A Rare Practical Example of
8...d5!? 9.OOO Be6 10.Nxd5!
by Michael Goeller
It is rare, except in correspondence play, to encounter the traditional main line of the Urusov Gambit which follows 1.e4 e5 2.Bc4 Nf6 3.d4 exd4 4.Nf3 Nxe4 5.Qxd4 Nf6 6.Nc3. In overtheboard encounters, 4...Nc6 (transposing to the familiar Two Knights Defense) is by far the most common choice, with 4...d5 a distant second. When Black does play into the traditional main line, he usually goes for building a solid center with 6...c6 and 7...d5 or trying to catch up in development by 6...Nc6 7.Qh4 Be7 8.Bg5 d6. Only very rarely does he risk 6...Nc6 7.Qh4 Be7 8.Bg5 d5!? which has long been considered suspect, though probably not for the right reasons.
The old book refutation suggests that White gets a strong attack that at least forces a draw after 9.OOO Be6 10.Rhe1, but close analysis suggests that Black may be more resilient in this line than previously thought. More promising, therefore, is NM Victor Baja's old recommendation of 10.Nxd5! which analysis suggests gains White a clear edge. The only problem is that 10.Nxd5 has never been seen in practice. So it is a rare treat to see a game with what should be the book refutation of 8...d5!?  especially when Black introduces an interesting twist of his own...
Remi Legendre  NN [C24]
Correspondence/France 2008
1. e4 e5 2. Bc4 Nf6 3. d4 exd4 4. Nf3 Nxe4 5. Qxd4 Nf6 6. Nc3 Nc6 7. Qh4 Be7
7... Bb4!? 8.
10. Nxd5!
This was California NM Victor Baja's suggestion and it seems to lead to at least a slight edge for White and would therefore be my choice. I am still undecided about the soundness of the "book" line beginning 10.Rhe1, when White may be able to maintain his attack with Kb1 and g4. However, the two "forced draw" lines given by older books actually favor Black, and I have my doubts about how well the 10.Rhe1 line will hold up to close scrutiny:
(a) 10... Qd6? 11. Bxd5 Qc5 12. Bxe6 fxe6 13. Bxf6 Bxf6 14. Rxe6+ Ne7 (14... Kf7 15. Rxf6+!) 15. Qa4+ Kf8 16. Ne4 10 Schuermans,RDe Jonghe,B/Belgium BEL 2006 (27).
(b) 10... h6?! 11. Bxd5! (11. Bxf6!? Bxf6 12. Qh5 10 TereschenkoRotlevi/St Petersburg 1909 (16) 12... Kf8!?) 11... Nxd5 12. Rxd5! Bxd5? (12... Bxg5+ 13. Nxg5 Qf6 14. Qh3
OO 15. Nxe6 fxe6 16. Ne4 Qf7 17. Rd3) 13. Bxe7 Nxe7 14. Nxd5OO 15. Nxe7+ SpagnuoloSmith, Email Tournament 2000.
11. Kb1!?
This refined move may transpose to the book line, which goes:
(a) 12. Bxh6?! Ne4! 13. Qf4 (13. Bg5 Bxg5+ 14. Nxg5 Qxg5+ 01 Fleischer,EKotzem,E/Bad Bevensen 1999 (40)) 13... Bd6! 14. Qe3 f5!? (14... Bc5 15. Qf4 Nxc3 16. Bxg7) 15. Bg5 Qd7 16. Bb5 Qf7 17. Bxc6 bxc6 18. Nxe4 fxe4 19. Nd4 Bd7 01 DegliErediEg, Correspondence 1998 (23).
(b) 12. Rxe6?! Keres thought this was a forced draw 12... fxe6 13. Bxh6 gxh6 (13... Nb4!?) 14. Qg3+ Kh8 15. Qg6 Rf7! (15... Bb4 16. Qxh6+ Kg8 17. Qg6+= 1/21/2 Leisebein,PMarkus,J/DDR 1989) 16. Qxf7 Qg8 with equality according to Forintos and Haag, but Black should actually have a slight edge due to his control of the center.
(c) 12. g4 Nxg4 13. Bxe7 Qxe7 14. Qg3 Qd6 Forintos and Haag.
12... Ne8 Euwe (12... Nd7!? 13. Bxe7 Qxe7 14. Qg3!?) 13. Bxe7 Qxe7 14. Qxe7 Nxe7 15. Nd4 Nc6 16. Nxe6 fxe6 17. Rxe6 Rxf2 18. Nxd5 Rxg2 19. Bc4 Kh8 20. Rde1 Nd6 21. Nxc7 Rd8? (21... Nxc4! 22. Nxa8 Nd4!?) 22. Bd3! Kg8 23. Bg6 10 LaesZittorio/corr 1972.
11... Re8 12. Bd3 h6 13. g4 Bf8?! (13... hxg5! 14. Nxg5 Bb4 Van der Tak) 14. Nb5 hxg5 15. Nxg5 g6 16. Bxg6! fxg6? (16... Bg7) 17. Nxe6 Rxe6 18. Rxe6 Bg7 19. g5 Nh7 20. Rxg6 Nf8 21. Rxg7+ Kxg7 22. f4 Qd7 23. Nc3 Ne7 24. Qh6+ Kg8 25. Ne4 Qf5 26. Ng3 Qe6 27. Qh5 Nf5 28. Qf3 Nxg3 29. hxg3 Re8 30. a3 c6 31. Rh1 Qe2 32. Qc3 Qe4 33. Rh8+ Kf7 34. Qf6# 10 Strijbos,MZagema,W/Breda 1997.
10... Nxd5
10... Bxd5 11. Bxd5 Nxd5 12. c4 Bxg5+ 13. Nxg5 Qf6 14. Rxd5 h6 15. Re1+ Kf8 16. Re3! Kg8 17. Nf3 Qxh4 18. Nxh4 transposes to one of the lines below, but leaves both players fewer options.
12. Rhe1!? is probably not a bad alternative.
12... Bxf3!?
I gave this move as doubtful in my original analysis, but Black has a hidden resource I had not seen.
a) 12... Nb4 13. Rhe1! (13. cxd5) 13... f6 14. Bxf6 gxf6 15. a3
b) 12... f6 13. Rxd5 Qc8 14. Bd2
OO 15. Rh5! h6 16. Bxh6! gxh6 17. Rxh6 Qf5 18. Re1
c) 12... Bxg5+ 13. Nxg5 h6 14. Rhe1+ Kf8 15. Rxd5 Qf6 16. Re3! Kg8 17. Nf3 Qxh4 18. Nxh4

14. Bxe7
14. gxf3!? is the only legitimate alternative, but it gets difficult after 14... Rd4 15. Qg3 Rxc4+ 16. Kb1 Bxg5 17. Qxg5
14... Rd4!
14... Nxe7? 15. gxf3! (even better than 15. Re1)
15. Qg3
15. Qg5!? Rxc4+ 16. Kd2 (16. Kb1 Be4+ 17. Ka1 Nxe7 18. Re1
From this point on it appears that Black can force a draw. The last chance for an improvement is:
16. Kb1! Be4+ 17. Ka1 Kxe7!? (17... Nxe7 18. Re1
Another attractive drawing line for Black goes 17. Ke3 Re4+! 18. Kxf3 f5!! 19. Qxg7 Nd4+ 20. Qxd4 (20. Kg3 Ne2+ 21. Kf3 (21. Kh3?? Nf4+ 22. Kg3 Nh5+)
21... Nd4+=)
20... Rxd4 21. Bf6 Rd3+ 22. Ke2 Rf8 (22...
17... Be4 18. Qxg7 Rd3+ 19. Kc4

19... a6!!
Basically, White has no defense against perpetual check by Na5+ and Nc6+ since the Rook at d3 and the pawn at a6 take away all of the other squares. Of course, 19... Kxe7 20. Qxh8 a6!! transposes to the game.
21. b4?? b5+ 22. Kc5 Kd7! and White gets mated.
21... Na5+ 22. Kb4 Nc6+ 23. Kc4
23... Na5+
1/21/2
A fascinating game! And I want to thank Remi Legendre for sharing it with me. But my impression is that White has more opportunities for improvement than does Black, and therefore the whole line beginning with 8...d5 remains simply too dangerous to repeat.