Wednesday, March 20, 2013

The Complete Caveman Caro-Kann

I have posted The Complete Caveman Caro-Kann (download PGN) analyzing 1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5, about which I have written several articles over the years.  This latest piece offers a complete overview of the system after 5.Bg5 and features annotations by Angus French of his great win with the line at the London Chess Classic Open (which was published in IM Malcolm Pein's column in The Telegraph).  If time allows, I will eventually offer up a complete repertoire for White, including interesting approaches against the topical 3...c5 and alternatives at move four.

I have carried the analysis as deeply as possible, but there are still some positions that are quite unclear, where "White has compensation" is all that can reasonably be claimed.  The complexity of the situation will definitely favor the well-prepared player and computers will offer no easy answers to either side, even in correspondence.  In this way the opening resembles the complex lines of the Botvinnik Semi-Slav and the King's Indian Defense and represents a very interesting challenge for analysts and players alike.  

One thing is certain to me now: White is definitely better if Black goes after the Rook by 5...Qb6 6.Bd3 Bxd3 7.Qxd3 Qxb2? 8.e6!  This is a common misstep among players who have not seen this line before (and who therefore assume White has simply lost his mind).

Position after 7...Qxb2? 8.e6!
Of course, the lines are quite complicated even after the Queen falls into its "mammoth trap," but it is eventually possible to discover White's winning ideas, which begin with 8...Qxa1 9.Qb3! -- threatening Qxb2 winning material, or Nf3, O-O, and Nbd2 trapping the Queen.

The most complex play results not from grabbing the Rook in the corner but instead from swiping the d-pawn right in the center of the board with 6.Bd3 Qxd4!? 7.Nf3 (Black can meet 7.Bxf5? with 7...Qxe5+ winning back the Bishop) and now 7...Qg4! -- a variation that I did not even consider in my initial analysis.

Position after 6...Qxd4 7.Nf3 Qg4!
White has lost a couple of recent games with this line (included in my notes), though they featured weak play.  However, there is no question that White is hard-pressed to prove more than "compensation" here, and so it is definitely one avenue for Black players going for the win and willing to take some tactical risks.  White's best in the diagram above may be to exchange Bishops immediately with 8.Bxf5! Qxf5 (the square f5 is not as good as g4 for the Queen) 9.Qe2 followed by a quick c4, though there are no games to illustrate.

A more strategic choice is to decline the gambit with 5...Qb6 6.Bd3 Bxd3 7.Qxd3 e6, after which Black should prefer a quick c5! over Anand's idea of Qa6 offering the trade of queens, which White can exploit or even accept with some advantage.  Note, for example, that White can meet Anand's 7...Qa6 (played before White is ready to block the diagonal with c4) with 8.Qh3! e6 9.Nc3 transposing to our main game in the 8.Nc3 which looked very good for White.  And 8.Qxa6!? Nxa6 9.e6!? should also favor White.

After 5...Qb6 6.Bd3 Bxd3 7.Qxd3 e6, the chief question is whether to play 8.Nd2 (which leaves open the option of a quick c4 advance) or 8.Nc3 (where White focuses on speedy development and a potential piece breakthrough).  In practice, 8.Nd2 has proven more popular, but White needs to think about how to meet the immediate 8...c5! -- when best must be to fight fire with fire by playing 9.c4!  The resulting play will emphasize White's lead in development.  

Position after 8.Nd2 c5! 9.c4!?
In the process of putting together this blog post, I stumbled upon a recent game that demonstrates some of the dangers Black faces in this line:
Hart, David (2228) - Turner, Jeff (2046)
San Diego CC Championship (1), February 20, 2013    
1.e4 c6 2.d4 d5 3.e5 Bf5 4.h4 h5 5.Bg5 Qb6 6.Bd3 Bxd3 7.Qxd3 e6 8.Nd2 c5 9.c4 Ne7 10.Ne2 Nd7 11.0–0 cxd4 12.cxd5 Nxd5 13.Nc4 Qc7 14.Rac1 Nxe5 15.Nxe5 1–0
While 8.Nd2 is attractive, my personal preference is for the less common 8.Nc3 (not considered by Lukacs and Hazai in their ground-breaking article in NIC Yearbook #100), which can be quite dangerous for Black unless he immediately challenges the center with 8...c5! 

Position after 8.Nc3 c5!
White should probably choose to castle kingside here after 9.Nf3 followed by dxc5 and O-O, as I discuss in my notes, leading to an interesting position where White still has good chances of fighting for the advantage.  This line awaits trials.  And it may well be that the right answer for White is to meet an early Qa6 with the Nc3 line and to choose the Nd2 line followed by c4 as the best remedy for Black's early c5 plan.  

In posting my analysis online for free, I hope to encourage more games and analysis that should eventually yield a more complete understanding of this fascinating line.  I welcome games and new ideas from my readers.



Anonymous said...

Hoho, nice stuff ! May be I'll switch from my Panov attack for this one. I hope to study this line seriously and teel you what I think.

Keep the good job going on !


Michel Barbaut

Hadron of said...

I took up the Advanced Variation of the French because it seemed more in line with what I play as I get older so it was a natural progression to look at the same for the Caro. I really like being able to add sharpness. However as I have continued to look at the "Caveman" I had to ask myself wouldn't be natural for some players to whip out 5..f6 (as ugly as it looks)? A number of the engines suggest that 6.Bishop retreats is ever so slightly in Blacks favour. However can I suggest 6.Bd3 !? because after 6…Bd3 7.Qxd3 fxg5 8.Qg6+ Kd7 9.e6+ Kc7 10.hxg5 Black is utterly clamped down and will have further trouble with Ng1-f3-Ne5

Michael Goeller said...

I think you are right -- White should be able to sac here. I have played some blitz games with this line and will try to dig them up. The possibility of ...f6 comes up in other lines, and I think White has to be prepared to just abandon the Bishop there too. In his book Attacking the Caro-Kann, Dreev analyzes the line 5.Bg5 Qb6 6.Bd3 Bxd3 7.Qxd3 Qxb3 8.e6 f6 9.Nf3! (his mark) 9...fxg5 (9...Qxa1 10.Qb3) 10.O-O g4 11.Nbd2 gxf3 12.Rfb1 Qxa1 13.Rxa1 Na6 14.Rb1 O-O-O 15.Nb3 Nc7 16.Nc5 b5 17.Qa3 Kb8 18.Qa5 Nf6 19.Rb3 +- (Dreev, p. 48).

Michael Goeller said...

By the way, there is promising game in the 6...Qxd4 line:

[Event "80A Championship Brasileiro Absoluto - Fase Final"]
[Site "?"]
[Date "2013.12.19"]
[Round "?"]
[White "Barbosa, Evandro Amorim"]
[Black "Molina, Roberto Junio Brito"]
[ECO "B12"]
[WhiteElo "2461"]
[BlackElo "2424"]
[Result "1-0"]

1. e4 c6 2. d4 d5 3. e5 Bf5 4. h4 h5 5. Bg5 Qb6 6. Bd3 Qxd4 7. Nf3 Qg4 8. Bxf5 Qxf5 9. O-O Qd7 10. c4 e6 11. Nc3 Ne7 12. Qe2 Nf5 13. Rfd1 Bb4 14. Rac1 Na6 15. Ne4 Be7 16. cxd5 exd5 17. Nf6+ gxf6 18. exf6 Qe6 19. Qf1 Nc5 20. fxe7 Ne4 21. Nd4 Nxd4 22. Rxd4 Qe5 23. Rb4 f6 24. Rxb7 Nd6 25. Re1 Qf5 26. Rc7 fxg5 27. Rxc6 Nc4 28. b3 Nb6 29. hxg5 Qxg5 30. Rc7 d4 31. f4 Qd5 32. Re5 Qd6 33. Qb5+ Nd7 34. Qb7 Nb6 35. Qxa8+ Nxa8 36. Rc8+ Kf7 37. Rxh8 Nc7 38. Rf8+ Kg7 39. Rc8 1-0

Michael Goeller said...

The latest statement on this line comes from Daniel Lowinger's "Opening Originals: Strong Sidelines for Club Cats" ( with its chapter on 1. e4 c6 2. d4 d5 3. e5 Bf5 4. h4 h5 5. Bg5 Qb6 6. Bd3 Qxd4 7. Nf3 Qg4, where Lowinger analyzes 8. Bxf5 Qxf5 9. O-O f6!? as the most likely solution for Black.

Michael Goeller said...

John Shaw's book "Playing 1.e4: Caro-Kann, 1...e5 & Minor Lines" (Quality 2016) is a good book to have for anyone serious about the Caveman and related lines. He covers Black alternatives very well, and he offers a different line involving Bg5 that makes a good alternative line against 4.h4 h5.