Refuting 5...Nxe4 in the Scotch Four Knights
by Michael Goeller
What opening should you choose when playing someone rated over 400 points higher than yourself? I have heard GMs (notably Yermolinsky) say that you should always play something thoroughly sound and well-tested, even against higher rated players, since that will do the most to "make them work." But it is hard to resist testing your opponent with a trappy line -- especially if you are a noted openings Expert like Pete Tamburro who probably has lots of those tricks under his sleave. The following game, however, demonstrates that the GMs are probably right. It also provides a very nice proof that a tricky line in the Scotch Four Knights is thoroughly unsound.
IM Enrico Sevillano (2500) - Pete Tamburro (2090) [C47]
US Open Championship/Cherry Hill, NJ USA 2007
This move has been known to theory for some time. It can be tricky, and though theory shows that White gets a clear edge with best play, the move did not seem really refuted... until now!
6. Nxc6?! Nxc3 7. Nxd8 Nxd1 8. Nxf7 Kxf7! (8... Nxf2? 9. Nxh8 Nxh1 is "surely one of the weirdest openings in the history of chess" according to Fred Reinfeld (Chess Life, July 1950), but likely losing for Black after 10. Be3! Bd6 (10... g5!? 11. Be2!)
A move with great merit, as the further course of the game shows. Interestingly, several older books (including Gary Lane's Winning with the Scotch from 1993 ) award this a question mark!
a) The "book refutation" of this line goes 7. f3 d5 8. Bb5! (8. Nb5? is called "interesting but unclear" by Reinfeld, but it seems losing after 8... dxe4 9. Bf4 exf3+ 10. Kf2 Qh4+ 11. g3 Bc5+ 12. Ke1 Qe7+ 13. Kd2
O-O) 8... Bd7 9. O-Odxe4 10. Bxc6 bxc6! (10... Bxc6 11. Nxc6 bxc6 12. Re1 "This variation, by Scheltinga, gives White a considerably superior game" is Reinfeld's conclusion.) 11. Re1 (11. fxe4!? also looks good enough for an edge) 11... O-O-O12. Rxe4 Qf6 13. Qe2 Bc5 (13... c5 14. Nb3 Bc6 15. Qa6+ Bb7 16. Qxf6 gxf6 yields an interesting ending, where Black's activity and two Bishops likely compensate for his awful pawn structure.) 14. Be3 Rhe8 when Burgess says that this "looks OK for Black," though White is clearly for choice after 15. Nb3 (15. b4!? Bb6 (15... Rxe4 16. bxc5!!) 16. Rd1 Rxe4 17. fxe4 Kb7! threatening ...Bg4 18. c3 Qe5 is survivable) 15... Bxe3+ 16. Qxe3 Rxe4 17. fxe4! Kb8 18. c3
Reinfeld simply mentions this line as "very unclear" and leading to "an uncomfortable game for Black, though he has the solace of a Pawn to the good."
Here, Graham Burgess's "101 Chess Opening Surprises" simply notes that Black "holds everything together," but he gives a longer line under 7.Nb5 Qxe4+ 8.Be2.
11... Qe5 has also been tried here, though it does not look like any improvement: 12. b4?! (12. Re1! a6 13. Nd4) 12... a6 13. bxa5 axb5 14. Qd2 Re8? (14... Nxa5) 15. Rfe1 d6 16. Rab1 Rxa5 17. Bb6! cxb6 18. Bf1 Be6 19. Rxe5 dxe5 20. Bxb5 Rd8 21. Qe1 Bxa2 22. Rb2 Bd5 23. h3 f6 24. Qe3 Ra1+ 25. Kh2 Rd7 26. Rd2 Be6 27. Rxd7 Bxd7 28. Qxb6 1-0 Belkhodja,S-Stromer,A/France 1991.
This appears to be a significant improvement on existing theory, though not exactly a novelty in the position.
12. Nd4 is mentioned by Burgess.
It is hard to recommend a satisfactory alternative. A previous game with the same line beautifully demonstrates what Black is up against: 12... Ne5!? 13. b4 Bb6 14. Bxb6 cxb6 15. f4 Qe3+ (15... Ng6 16. Qd2 also looks better for White.) 16. Kh1 Qxc3 17. fxe5 Qxa3 18. Qd6 a5 19. b5! Qxd6 20. exd6 (Black's pieces are entombed) 20... a4 21. a3 Re8 22. Bc4 Re6 23. Rad1! 1-0 Burmeister,R-Giesen,D/Germany 2003.
Black's Queen has no good square.
Desperation, but it is all over.
White wins. I think 5...Nxe4? in the Scotch Four Knights is now completely refuted.1-0
Game in PGNCopyright Michael Goeller © 2007