for your complaint that Silman is calling you "Empty" - you should change your name to something that does not include "empty" if you don't want to be called that.

  • 6 years ago


    I love your articles because of how they're written. The irony and the sarcasm make it so much fun to read! Thanks for another article.

  • 6 years ago


    Black is going to be attacking the central pawns of white.  In particular, the d4 pawn.  If white isn't going to defend it, then it will be lost.  Bg5 is too early, it needs  to be played, but not that early.  As Black, when I play the Caro-Kann, which I do quite a bit, the earlier I strike at white's pawns, the better.  Also, e6 needs to be played as well.  Nf3 should be played to defend the central pawns for white and then Bg5 would work for white.  White also needs to realize that c5 is something black needs to play to free its game.


  • 6 years ago


    jemptymethod, Silman isn't arguing whether or not your engine's analysis of the Fajarowicz is wrong, but simply warning readers of the dangers of depending on an engine in general. I do not understand why your replies need to be so pointed.

  • 6 years ago


    haha! that tongue and cheek letter and response was quite entertaining - keep up the good work silman.

  • 6 years ago


    I play the caro-kann defence a lot (I'm approximately 1500 rating) but never have I encountered 4. h4 h5 it just seems like a very weird sort of move to do... What is the reason for it?

Back to Top

Post your reply: