The Spanish Four Knights, Part Two

1.e4 e5 2.Nf3 Nc6 3.Nc3 Nf6 4.Bb5 Bb4 Main Line

by Michael Goeller

The main line of the Spanish Four Knights has developed an undeserved reputation for drawishness. That reputation has arisen because many players adopt the Four Knights as White to draw easily and so play the most quiescent lines. But the Four Knights can also be played as a fighting game where the players embrace the struggle for control of key squares and the initiative. The game Apsenieks-Fine, Stockholm Olympiad 1937, is a case in point. Both players tried their best to win and the contest became a thrilling struggle throughout. The ethic of the Olympiad (at least at that time) tended to discourage drawish play. Both sides desired to win, since not only team prizes but board prizes were at stake. That may be why Fricis Apsenieks (also spelled "Apscheneek") chose the Four Knights and why Reuben Fine chose to play into the main line. This was "manly chess" (as Fred Wilson calls it) of the highest order...

Fricis Apsenieks - Reuben Fine [C49]

Stockholm ol (Men)/Stockholm (14) 1937

1. e4 e5 2. Nf3 Nc6 3. Nc3 Nf6 4. Bb5 Bb4

In my last article, I recommended that White meet the Rubinstein
Variation 4... Nd4 with Sutovsky's favorite
5.

(White can also accept relative equality with either 7. d3!? b5!= Bach-Georgadze, Ampuriabrava 1997 or the immediate 7.Bb3 = Spassky-Korchnoi, St Petersburg 1999 )

7... d6

(a) 7... Bb4?! 8. Nxd4! exd4 9. e5 Ng8 10. Qg4 (10. Bb3!? dxc3 11. dxc3 Be7 12. Qg4 Kf8 13. Qc4 Qd5 14. Qf4 g5 15. Qg3 Qa5 16. Qf3 d5 17. exd6 Bf6 18. Re7 Be6 19. Bxe6 fxe6 20. Rxb7) 10... dxc3 11. Qxg7 cxd2 12. Bxd2 Bxd2 13. Qxh8 (13. Re4!? Emms 13... d5 14. exd6+ Be6 15. Bb3

O-O-O 16. Bxe6+ fxe6 17. Qxh8 Nh6 18. Qf6 Rxd6 19. Rxe6) 13... Bxe1 (13... Qxa4 14. Qxg8+ Ke7 15. Re2) 14. Qxg8+ Ke7 15. Qg5+ Kf8 and White need not take a perpetual check as in Riemersma-Gausel, Gausdal 1993, but can play for a win with 16. Bb3!)

Now an interesting alternative to the main line that yields excellent practical chances for White (in a likely Rook and pawn(s) vs. two minor piece ending) is:

8. Nxd4!?

(Not so promising is the main line 8. h3 b5! (8... Be7 9. d3

O-O 10. Bd2! Qc7 11. Nxd4 exd4 12. Ne2 1-0 Vehi Bach,V-Sanchez Jerez,E/Barbera del Valles 2006 (34)) 9. Bb3 (9. Nxd4? exd4 10. e5 dxe5 11. Qf3 Bd7 (11... Rb8) 12. Rxe5+ Kd8 13. b4 Qc7 14. Rxb5 dxc3 15. dxc3 cxb5 16. Qxa8+ Qc8 17. Qxa7 bxa4 1-0 Tarakanov,O-Sepp,O/Parnu 2002 (36)) 9... Nxb3 10. cxb3 b4! 11. Ne2 c5 1/2-1/2 Shirov,A-Hansen,L/Moscow 1994 (26))

One might ask, "why not the simpler 9... Qxa4?! 10. Nc7+ Kd8 11. Nxa8?! b6 12. b3 Qa6 13. Bb2 Bb7 14. Bxd4 Bxa8 15. Bxf6+?! (better 15.Qf3 Be7 16.Qg3) 15... gxf6 16. Qh5," but White has many ways to improve on this line, including 11.b3! Qa5 12.Nxa8 b6 13.Bb2 Bb7 14.Nxb6 (better than Pinski's suggestion 14.Bxd4 Bxa8 15.e5?! dxe5?! 16.Rxe5 which, he writes, "would give White a very strong attack for the piece" [sic]) 14...Qxb6 15.c3 which should lead to open lines in the center, for example.

Pinski says this leads to "a mess" and suggests instead 10...Qd8! 11.Nf4 Ne5 "with equality," which is probably right -- though the position resembles lines of Bird's Variation of the Spanish / Ruy Lopez that are hardly bad for White. Probably 12.d3 first followed eventually by the c3 break and d4 push, though White does have to watch out for tricky tactical ideas on the kingside.

11. Nc7+ Kd8 12. Nxa8 b6 13. Rb1!?

The most forcing of White's alternatives.

(14... Bxa8? 15. Ra3 Qb5 16. Rxa7)

15. Ra3 Qc6 16. Rxa7 Bxa8 17. b5 Qc8 18. c3 Qb8

A difficult position for both sides that therefore should be full of practical chances for the well-prepared....

19. Ra4

(19. Qa4!? Bb7 20. cxd4 cxd4 21. Bb2 Nc5 22. Qa3 Ne6 23. Ra4 d5 24. Qf3 Qf4 25. Bxd4)

19... dxc3 20. dxc3 Be7 21. Bf4 f6 22. Ra3?!

The Rook is better placed at a4 for now. White must have better.

(Perhaps 22. Qb3!? g5 23. Bg3!? h5 (23... Ne5!?) 24. Rd1! Re8 (24... h4? 25. Bxd6 Bxd6 26. Qe6) 25. Qe6 Bf8 26. Bxd6! Rxe6 (26... Bxd6 27. Qxd6 Qxd6 28. Rxa8+ Kc7 29. Ra7+ Kb8 30. Rxd6 Kxa7 31. Rxd7+) 27. Bxb8 Bxe4 28. Ra7 Bd3! 29. Rxd7+ Kxd7 30. Rxd3+ Kc8 31. h4! with a fascinating endgame that should definitely yield winning chances to the well-prepared.)

22... Re8 (22... g5!?) 23. Qh5 h6 24. Qg6 Bf8 25. f3 Bb7 26. Rd1 Qc7 27. c4 Ne5! 28. Bxe5 Rxe5 29. Rad3

White has achieved his ideal position, envisioned on move 22, but Black now has the better dynamic chances and soon gains the initiative.

29... Rg5 30. Qh7?! Qf7 31. a4 Kc7 32. Qh8? Re5 33. a5 bxa5 34. Ra3 Kb6 35. Rda1 Re8 36. Rxa5 g6 37. Ra7 Bg7! 38. R1a6+ Bxa6 39. Rxf7 Bxh8 40. bxa6 Kxa6 41. Rh7 h5 42. Rh6 Ka5 43. Rxg6 Kb4 44. Rh6 f5 45. exf5 Bd4+ 46. Kf1 Kxc4 47. g4 Kd3 48. Re6 Rg8 0-1 Smagin-Malaniuk, Tilburg 1993 (48)

In his notes to this game in *The Lost Olympiad*, W. H. Cozens writes that this line is not as drawish as its symmetry might suggest, noting: "The masters of the early 20th Century played the Four Knights with aggressive intent and extracted some fine wins with it." As my notes suggest, a lot of that "aggressive intent" is masked by modern, well-studied defensive technique on Black's part, which side-steps the most dangerous lines.

Making room for the well-known "Metger unpin" Knight maneuver Nc6-d8-e6.

The more usual order of moves is 6... d6 7. Bg5 Bxc3 8. bxc3 Qe7 9. d4 when Black should not play 9... h6 (better 9... Nd8 transposing to the game continuation) 10. Bh4 g5?! because 11. Nxg5! hxg5 12. Bxg5 yields a dangerous attack, e.g.: 12... exd4 (12... Bg4?? 13. Bxf6! Bxd1 14. Bxe7 Nxe7 15. Rfxd1) (12... Nd8!? 13. Qf3 (13. f4!? Ne6 14. Bxf6 Qxf6 15. fxe5 Qg5 16. Rf3 Ng7 17. Rg3) 13... Kg7 14. Qg3 Qe6 15. Bxf6+ Kxf6 16. f4) 13. cxd4 Kg7 14. Re1 Qe6! 15. f4 (15. d5 Qe5 16. Qd2! Rh8! 17. Bf4 Qh5 18. Be2 Nxe4) 15... Nxe4! 16. d5 (16. Bd3 f5) 16... Qf5 17. Bd3 Qxd5! 18. Bxe4 Qd4+ 19. Kh1 (19. Qxd4+!? Nxd4 20. Bh4) 19... Qxd1 20. Raxd1 Bg4 21. Rb1 Rae8 22. f5 Re5 23. Bd2 d5 24. Bd3 Rfe8 25. Rxe5 Rxe5 26. Rxb7 1/2-1/2 Imanaliev-Yuneev, Frunze 1989 (26)

7. Bg5

An alternate method (especially with ...Bg4 not possible) is 7. Ne2 d6 (7... d5!? 8. exd5 Nxd5 9. Bxc6 bxc6 10. Ng3) 8. c3 Bc5 9. d4 Bb6 10. d5

7... Bxc3

Forced, to prevent Nd5.

The most aggressive line for White.

a) White typically first defends the e-pawn by 9. Re1 when might follow 9... d6 10. d4 Ne6 (10... Bg4=) 11. Bc1 c6 12. Bf1! Rd8 13. g3 which we may examine in Part Three of this series.

b) One point of Fine's nuanced opening play may be that 9. Nd2 looking to play f4 can be answered powerfully by 9... Ne6 10. Bh4 c6 11. Bc4 Nf4! and Black will get in ...d5 in one move, with advantage.

9... d6

Black must defend the e-pawn, since surrendering the center is no option: 9... exd4? 10. e5 Qc5 11. Bd3! Nd5 12. cxd4 gives White a powerful position.

10. d5!

"Spiking the Metger plan," notes Cozens. But Black can now open the c-file for play against White's doubled pawns.

a) Still too dangerous is 11... g5?! 12. Nxg5 hxg5

( Perhaps better 12... Nxe4!? 13. Nf3 f6 (13... Ng5!?) 14. Qe1 Nc5 15. Nd2 Bf5 16. Qe3 Qg7 17. f4 e4 18. Rae1 Kh7 19. Qd4 a6 20. Be2 b5 21. Bh5 Ndb7 22. Re3 Rad8 23. Rfe1 Rd7 24. Nf1 Re7 25. Ng3 Bg6 26. f5 Be8 27. Bxe8 Qg4 28. Bg6+ 1-0 Gronroos,M-Alkkiomaki,J/Helsinki 2003 (28))

13. Bxg5 c6 14. Qf3! Kg7 15. Bd3 Rh8 16. h3?! (16. Qg3!) 16... Rg8?! 17. Qg3! Kf8 18. Qh4 (18. f4!?) 18... Rxg5 19. Qxg5 cxd5 20. f4?! (20. Rae1) 20... Nxe4! 21. Qh6+ 1-0, Leite-Coimbra, POR-ch 2000 (46)

b) 11... Bg4 12. h3 Bxf3!? ( better 12... Bh5) 13. Qxf3 g5 14. Bg3

12. Ba4!?

The chief advantage of this move is that it preserves a good square for the Queen at d3.

a) The more typical retreat is 12. Bd3 cxd5

(12... Re8 13. c4 Bd7 14. h3 a6 15. a4 b6 16. g4 Nb7 17. g5 hxg5 18. Nxg5 Nh7 19. Qh5 Nxg5 20. Bxg5 f6 21. Be3 Nc5 22. Kh2 Qf7 23. Qh4 Qg6?! 24. Rg1 Qh7 25. Bh6! Re7 (25... g5 26. Rxg5+!! fxg5 27. Qxg5+ Kf7 28. Be2) 26. Qxf6 Rf7 27. Qxd6 Rxf2+ 28. Kh1 Rf7 29. Raf1 Rxf1 30. Bxf1 Nxe4 31. Qxd7 Qxh6 32. dxc6 Nf2+ 33. Kh2 Rf8 34. c7 Ne4 35. Qd5+ 1-0 Curdo,J-Conner,M/Queen City 1998 (35))

14. Qd2 b6 15. Qe3 Bh5 16. Nd4 Bg6 17. Rae1 Qb7 18. Bxf6 exd4 19. Bxd4 Qxd5 20. c4 Qh5 21. f4 Bf5 22. Be2 Qh4 23. Bf3 Rc8 24. Bd5 Ne6 25. g3 Qh5 26. Bb2 Qg6 27. Qf3 Rc7 28. Kg2 Rcc8 29. h3 h5 30. Kh2 Kh7 31. Re2 Rc7 32. Rg2 Nc5 33. g4 Be4 34. Bxe4 Qxe4 35. Qg3 Qg6 36. f5 Qg5 37. Qxd6 Rd7 38. Qxf8 h4 39. Bc1 Qf6 40. Qa8 Qe5+ 41. Bf4 1-0, Carvalho-Leite, POR-ch 2000 (41)

b) less good is 12. Bc4?! cxd5 13. Bxd5 Bg4 14. h3 Bxf3 15. Qxf3 g5 16. Bg3 Nxd5 17. exd5 f5 0-1 Karlsson,B-Vnukov,D/IECG email 1999 (43)

12... Bg4 13. Qd3 cxd5 14. exd5 Rc8

The natural square for the Rook, facing White's doubled pawns, but W. H. Cozens roundly criticizes this move since the Rooks position is quite unfortunate in the main combination that arises later.

a) 14... e4 Fritz suggests 15. Qd4! exf3 (15... g5 16. Nxg5 Qe5) 16. Qxg4

b) 14... Bxf3 15. Qxf3 g5 16. Bg3 Nh7 17. Qf5! a6 18. c4 b6 19. Rab1 Qc7 20. c3 Kg7 21. Bc2 Rh8 22. Rfe1 Qxc4 23. Rxe5!! Qc7 24. Rbe1 Qxc3 25. Re7 1-0 Petro,J-Chladek,V/ICCF corr 1994 (25)

This Knight might prove annoying.... If 16... b6 17. Nd4!? (17. Rae1 Nb7 18. Nd4!)

Cozens writes: "Not only wild on principle but tactically quite unsound; in short, a clanger. Fine has forgotten the undefended rook." However, the further course of the game suggests that Fine played this move as an intentional sacrifice.... There is not much appeal for Black in 17... b6 18. Qe3 Re8 19. f4

18. Bxg5 hxg5 19. Qh3+ Kg7 20. Qxc8

Cozens writes that Apsenieks "has won the exchange and a pawn. He has hooked his fish and just needs to reel it in. What he has at the end of the line, however, is a man-eating shark."

20... Ne6!

Simultaneously springing the Knight into attacking position and cutting off the Queen's most natural rereat to f5, where it might be useful for defense.

White must switch immediately to defense with 22. Rfe1! Qd7 23. Qf1 Rh8 24. f3

22... Qd7 23. Rfe1 Qg4 24. Qf1 Rh8

25. Rd3

Giving back the exchange, if Fine will take it!

No better is 25. Re3 Qh5 ( Cozens mistakenly suggests 25... Nh3+? overlooking 26. Rxh3! Rxh3 27. f3) 26. h3 g4 27. Rg3 Kf8

25... Qh4?!

Mistakenly testing White before recouping the Exchange.

Best was the immediate 25... Nxd3! since White must then recapture with the Queen 26. Qxd3 (because here 26. cxd3? allows 26... Rxh2!! 27. Re3 (27. Kxh2? Qh4+ 28. Kg1 Ng4 29. Qe2 Qh2+ 30. Kf1 Qh1#) 27... Qh4 28. Rh3 Rxh3 29. gxh3 Nh5) 26... Qh5 27. h3 g4 28. Re3 Rh6 and Black still has some chances despite his pawn deficit.

The improvement of White's pawns over the line above is sufficient to give him the edge. Now the White Bishop can re-enter the game via c2 or d1.

28... Rh6 29. Qe2 (29. f4?! gxh3 30. Rxh3 Qxf4=) 29... Rg6 30. hxg4 Rxg4 would certainly still be quite difficult for White to win due to Black's kingside attacking chances.

29. Rg3+!!

A deadly interpolation! Now Black must surrender material, since, if the King retreats to f8, the Rook will be undefended after 30.Rxh3. And, of course, the Queen is lost if the King goes to the h-file. Fine must have expected to enter a difficult Queen ending after 29. Rxh3 Qf4 30. Rxh8 Kxh8 which would be hard to win.

29... Ng4 30. gxh3 f5 31. Qg2!

31... Qh6 32. Qf3 Kf8 33. Rxg4

"A thrilling game" writes Cozens. I like how well it illustrates the type of kingside attacks that can develop very quickly out of the Four Knights for both sides.

1-0