How I beat Tal: the Caro can

The Caro-Kann defence suffers from a somewhat pacifist reputation when compared with the famously combative Sicilian, but this reputation is unfair. Several world champions have played the Caro-Kann to win as Black, notably Capablanca, Botvinnik, Smyslov, Petrosian, Karpov and even Kasparov. It should also be observed for Sicilian devotees that after 1. e4 c5, the continuation 2. c3 makes it virtually impossible for Black to prevent White stonewalling a draw, whatever contortions they may attempt to impose.

This week, I present four of my own wins on the Black side of the Caro-Kann, with comments augmented by the latest analysis engine technology.

Mikhail Tal vs. Raymond Keene

Simul, 20b (exhibition), London, 1964

1. e4

I sit down to face Tal. General expectation: a gory loss embellished with Tal-like sacrifices.

1… c6

He plays 1. e4, I played the Caro-Kann. It was my main defence at the time. I had just won the London under-18 championship and I had played it several times during that event — notably against Kenny Harman. I also played Taimanov and Lowenthal Sicilians at the time. I used the Lowenthal against Kotov and drew, but thought this might be too provocative against Tal.

2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. N1e2 e6 7. h4 h6 8. Nf4 Bh7 9. Bc4 Nf6 10. O-O

10… Nd5

Instead, 10… Bd6 allows the famous Nxe6 sacrifice from Tal’s first match against Botvinnik which Tal was still prepared to repeat in games as late as 1978. I wanted to avoid Tal’s vastly superior knowledge of this line so chose something to try and dampen down White’s attacking prospects.

Interestingly, our pet engine ‘Pancho’, debunks both the threat and the execution of this possible sacrifice. In the first instance, after 10… Bd6 it suggests as best, 11. Ngh5 (11. c3 is also good) 11… O-O 12. g3 Bf5 13. Bd3 Nbd7 14. Bxf5 exf5 15. Qd3 Ne4 16. Kg2 Ndf6 17. Be3 with equality. Pancho further suggests that 10… Nd5? is simply an error. Also, if the threat after 10… Bd6 is activated with 11. Nxe6, it again deviates from Botvinnik’s line quite quickly: 11… fxe6 12. Bxe6, when instead of 12… Qc7 13. Re1 Nbd7 14. Bg8+ Kf8 15. Bxh7 Rxh7, played in our stem game, it prefers, 12… Nbd7 13. Re1 Qc7 14. Bh3+ Kd8 15. Nf5 Bxf5 16. Bxf5 Re8 17. Rxe8+ Kxe8, when Black has a pleasant edge, but no more than that.

11. Qg4

This is the error, although it looks absolutely plausible. It was later discovered that 11. Bxd5 cxd5 and now either Qg4 or Qh5 gives White a huge attack.

But Pancho again pushes the boundaries in positing that much the best continuation here is 11. Re1! when, 11… Qxh4 12. Qf3 Bb4 13. c3 Nxf4 14. Bxf4 Be7 15. Nh5 O-O 16. Be5 Bf5 17. Nxg7 Bg4 18. Qe4, delivers a highly promising position close to a winning breakthrough.

11… Nxf4 12. Bxf4 Nd7 13. Be5

An inferior move. White can maintain momentum with 13. Nh5, but after the text, the position is level.

13… Nxe5 14. dxe5 Qc7

Black has better available in, 14… Bxc2 15. Rac1 h5 16. Qf4 Bg6 17. Rfd1 Qc7, with parity.

15. Rad1

All this looks like vintage Tal, to keep Black’s king stuck in the centre, and my next move looks like I am cooperating in my own demise by grabbing a hot pawn against Tal, of all people!

15… Bxc2

This almost seems like desperation but in fact I was no longer very worried. White’s c-pawn was a very useful unit and its loss now causes White problems. My main worry was that Tal would still find some way to sac on e6 and explode my position before I could get castled.

16. Rd2 h5

A decoy to gain time after 17. Nxh5 Bg6 18. Nf4 Bf5. Tal hardly even thought about taking the pawn and slowing down–he kept his queen fixed on e6 where he clearly still wanted to sacrifice.

17. Qh3

Pancho thinks this a serious mistake by White. Fixated by the threats to the e6-square, he overlooks the much stronger, 17. Qg5 Bg6 18. Bxe6 Be7 19. Rd7 Bxg5 20. Rxc7 fxe6 21. hxg5 O-O 22. Rd1 Rad8 23. Rd6 Rxd6 24. exd6, with a significant advantage.

17… Bh7 18. Rfd1 Bb4 19. Rd7

The only way forwards now, but Black’s position is secure.

19… Qxd7

Pancho calls this out as a mistake. Black’s position was secure after, 19… Qxe5 20. Rxb7 Bc5 21. Ne2 O-O, but after the move played, Black hands the initiative back to his opponent.

20. Rxd7 Kxd7

So it has come to pass, and Black never got castled. White now also starts to rip up the Black kingside, but I was happy with two rooks for a queen and the bishop pair, and soon some open lines too.

21. Nxh5 Kc7 22. Nxg7 Rag8 23. Nh5 Bf5

21. Qf3

An awful mistake from the master, according to Pancho. Necessary is 24. Qb3 when the game continues with chances for both sides. After the text, White is facing serious threats.

24… Rg4 25. Nf4 Bd2

Now I knew I had him on the run. Tal thought for a second and gave up his knight.

26. Qa3 Bxf4 27. Qd6+ Kc8 28. Ba6 Be4

Defending and attacking. White’s threats are now illusory.

29. Qc5 Bxg2 30. Qxa7 Bh2+

A sac finishes White off. I was very pleased, of course, to beat Tal in a war of movement (not just simul-induced blunders) and I felt the choice of opening had been psychologically a good one. I was fortunate though, that Tal did not discover the probable refutation of Black’s innovation with 11. Bxd5.

Pancho is in complete accord with Black’s powerful focussing of forces after White’s disastrous twenty-fourth move, culminating in this very final and forced mate in two, 31. Kxh2 Rhxh4+ 32. Kg1 Rh1#.

White resigns 0-1

Korom vs. Raymond Keene

International junior team tourney, The Hague, 1965

1. e4 c6 2. d4 d5 3. exd5 cxd5 4. c4 Nf6 5. Nc3 g6 6. Qb3 Bg7 7. cxd5 O-O 8. Nf3 Nbd7 9. Bg5 Nb6 10. Bxf6 exf6 11. Be2 a5 12. a4 Qd6 13. O-O Rd8 14. Nd2 Qb4 15. Rfd1 f5 16. Nb5 Nxd5 17. Bc4 Be6 18. Qxb4 axb4 19. Nf3?

Having negotiated the opening without either error or advantage, White’s move commits the knight before knowing its correct position. 19. Bb3 or g3 are better waiting moves.

19… Rac8?

Black fails to capitalise on White’s last inaccuracy. A better try is, 19… Nc3 20. bxc3 Bxc4 21. cxb4 Bb3 22. Rdb1 Rxa4, which consolidates an advantage for Black.

20. Bb3 Bh6 21. Ne5 Nf4 22. Bxe6 fxe6 23. g3 Nd5 24. Nd3 Rc2 25. Re1 Nc7 26. Nxb4 Rc4 27. Nxc7 Rxc7 28. Rxe6 Rxd4 29. Rb6??

Despite a few missteps around time control at move 40, it is this move that proves White’s downfall, in a game that should have ended about twenty moves earlier than it does. The move played is predicated on the untenable defence of his queenside pawns. Instead, White should find a method of redeploying his knight. For example, after 29. Na2 Rxa4 30. Re2 Ra5 31. b3 f4 32. Kg2 Rf5 33. f3 Rb5 34. Rb1 Kf7 35. g4, the position is completely equal.

29… Bf8 30. Na2 Rxa4 31. Kg2 Rc6 32. Rxb7 Rca6 33. Rd1 Rxa2 34. Rdd7 h6 35. h3 R2a4 36. h4 Rb4 37. h5 Rxb7 38. Rxb7 g5 39. Kf3 Ra4 40. Rb5 Rb4?

A weak move. Better is 40… f4 41. Rb6 Rd4 42. g4 Rb4, when Black retains a significant lead.

41. Rxf5 Rxb2 42. Kg2 Rc2 43. Rf6 Kg7 44. Rf5 Rc5 45. Rf3 Be7 46. Re3 Kf6 47. Kh3 Rf5 48. Kg2 Bc5 49. Re2 g4 50. Rc2 Kg5 51. Re2 Kxh5 52. Kf1 Kg5 53. Kg2 Kf6 54. Kf1 Re5 55. Ra2 Rf5 56. Re2 h5 57. Kg2 Kg5 58. Rd2 h4 59. gxh4+ Kxh4 60. Rc2 Rxf2+ 61. Rxf2 Bxf2 62. Kxf2 Kh3 63. Kg1 Kg3 White resigns 0-1

 

John Joseph Carleton vs. Raymond Keene

British Team Championship, London, 1978

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nf6 5. Nxf6+ exf6

I now regard 6. c3! as best.

6. Bc4 Qe7+

6. Bc4 was the only move given in the Encyclopaedia of Chess Openings as leading to White’s advantage against 5… exf6, but 6… Qe7+ is an adequate counter and far superior to 6… Bd6 or 6… Be7. 6… Qe7+ was first conceived by Bronstein and was later propagated by Bob Wade who used it to draw with Matanovic in 1972. After 6… Qe7+ Black’s threat of …Qb4+ more or less forces White to play 7. Qe2 (7. Be2 Bg4 is easy for Black, while 7. Kf1?! looks artificial) when the most reliable procedure is 7… Be6!, as played in this game.
7. Qe2 Be6! 8. Bb3 Nd7

8… Bxb3 9. axb3 Qxe2+ 10. Nxe2 (Kuijpers – Lechtynsky, Nice Olympiad 1974) gives White a nagging edge in the endgame. If light-squared bishops are to be traded, Black would prefer this to take place on e6; e.g. 8. Bxe6 (in place of 8. Bb3) 8… Qxe6 9. Bf4 Na6 10. c3 O-O-O (Gaprindashvili – Andersson, Dortmund 1978) and now if White swaps queens Black can straighten out his pawn structure, while if White avoids the exchange Black’s queen is excellently posted on e6.

9. Bf4 Nb6 10. O-O-O Nd5

Here White has two ways of moving his queen’s bishop – either to g3, as in the game, or retreating it to d2. Thus, 11. Bd2 b5 12. Nf3 Qd7 13. Ne1 Be7 (Tal proposed 13… Bd6) 14. Nd3, and here I suggest the aggressive 14… a5! In the game Tal – Bronstein, USSR 1974, Black proceeded carelessly with 14… O-O? when the blockading coup 15. Ba5! utterly stymied Bronstein’s intended queenside pawn offensive.

11. Bg3 g6

Worth consideration is 11… b5 as an immediate remedy against White’s playing c2-c4.

Pancho gives, 11… a5 12. c4 a4 13. Bc2 Nb4 14. a3 Nxc2 15. Qxc2 Qd7, as the best line. 

12. Nf3?!

If 12. c4!? Bh6+ 13. Kb1 Nf4!? (13… Nb4 14. d5 Bf5+ is unclear) 14. Bxf4 Bxf4 15. d5 Bf5+ 16. Bc2 Bxc2+ 17. Qxc2 O-O 18. Ne2 Be5 with an obscure position.

12… Bh6+ 13. Kb1 O-O 14. Rhe1

If 14. Bxd5 cxd5 15. h4 Rac8! 16. h5 Rxc2!! 17. Qxc2 (17. Kxc2 Bf5+ is winning for Black) 17… Bf5 18. Rd3 Rc8 19. Qb3 (or 19. Qd1 Rc1+) 19… Qe2 etc.

14… Rfe8 15. Qe4 Qd7 16. Qh4 Bg7

Intending …g6-g5.

17. h3!

The only move.

17… a5 18. a3 b5 19. Nd2 a4 20. Ba2 b4 21. axb4 a3 22. Bb3

If 22 Bxd5 cxd5! planning …Qa4, which is better for Black.

22… g5! 23. Qh5 axb2 24. Kxb2 Nxb4

Black evidently enjoys a devastating attack against the exposed white king, but White’s next move makes things even worse.

25. Kc3?

If 25. Nc4 Bxc4 26. Rxe8+ Rxe8 27. Bxc4 Qb7 28. Bb3 c5! (better for Black) 29. Bd6 c4! 30. Bxc4 Nd3+ 31. Kc3 Nxf2 threatening …Nxd1+ and …Ne4+.

25… Na2+ 26. Bxa2 Rxa2 27. Ra1 Rxa1 28. Rxa1 c5 29. Nb3 Bxb3 White resigns 0-1

If 30 cxb3 Qxd4+, or 30 Kxb3 Qb5+ 31 Kc3 Qb4+.

 

Lawrence Day vs. Raymond Keene

Windy City International, Chicago, 1985

1. e4 c6 2. Nc3

I had expected 2. d4 d5 3. e5!? Bf5 4. Nc3, e.g. 4… e6 5. g4 Bg6 6. Nge2 Bb4 7. h4 Be4 8. Rh3 h5 9. Ng3 c5 10. Bg5 Qb6 as in Day – Vranesic, Canadian Championship 1981; here Day himself in BCO gives 11. a3! ’better for White’. Therefore I wanted to try 4… Qb6!? 5. g4 Bd7 6. Na4 Qc7 7. Nc5 e6 8. Nxd7. Velimirovic – Kasparov, Moscow Interzonal 1982, proceeded with 8… Nxd7. My new idea was 8… Qxd7 intending a quick …c6-c5 in conjunction with …Nc6. The black Ng8 can go to e7 to hold up White’s thematic plan of f2-f4-f5 which caused Kasparov trouble in his game.

2… d5 3. Nf3 Bg4 4. h3 Bxf3 5. Qxf3 Nf6 6. b3!?

This must be a novelty, though it’s quite in Day’s inventive style. [I have recently found a record of this having been played in Zapata-Burger, New York, 1983]. I anticipated White to follow up with O-O-O and then g4, h4, g5, h5, trying to mince my kingside in cooperation with the Bb2.

6… e6 7. Bb2 Nbd7 8. exd5?

From now on White starts to play feebly. He should maintain the tension with 8. O-O-O, although 8… Bb4 is satisfactory for Black.

8… cxd5 9. Qg3

With threats of Nb5, also trying to impede the development of Black’s dark-squared bishop. A broadly similar concept was seen in Smyslov – Botvinnik (first match game, World Championship 1958), which duplicated this game for the first five moves and then varied with 6. d3 e6 7. Be2 Nbd7 8. Qg3, but Botvinnik proved that 8… g6! is possible, even with …e7-e6 already played. Black developed the Bf8 on g7 and went on to win. So in this game I always had …g7-g6 in mind as a reserve development.

White does better to continue with either 9. g4 or O-O-O, according to Pancho.

9… a6 10. O-O-O Rc8 11. d4?

A losing positional blunder which weakens c3 and blocks his dark-squared bishop. Perhaps White was bothered by distant threats of …d5-d4 from Black. Anyway, 11. Re1 is superior.

Pancho suggests that 11. Kb1 or Bd3 is even better than 11. Re1.

11… Qa5

Black swiftly works up an annihilating initiative.

12. Kb1 Ba3 13. Ba1

Perhaps 13. Qe3 and Bxa3 are better tries.

13… Bb4 14. Rd3

Of course 14. Qxg7 loses a piece to 14… Rg8.

14… O-O 15. Qe3

The threat was 15… Bxc3 and 16… Ne4.

15… Rc6 16. Nd1 Ne4

Also tempting but complicated is 16… Rxc2 17. Kxc2 Qxa2+ 18. Bb2 Rc8+ 19. Rc3! Bxc3 20. Nxc3 Ne4 21. Bc4! Nxc3 22. Ra1 Nd1!

Better than 18… Rc8+ in the above analysis, is the continuation suggested by Pancho, 18… Ne4 19. Rc3 Nxc3 20. Nxc3 Rc8.

17. c3 Be7 18. h4 b5

The rest is automatic. White’s queenside pawns are structurally dead, and Black enjoys the luxury of having the white king as an additional target.

19. b4

An awful move, opines Pancho. White should prefer, 19. g4 b4 20. c4 dxc4 21. Qxe4 Rfc8 22. bxc4 Rxc4, when White remains worse, but there remains play in the position.

19… Qc7 20. f4

I didn’t understand this, but it is too late to make any difference.

20… Rc8 21. Bb2 Nb6 22. Rh3 Nc4 23. Qe1

A series of brutal sacrifices now forces mate.

23… Bxb4 24. cxb4 Na3+ 25. Bxa3 Rc1+ White resigns 0-1

Ray’s 206th book, “  Chess in the Year of the King  ”, written in collaboration with Adam Black, and his 207th, “  Napoleon and Goethe: The Touchstone of Genius  ” (which discusses their relationship with chess) can be ordered from both Amazon and Blackwells. His 208th, the world record for chess books, written jointly with chess playing artist Barry Martin,  Chess through the Looking Glass ,  is now also available from Amazon. 

A Message from TheArticle

We are the only publication that’s committed to covering every angle. We have an important contribution to make, one that’s needed now more than ever, and we need your help to continue publishing throughout these hard economic times. So please, make a donation.