I continue to publish material from the pgn file, for people who hate to download pgn files (for whatever reason). The following was my proposed main line, though I admit that 7...d5 and 7...Bxd2+ also have their points. The text is edited for better readability. Games quoted in caps are referring to the pgn file. Analysis by: Stefan Bücker.
1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.c3 Nf6 5.d4 exd4 6.cxd4 Bb4+ 7.Nbd2 Nxe4 For 7...d5 see CHUPROV - GENBA; for 7...Bxd2+ see SAGLIONE - MAGALLANES; for 7...0–0 see STÜMPFIG - PIELECK.
8.d5! (a) 8.a3? forces an exchange...
(b) ...but 8.0–0 does the same more efficiently: 8...Bxd2+ 9.Bxd2 Nxd2 [0–1, 38] in Blaskan,I-Moruzzi,M, Chalkidiki 2000, and [0–1, 64] in Chang,A-Richter,S, Box Hill 2000.
(c) 8.Qe2? d5 9.0–0 (9.Bd3 0–0 10.0–0 Re8–+ [0–1, 22] Delgado,M-Castaneda,K, Bento Goncalves 2000) 9...Bxd2 10.Bxd2 f5? (Instead, 10...Bg4! would have been more difficult to meet, for example: 11.Bb5 0–0 12.Bxc6 Bxf3 13.gxf3 Nxd2 -/+) 11.Bb5 0–0 12.Bxc6 bxc6 13.Ne5 Qd6 14.Bf4 with compensation [1/2–1/2, 29] Ptacek,J-Schovanek,J, Most 1999.
(d) The only real alternative to the text move, though in my opinion less ambitious, is 8.0–0:
(d1) 8...d5 9.Nxe4 (9.Bd3? Bf5 -/+) 9...0–0! (9...dxe4? 10.Ng5 +/- 0–0?? 11.Qh5 +-) 10.Bd3 dxe4 11.Bxe4 +=.
(d2) 8...0–0 9.Nxe4 d5 transposes to variation "d1": 10.Bd3 (10.Bxd5 Qxd5 11.Nc3 Qd8? 12.d5 Klesa,J-Drab,T, Frymburk 2002; 12. Bg5 +=; 11...Qc4! =+) 10...dxe4 11.Bxe4 +=.
(d3) 8...Bxd2!? 9.Nxd2 (9.Bxd2? d5 loses a pawn without compensation, Merino,F-Ramirez Rio,J, Spain 1996) 9...d5 (9...Nxd2? 10.Re1+! Ne7 11.Bxd2 0–0 12.Bg5 Re8 Simlova,K-Rehak,M, Plzen 1996; 13.Qb3 +-) 10.Nxe4 0–0 (10...dxe4 11.d5 Ne5 12.Re1 f5 13.Qb3; 11.Re1!?) 11.Bg5 Qd7 (11...f6? 12.Bxf6 gxf6 13.Bb3) 12.Bd3 dxe4 13.Bxe4 Qxd4 14.Qc2 h6 15.Be3 Qf6 16.Rad1 Re8 17.Rfe1 Nb4 18.Qc4 (18.Qxc7 Rxe4 19.Rd8+ Kh7 20.Rxc8 Re7!) 18...Nc6, about =.
(d4) 8...Nxd2 9.Re1+ (9.Bxd2 Bxd2 10.Qxd2 d5 11.Rfe1+ Ne7 leads to the main line; 11.Bb3 0–0 12.Bc2 Qf6 13.h3? Bxh3 Petimezas,M-Gerrits,B, Hengelo 1994) 9...Ne7 10.Bxd2 Bxd2 11.Qxd2 d5! (11...0–0 12.Qg5! d5 13.Rxe7 dxc4 14.Rae1 with an interesting situation; weaker was 12...Ng6? Sell,G-Rasch,W, Hassloch 1998 13.Bxf7+! Kh8 14.Qxd8 Rxd8 15.Bxg6 hxg6 16.Re7 +-) 12.Bxd5 (12.Bd3?! 0–0 13.Rac1 c6 14.h3 Nf5 15.b4 =+ is a dubious kind of "minority attack", rather optimistic concerning White's material deficit. White's activity here is probably not enough to compensate for the pawn) 12...Qxd5 13.Re5 Qd8 14.Rae1 Be6 15.d5 about =.
(e) For the weak 8.Bd5? see the stem game of the variation, DR. POMTOW - SCHALLOPP.
8...Ne7 8...Na5?! 9.Bd3 Nxd2 10.Bxd2 Qe7+ is original, but could have been punished: 11.Kf1! (11.Be2? Nc4 12.Bc3 Bxc3+ 13.bxc3 0–0 14.0–0 Nb6 and Black was much better, Viennois,R-Cristofari,A, Montpellier 1997) 11...0–0 (11...b6 12.Bxb4 Qxb4 13.Qe2+ Qe7 14.Qc2 Qc5 15.Re1+ Kf8 16.Qe2 Kg8 17.a3 Qf8 18.b4 Nb7 19.Qc2 +/-) 12.Bxh7+ Kxh7 13.Ng5+ Kg6 14.h4 +/-.
9.0–0! 9.a3?! Bxd2+ 10.Bxd2 d6 loses time, but won a game [1–0, 40] in Kartseva,N-Broeker,S, Wolfsberg 2004. - With the text move we are leaving the area for which games have been available.
9...Nxd2 9...Nf6 seems to be weaker: 10.Qb3 Bd6 11.Re1 0–0 12.Ne4 Nxe4 13.Rxe4 Ng6 (13...Rb8 14.Bg5) 14.Bd2 c5 15.Rae1 Rb8 16.Qd3 a6 17.a4 b5 18.axb5 axb5 19.Bxb5 Qb6 20.Bc4 Qxb2 21.Rg4 Ra8 (21...Ne5? 22.Rxg7+ Kxg7 23.Bc3 +-) 22.Bc3 Qa3 23.Nh4 +/-.
The main alternative is 9...Bxd2, e.g.:
(a) 10.Bxd2 0–0 (10...Nxd2 11.Qxd2 d6 12.Rfe1 0–0 13.Re3 h6 14.Rae1 Ng6 15.Nd4 Ne5 16.Rc1 Ng4 17.Rg3 Kh8 18.Rgc3 Nf6 19.Qc2 Re8 20.Bb3 Re4 21.Rxc7 Rxd4 22.Rxc8 Rd1+ 23.Qxd1 Rxc8 24.g3 Kg8 =; 15...Qf6!?; 12...Bg4? 13.Rxe7+) 11.Bb4 d6 12.Re1 Nc5 13.h3 (for 13.Qd2!? see the correspondence game SAGLIONE - MAGALLANES, which White won in 42 moves) 13...b6 14.Bc3 a5 15.Qd2 Bd7 16.b3 Ng6 17.Bb2 Re8 18.Rxe8+ Bxe8 19.Re1 with compensation.
(b) 10.Nxd2!?, for example:
(b1) 10...Nd6?!. To refute this move beyond any doubt may be difficult, but White gets fine chances for an attack: 11.Bd3 0–0 12.Qc2 h6 13.Ne4 Ne8 14.d6 cxd6 15.Bf4 d5 16.Nd6 Qb6 17.Qe2 Ng6 18.Bxg6 Nxd6 (18...fxg6 19.Nxe8 Rxf4 20.Qe5 Qd4 21.Qd6 a5 22.Nc7 Rf6 23.Qg3 Raa6 24.Nxa6 +/-; 20...Rf7 21.Nc7) 19.Bc2 Re8 20.Qd3 Ne4 21.Rae1 Re6 22.Qxd5 Nxf2! 23.Be3 Rxe3 24.Rxe3 Qxe3 25.Rxf2 Qe1+ 26.Rf1 Qe3+ 27.Kh1 Qe6 28.Qd3 (28.Bb3) 28...Qg6 29.Qxg6 fxg6 30.Bb3+ Kh7 31.Rf8 g5 32.Bc4! +/- b5 (32...b6? 33.Bd3+ g6 34.Ba6 Bb7 35.Rf7+ Kg8 36.Bxb7) 33.Bxb5 Kg6 34.Kg1 Bb7 35.Bd3+ Kh5 36.Rf7 +/-.
(b2) 10...Nf6 11.Nf3 0–0 (11...d6? 12.Bg5) 12.Bg5 Ng6 13.d6 cxd6 14.Qxd6 Ne4 15.Bxf7+ Rxf7 16.Qxg6 Qe8 17.Qh5 b6 18.Rfe1 Bb7 =.
10.Nxd2 An attempt to keep White's bishop pair intact. It isn't clear whether this is better or worse than to take with the bishop: 10.Bxd2 Bxd2 11.d6 (11.Qxd2 transposes to the variation 9...Bxd2 10.Bxd2 Nxd2 11.Qxd2 above. In that line 11.d6 obviously would not have been possible because of 11. ..Nxf3+, so the text move is an attempt to profit from the difference) 11...Bb4 (less risky: 11...cxd6 12.Qxd2 0–0 about =) 12.Ng5 (12.dxe7 Bxe7 13.Qb3 0–0 14.Ne5 c6 15.Bxf7+ Kh8 16.Bg8 d5 17.Nf7+ Rxf7 18.Bxf7 Qc7 =) 12...cxd6 (12...Bxd6?? 13.Nxf7 loses the queen) 13.Nxf7 Qc7 14.Qb3 d5 (14...Rf8? 15.Qxb4 Rxf7 16.Bxf7+ Kxf7 17.Qf4+ Kg8 18.Rae1 Qd8 19.Qh4 +-) 15.Bxd5 Rf8 16.Qxb4 Nxd5 17.Nd6+ Kd8 18.Nxb7+ Ke8 19.Nd6+ Kd8 20.Qa3 (20.Qh4+ Rf6 21.Ne4 Qf4 22.Qxh7 Rh6 23.Qg8+ Kc7 24.Rfc1+ Rc6 =) 20...Rf6 21.Ne4 Re6, for example:
(a) 22.Rfe1?! Bb7 23.Rad1 (23.Ng5 Rxe1+ the difference to the main line is that this exchange is accompanied by a check, so White has no time for 24.Qf8+ etc., mate. So: 24.Rxe1 Qf4 -/+) 23...Bc6.
(b) 22.Ng5 Rf6 23.Ne4 = (23.Rad1 Ba6 24.Rfe1 Bb7).
(c) 22.Rae1 Qc6. Of course there are several other continuations; what follows is only a sample line. Black seems to hold. 23.Rc1 Qb6 24.Ng5 Rf6 25.Ne4 Re6 26.Qf3 Bb7 27.Nc5 Rf6 28.Qh3 Bc6 29.Qxh7 Qxb2 30.Qxg7 Rxf2 31.Qg8+ Kc7 32.Qg3+ Nf4 33.Na6+ Kb7 34.Nc5+ Kc7 35.Na6+ Kb7 =.
10...0–0 10...Bxd2 11.Bxd2 (11.d6?!) 11...d6 12.Re1 0–0 13.Rc1 h6 (13...Re8). White has sufficient compensation for the pawn. Of course there is no forcing variation, the following moves only serve as an example: 14.h3 Re8 (14...Bf5) 15.Qh5 Ng6 16.Rxe8+ Qxe8 17.Bd3 Ne5 18.Bb1 c6 19.f4 Ng6 20.Re1 Qf8 21.f5 Ne5 22.f6 g6 23.Qh4 +/-.
11.a3 For analysis on 11.Nb3, see
Luzin's reply #32 in the "Chat" thread.
11...Bc5 (a) 11...Bxd2 is better than the immediate Bxd2 in the last note, since the inclusion of a3 and 0–0 improves Black's chances. I believe that White still gets enough compensation. The idea was advocated by
Luzin in reply #32 in the "Chat" thread, as being clearly favourable for Black. In reply #33 I proposed another treatment for White, which seems to give good compensation.
(b) 11...Ba5 12.b4 Bb6 13.d6 cxd6 14.Ne4 d5 15.Bxd5 Nxd5 16.Qxd5 Qe7 (16...d6) 17.Nd6 Qe6 18.Qd3 Bc7 19.Bf4 Bxd6 20.Bxd6 Re8 21.Qd2 Qf5 22.Rae1 Re6 23.f4 b6 24.Rxe6 fxe6 (24...dxe6 25.Bb8!! Qd5 26.Rd1 +=) 25.Be5 Qe4 26.Re1 Qg6 about =.
12.Ne4 d6 13.Nxc5 dxc5 14.Bf4. With chances for both sides. For the sacrificed pawn, White has active pieces and the bishop pair.