IM Silman
May 9, 2011, 12:00 AM |
33 | Opening Theory

member Matt Chapman asked:

I’m looking not so much for an answer to this question (unless you happen to have the knowledge, which would be fantastic) but a reference to someone who might know (and indeed, how to contact them). You see, I’m having trouble with one particular variation of the mainline Caro-Kann, and I can’t see any way out of it for Black!

The line begins, 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 e6 11.Bf4 Qa5+ 12.Bd2 Qc7 13.O-O-O Ngf6 14.Ne4 O-O-O 15.g3 Nxe4 16.Qxe4 Bd6 17.c4 c5 18.d5 Nf6 19.Qc2

In GRANDMASTER SECRETS: THE CARO-KANN, Peter Wells suggests Black is holding after 19...exd5 20.cxd5 Rhe8 21.Bc3 Qd7 22.Bxf6 gxf6 since 23.Nh4 fails to deliver against 23...Re5!

The problem is, White can delay Nh4 a move by 23.Kb1!? and now Black has some real problems. All roads now lead to Rome for White, as he seems to have an overwhelmingly strong position and nothing Black tries, be it 23...f5, 23...Kb8, 23...Qg4, or my own try 23...Re7 give Black anything I could really define as “hope”.

Apologies for no diagrams with a fair few moves, but perhaps you can see my dilemma. This seems to be quite an issue to me since it arises from the main line of the Caro-Kann, but there must surely be improvements along the way, or else this line seems quite dead.

Dear Mr. Chapman:

To me, this is a clear case of too much opening. In other words, the opening demands more from you than you’re able to give. Many players love the idea of playing main lines, and if they have good memories they can follow theory quite far. However, these openings are often very cutting edge, and very tricky. This means that the theory is often changing day by day, and if you’re not strong enough to stand up for the system, then it’s not the right opening for you.

Please don’t think I’m putting you down. I have no idea what your rating is, but even a master (2200) would likely find that he’s biting off more than he can chew here. Personally, I encountered many lines that I felt were more opening than I was willing to deal with – they demanded a perfect memory and non-stop work to keep them afloat. Instead, I used systems that were more in line with my memory and available time.

Another problem with this kind of system is that, unless you play in very strong events, you’re rarely going to encounter a theory-fest that will take you to a starting position that begins on move 22 and continues on – at times even to an endgame! But if you play these systems, you’ll need to be prepared for just that (in case)!

If you wish to employ monster-theory openings, you need up-to-date books and databases, and other theory sources too (like Informants and New In Chess Yearbooks). You mentioned the Caro-Kann book by Peter Wells (who is a great chess writer), but though it came out in 2007, it’s already outdated. You simply can’t count on older sources to safely get you through modern mega-theory! 

Having said all that, I’m also a lover of the Caro-Kann and can understand one wanting to plumb the depths of these positions. Nowadays kingside castling for Black is highly thought-of (leading to a very sharp and interesting battle), but of course the classic queenside castling lines are still completely viable. Though I’m not up to snuff on my theory anymore, I do keep an eye on theoretical changes (reading through various opening books blindfolded as I walk my cat).

The line you’re discussing made a big splash when it first appeared, and it terrified many Caro-Kann devotees. However (to the best of my knowledge … of course, I might be wrong), it’s no longer the fearsome beast that it once was. I rarely do this, but I’m going to feed you the theory avalanche that you so badly want. I try and avoid doing so since it’s so advanced that it can be viewed as totally useless to the vast majority of players. But perhaps it can serve a non-theoretical function too: perhaps it will let members get a glimpse into the kind of preparation one must do at a professional level.

In general (when I was still playing), I tried to know all the basics of an opening, understand what all the typical pawn structures meant, knew all the typical middlegame plans, and also was up on many typical endgames. But when it came to cutting-edge theory lines, just memorizing the basics wasn’t good enough: I also had to have new moves and ideas ready for over-the-board tryouts. And to get these new moves, I would sometimes spend weeks working on a new idea (imagine how depressing it was if you worked for 2 weeks on something, only to conclude that it sucked).

The following analysis makes use of the latest theoretically important games I could find, but even so, I had to do a lot of personal exploration too. Since I certainly don’t have time to do a proper job of it, use my stuff as a guideline, but don’t take any of it as gospel (there will, without any doubt, be errors)! My conclusion: black’s okay!

For those that hate this kind of stuff, I promise that next week’s column will be far more “human.”

C. Sandipan (2522) - K. Asrian (2646) [B19]

4th Aeroflot Open, Moscow 2006

1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 e6 11.Bd2

11.Bf4 Qa5+ 12.Bd2 Qc7 transposes, though the move numbers show one move more with this order. However, White often avoids this move order since instead of 12…Qc7 Black can also play the theoretically important 12…Bb4. Of course, that’s a completely different animal, which has nothing to do with our reader’s question.


Transposing to the previous note (11.Bf4 Qa5+ 12.Bd2 Qc7) and also showing that he’s going to castle queenside. Nowadays, many black players prefer to castle kingside via 11…Nfg6 (…Qc7 isn’t necessary if you’re going to castle kingside!) 12.0-0-0 Be7 followed by …0-0. This leads to a very sharp game where White can go right for black’s King and Black can initiate sharp play on the queenside against white’s monarch. In general, Black is doing quite well with the plan of kingside castling.

12.0-0-0 Ngf6 13.Ne4 0-0-0 14.g3 Nxe4 15.Qxe4 Bd6

Another sharp, risky, but highly interesting line is 15…Nf6 16.Qe2 Rd5, taking aim at the h5-pawn. 

16.c4 c5 17.d5 Nf6

The main line, but 17…Rhe8 is a very important alternative. White has the following replies:

* 18.Rhe1 Nf6 19.Qc2 exd5 20.cxd5 Qd7 and White was already worse due to the weakness of his pawns on d5 and h5, M. Golubev (2531) - O. Peyrat,Olivier (2272), Paris 2004.

* 18.dxe6 Rxe6 (18...Nf6, =) 19.Qf5 Rf6 20.Qg4 Qc6 21.Nh4 Rxf2 22.Be3 Rf6 23.Nf5 (23.Qxg7? Qe4 is much too strong) 23…Re8 (23…Bf8 is pretty good) 24.Rhe1 Bf8 (24...Re4 is also good) 25.Nxg7, S.S. Ganguly (2456) - G.B. Prakash (2453), New Delhi 2001, and now 25...Re4! 26.Qxd7+?? (26.Bf4 Rxc4+ 27.Kb1 Re4 28.Kc1 Rc4+, =) 26...Qxd7 27.Rxd7 Kxd7 28.Rd1+ Kc6 wins for Black since the g7-Knight is trapped.

* 18.Bc3 exd5 19.Qxd5 Be5! is fine for Black: 20.Nxe5 Nxe5 21.Bxe5 Rxd5 22.Bxc7 Rxd1+ 23.Rxd1 Kxc7 24.Rd5 Kc6 25.Rf5 f6 26.Kd2 Re4 27.Kd3 Rd4+ 28.Kc3 Rg4 29.Kd3 a6 30.a3 b6 31.Kc3 Re4 32.Kd3 Rd4+ 33.Kc3 Rg4, 1/2-1/2, O. Zimina (2376) - A. Iljin (2456), St Petersburg 2002.

* 18.Qg4! is white’s most challenging reply. Black has two ways to deal with it, and both demand a lot of personal analysis:

* 18…Ne5 19.Nxe5 Bxe5 20.Rhe1 Bd4 21.dxe6 fxe6 22.Be3 Bxe3+ (22…Bf6 23.Rxd8+ Qxd8 24.Qg6 [24.Rd1!?] 24...Re7 and Black’s pawn structure gives White a slight advantage, but no more than that.) 23.Rxe3 Rxd1+ 24.Kxd1 Qd7+ 25.Ke2 e5 26.Qxd7+ Kxd7 27.Re4 Ke6 28.Rg4 Kf7 29.Rg6 Rd8 30.b3 Rd4 31.f3 e4 32.f4 Rd3 33.f5 Rc3 34.Rd6 Ke7 35.Rg6 Kf7 36.Rd6 Rxg3 37.Rd7+ Kf6 38.Rxb7 Kxf5 39.Rxa7 Kf4 40.Rf7+ Ke5 41.Ra7 g5 42.hxg6 Rxg6 43.Ra5 Kf4 44.Rxc5 Rg2+ 45.Kf1 Rxa2 46.Rh5 Rb2 47.Rxh6 Rxb3 48.c5 Rc3 49.c6 Ke5 50.Ke2 Kd5 51.Kd2 Rxc6 52.Rxc6 Kxc6 53.Ke3 Kd5 54.Ke2 Kd4 55.Kd2 e3+ 56.Ke1 Kd3 57.Kd1 e2+ 58.Ke1 Ke3, 1/2-1/2, D. Neelotpal (2370) - F. Izeta Txabarri (2475), Benasque 1998.

Well, that seems playable for Black. But the real problem rears its ugly head after 18…Ne5 19.Nxe5 Bxe5 20.Rhe1 Bd4 21.dxe6 fxe6 22.Rxe6! Qd7 23.Rg6 Qxg4 24.Rxg4 when one really has to put in some overdrive analysis if they want to see if it’s possible to save the game. I certainly can’t give a definitive answer to this, but any endgame stand from this position demands some super-grandmaster endgame knowledge. For example: 24…Re2 25.Bc3 Bxc3 26.Rxd8+ Kxd8 27.bxc3 Rxf2 28.Rxg7 Rxa2 29.g4 Ke8 30.Rxb7 a5 31.Rb5 a4 32.Rxc5 Rg2 33.Ra5 Rxg4 34.Rxa4 Rg5 35.Kc2 Rxh5 with a theoretical draw, even if Black doesn’t have his h-pawn!

Another similar position that might occur is the following:











White to move, draw. However, if black’s King is on g6 with White to move, then White wins.

Even if Black can prove a forced draw after 18…Ne5 19.Nxe5 Bxe5 20.Rhe1 Bd4 21.dxe6 fxe6 22.Rxe6!, it’s not fun at all and requires extremely advanced skills plus endless hours of home preparation. Thus our next move needs to be looked at.

* 18…exd5!? This obvious move can lead to some highly tactical situations that, again, demand reams of home analysis before it can be tried. Here’s a quick look by me, which almost certainly contains all sorts of errors. However, it’s not meant to be definitive, it’s meant to give you an idea of the types of positions that can occur. White can answer 18…exd5 with 19.cxd5 or 19.Qxg7:

** 19.Qxg7 dxc4 20.Qxh6?! (20.Rhe1 Qc6 21.Rxe8 Rxe8 22.Qxh6 Qxf3 23.Qxd6 Re6 24.Qf4 Qxh5 25.Qxc4 is more or less equal.) 20...Qc6 21.Ng5? (Better is 21.Rhe1 though Black has an edge after 21…Rh8 22.Qe3 Rxh5) 21…c3! 22.Nxf7 (22.bxc3 Ne5 is extremely strong for Black, while 22.Bxc3?? Bf4+ picks up the White Queen) 22…Ne5!! (22...cxd2+ 23.Qxd2 Bc7 24.Nxd8 Rxd8 25.Qd5 favors White due to his mass of kingside pawns.) 23.Nxe5 (23.Bxc3 Nxf7 gives Black an extra piece, though this is probably white’s best chance; 23.Nxd8 Nd3+ 24.Kb1 Qe4 wins for Black) 23…cxb2+! (23…cxd2+ 24.Qxd2 Bxe5 also wins, but 23…cxb2+! is much crisper) 24.Kxb2 Bxe5+ 25.Kc1 Qb5 and white’s going to get mated. Fun!

** 19.cxd5 Qb6 (I don’t trust passive moves here, so this Queen move injects a bit of spice into the position) 20.Bc3 (Black’s okay after 20.Qxg7 Re2 21.Qxh6 Qa6 22.Rhf1 [22.Rh4!?] 22…Qc4+ 23.Kb1 Ne5 24.Ng1 Ng4 25.Qg5 Re5 26.Qf4 Qxf4 27.Bxf4 Rxh5 28.Ne2 Rh2 29.Rc1 [29.Bxd6 Rxd6 30.Nc3 Nxf2 31.Rd2 Rf6 is also interesting] 29…b6 30.b4 Kb7 31.Bxd6 Rxd6 32.bxc5 Rxd5, =) 20…Qa6 21.a3 Re2 22.Rd2 Bf8 23.Qf5 (23.Bxg7?? f5) 23…f6 24.Rxe2 Qxe2 25.Nh4 Kb8 26.Qf4+ Ne5 27.Re1 (27.Ng6? Bd6 28.Nxe5 Bxe5 29.Bxe5+ fxe5 favors Black.) 27...Qc4 28.Qxc4 Nxc4 29.Ng6 Bd6 and though White has an edge here (an eventual Ng6-h4-f5 will prove annoying), it’s certainly not the end of the world for Black.

There’s still a lot to discover with 17…Rhe8! The position is rich and interesting.

18.Qc2 exd5

Worse is 18...Rhe8? 19.Bc3 exd5 (19...Kb8 20.Rhe1 Nxh5 21.dxe6 Rxe6 22.Rxe6 fxe6 23.Qg6 was clearly better for White in Zarnicki - Hernandez Penna, Buenos Aires 1994) 20.Bxf6 gxf6 21.Rxd5 with an obvious plus.

19.cxd5 Rhe8 20.Bc3 Qd7 21.Bxf6 gxf6

To quote grandmaster Lukacs: “At first glance white’s position is clearly better because of the weak black f-pawns. However things are far from clear. Black can organize good counterplay on the e-file and white’s h5-pawn may become weak.”


Wow … we’ve finally reached the position you were asking about!

Not as good are:

22.Rhe1 Rxe1 23.Rxe1 Qg4 24.Nh4 Qg5+ (now white’s struggling to equalize) 25.Kb1 (25.Qd2 Bf8) 25...Qxd5 26.Nf5 Bf8 27.Qe4 Qxe4+ 28.Rxe4 Rd5 29.Re8+ Rd8 30.Re4 Rd5 1/2-1/2, D. Boskovic (2475) - B. Bournival (2247), Foxwoods Open 2007.

22.Nh4 Re5! – This position has been reached in a ton of games and has pretty much been analyzed to a draw. One example: 23.f4 Rxh5 24.Qe2 Rxd5 25.Rxd5 Bxf4+ 26.gxf4 Qxd5 27.Rd1 Qxa2 28.Rxd8+ Kxd8 29.Nf5 Qe6 30.Qd3+ Kc7 31.Nxh6 Qe1+ 32.Kc2 Qh4 33.Nxf7 Qxf4 34.Qd8+ Kc6 35.Qe8+ Kb6 36.Qd8+ Kc6 37.Qc8+ Kb6 38.Qd8+ Kc6, 1/2-1/2, Z. Hracek (2606) - I. Khenkin (2613), 18th Croatian Team Ch. 2009.


This seems to me to be perfectly playable.

Lukacs mentioned 22...Re7!? 23.Nh4 Rde8 24.Qf5 Re2 25.f4 Qxf5+ 26.Nxf5 and he feels White is better since the Knight looks like a god. However, after 26…Bf8 Black is holding everything together. Black will follow up with …Kd7 when it’s hard to see how White can make progress against black’s solid defense and seventh rank domination. However, since I’ve never seen one single game with 22…Re7, I can only guess that it’s badly flawed. I would love to see what that flaw is, though!

Another interesting (but apparently flawed) move is 22…Kb8, which was largely based on a long tactical sequence: 23.Nh4 Qg4 24.Nf5 Re2 (this appealed to a lot of players since Black seems to have very active pieces) 25.Nxh6 Qf3 26.Qf5 Rxb2+ (this was long thought to force a draw, but it seems that this assessment was incorrect) 27.Kxb2 Be5+ 28.Kc1 Qa3+ 29.Kd2 Rxd5+ 30.Ke2 Qxa2+ 31.Kf3 Qb3+ 32.Kg2 Rxd1 33.Rxd1 Qxd1 34.Ng4 Qd5+ 35.Kh3 Qh1+ 36.Nh2 Qc1 37.f4 Bd4 38.h6 Qe1 39.h7 Qe8 40.Ng4 Qf8 41.Qh5 f5 42.Ne5 Qh8 43.Qxf7 Bxe5 44.fxe5, 1-0, L.H. Coelho (2392) - J. Molina (2265) Sao Jose de Rio 2003.

23.Nh4 Re4 24.f3 Re5 25.f4 Re4 26.g4 Rxf4 27.Nxf5

27.gxf5? leaves the poor Knight sitting dead on h4 after 27…Re8, when white’s busted.

27…Rxg4 28.Nxh6

A critical alternative is 28.Rhf1, but that doesn’t seem overly dangerous either (if you know what you’re doing, of course!):

* 28…Bf8 29.d6 Rg5 30.Rd5 (30.Qxc5+ Kb8 31.Qc2 Rxh5 32.Ne7 a6 33.Nd5 Qc6 34.Qxc6 bxc6 35.Ne7 Rd7 36.Nxc6+ [36.Rxf7 Bxe7 37.Rxe7 Kc8 38.Re6 Rd5 39.Rxd5 cxd5 40.Rxh6 Kb7 41.Kc2 Kc6 42.Kc3 Rxd6 43.Rxd6+ Kxd6, =] 36...Kc8 37.Rf6 Rc5, =) 30...b6 31.Qe4 Kb8 32.Re5 Rxh5 33.Re1 Rc8 (33…Qb7!?) 34.a4 Rg5 35.Ne7 Rxe5 36.Qxe5 Rd8 37.Rd1 h5 38.Qd5 Qxd6 39.Nc6+ Qxc6 40.Qxd8+ Kb7 41.Rd7+ Ka6 42.Qe8 Qg6+ 43.Ka1 Qc2 44.Rd8 Qc1+ 45.Ka2 Qc4+ 46.Kb1 Bg7 47.Rd7 Qf1+ 48.Kc2 Qc4+ 49.Kb1 Qf1+ 50.Kc2 Bd4 51.Qe7 Qc4+ 52.Kb1 Qf1+ 53.Ka2 Qc4+ 54.Kb1, 1/2-1/2, S. Ganguly (2586) - S.J. Alavi Moghaddam (2493), Cebu City 2007.

* 28...Rg5!? 29.Nxh6 occurred in the game Leonid Kritz - Igor Khenkin, Germany 2005. That game went 29…Rxh5 (29…f6!?) 30.Nxf7 Rg8 31.Qc3 Rh2 32.Nxd6+ Qxd6 33.Rf6 Qe7 34.d6 Qe4+ 35.Kc1 (35.Ka1 Qc2! 36.Qxc2 Rxc2 37.Kb1 Rcg2 38.Rd5 R2g5 39.Rxg5 Rxg5 40.Rf7 Rd5 41.Rc7+ Kb8, =) 35…Qb4? (This gives White some chances. Instead, 35…Rh1! is just equal: 36.Qxc5+ Kb8 37.Rxh1! [37.Qc7+?? loses to 37…Ka8 38.Rxh1 Qxh1+ 39.Kc2 Qe4+ forces mate] 37…Qxh1+ 38.Kd2 Rg2+ 39.Rf2 Rxf2+ 40.Qxf2 Qd5+ 41.Kc3 Qxd6 with a draw.) 36.Qxb4 cxb4 37.Rf4 Rd8 (Giving white more chances. Instead, 37…b3! draws: 38.axb3 Rh6 39.Rc4+ Kb8 40.Rcd4 Rd8 41.Kc2 Kc8 followed by …Kd7, =) 38.Rxb4 Rh6 39.Rc4+ Kb8 40.Rcd4 a5? (40…Kc8 should draw) 41.Kc2 and White was better and, after some more errors by Black, went on to win the game.

28…Rf4 29.Rhf1 Rxf1 30.Rxf1 Rh8 31.Nxf7

31.Rxf7 looks threatening, but it’s actually harmless: 31…Qe8 32.Nf5 Bf8! (32…Qe1+ 33.Qc1 Qxc1+ 34.Kxc1 Bf4+ 35.Kc2 Rxh5 36.Kd3 Rh3+ 37.Ke4 seems unpleasant for Black.)  33.Rxf8!? (33.Qf2 Qe4+ 34.Kc1 Qh1+ 35.Kc2 Qe4+, =; 33.Rf6 Rxh5 34.Qg2 Qd8 35.Qf3 [35.Rf7 Qe8 36.Rf6 Qd8, =] 35...Qxf6 36.Qxh5 Qe5, =) 33…Qxf8 34.h6 (34.d6? Rxh5 35.Qxc5+ Kb8 36.Qc7+ Ka8 doesn’t work for White because his King is too vulnerable: 37.Ne3 Rh1+ 38.Kc2 Rh3 39.d7 (39.Nd1 Rh6 when 40.d7 fails to 40...Rc6+) 39...Qf2+) 34...Rg8 and black’s okay: 35.a3 (35.Qe4 Rg5 36.h7 Rxf5 37.h8=Q?? [37.Qg4, =] 37...Rf1+) 35...Rg5 36.h7 Rh5 37.Ng3 Rxd5.

31...Rxh5 32.Qe4

32.Qg2 Be7 33.Qg8+ Kc7 34.Qg3+ Kc8 35.Ne5 Qh3 (35...Qxd5?? 36.Qg4+) 36.Qxh3+ Rxh3 37.Rf7 Bd8 38.Nc4 Rh6 39.a4 Bc7, =.

32...Bc7 33.d6 Bxd6 34.Nxd6+ Qxd6 35.Qg4+ Kb8 36.Qg8+ Kc7 37.Qg7+ Qd7 38.Qg3+ Qd6 39.Rf7+ Kc6 40.Qxd6+ Kxd6 41.Rxb7 a5 42.Ra7 c4 43.Ra6+ Kc7 44.Kc2 Kb7 45.Rg6 Rh3 46.Rg4 Kb6 47.Rxc4 Kb5 48.Rd4 Rh2+ 49.Kc3 Rh3+ 50.Rd3 Rh5 51.b3 Rc5+ 52.Kb2, 1/2-1/2.

More from IM Silman
Chess Dreams Do Come True

Chess Dreams Do Come True

How To Learn Chess Patterns

How To Learn Chess Patterns