After 1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3 Bc5 my opponent played 5.c3.
This is White's commonest continuation in ChessBase's 2022 Mega database, although in my praxis I have found 5.0-0 to be almost 50% more popular.
After 5.c3 the main line is 5...0-0, but I played 5...d5!?, a relatively new idea that has been tried by Kramnik, Aronian, Nakamura, Grischuk and other strong players.
 |
| Position after 5...d5!? |
*****
*****
*****
*****
The main line in Mega22, albeit in a variation that is fast evolving, runs 6.exd5 (first appearance in 2015) Qxd5 7.Bc4 Qd6, and either 8.b4 or 8.Nbd2. Stockfish14.1 reckons both moves give White a slight edge, while Komodo12.1.1 reckons Black has equalised.
In Alicante my opponent - the full game can be seen at https://beauchess.blogspot.com/2021/12/alicante-round-one.html - played 6.Nxe5, which tended to be White's response in early days of the variation in the 2010s.
It is easy to see why this looks very strong, especially as 6....Qd6? 7.d4 Bb6 8.Nxc6 bxc6 9.e5 Qe6 10.Bd3 leaves White a sound pawn up. Black could try 7...Nxe4 but 8.Bxc6+ bxc6 9.f3 Nf6 10.Qe2 wins a piece.
The correct reply to 6.Nxe5 is 6...0-0, when Black is a pawn down but leads in development at a point where White has not castled.
 |
| Position after Black has castled |
*****
*****
*****
*****
Three moves are known to have been tried.
The worst of them, 7.d4?, played by a 2070, runs into 7...Nxe5 8.dxc5 Qe7 (Black also seems to be better after 8...Nxe4 and 8...dxe4) 9.0-0 dxe4, when Black is a pawn up, has the better pawn-structure and is more developed.
The engines reckon best is 7.Bxc6 bxc6 8.d4 Bb6 with an equal position. Note that 8.Nxc6?! Qe8 9.Nd4 dxe4 is very promising for Black, eg 10.0-0 Ba6 or 10.Be3 Ba6. A 2302 tried 9.d4?! only to be hit with 9...Qe4+ 10.Kf1 Ba6+ 11.Kg1 Rfe8, when White is a pawn up but lost, according to the engines.
That leaves 7.Nxc6 bxc6 8.Bxc6, when early games, albeit featuring blacks rated under 1500, continued with 8...Rb8.
Then came the discovery of 8...Bxf2+!, which first appears in Mega22 in a game from 2017.
White has to capture as 9.Kf1? dxe4! 10.Bxa8 leaves Black with a pleasant choice between two bishop moves, 10...Bg4 and 10...Bb6, eg 10...Bg4 11.Qc2 Bh4 12.Bxe4 Nxe4 13.dxe4 Qf6+ 14.Kg1 Qb6+ 15.Kf1 Rd8 16.g3 Rd1+ 17.Qxd1 (this is best, or rather least-worst) Bxd1, when White is arguably ahead on material - two rooks, knight and pawn for queen and bishop - but Black seems to be winning, one line running 18.gxh4 Qf6+ 19.Kg1 Bf3 20.Nd2 Qg6+ 21.Kf2 Bxh1 22.b3 Qg2+ 23.Ke3 Qh3+ 24.Kd4 Qxh4 with an overwhelming position for Black. This is an extremely long line given by the engines, and there are many alternatives along the way, but Black seems to be winning in all of them.
After White captures with 9.Kxf2, Black plays 9...Ng4+, and now White's best continuation is not clear.
A) 10.Kf1?! looks strange, but two of the four games to reach the position after 9...Ng4+ in Mega22 saw this move. Both games continued 10...Qf6+ 11.Qf3 Qxc6 12.exd5 Qg6, when there is a split. Dadi Ómarsson (2266) - Einar Hjalti Jensson (2323), Icelandic Team Championship 2018, went 13.h3 Ne5 14.Qe4 Nxd3 15.Nd2 Qg3 16.Qe2 Bg4!? 0-1. Thijs de Bock (1959) - Lucas van Foreest (2488), Brugge 2018, saw 13.Na3, when 13...Bf5 would have given Black a winning advantage, according to the engines. The game continuation of 13...Rb8?! 14.d4 Re8 15.h3 Nf6 was not so clear, but Black won anyway by delivering checkmate on move 26.
B) 10.Kg3 was tried in Gulrukhbegim Tokhirjonova (2362) - Richard Yi (2103), Pan American College Team Rapid Championship. The game continued 10...Qd6+?! 11.Bf4 Qxc6 12.h3, when Black may not have enough compensation for being a pawn down. The engines suggest 10...Qf6!? 11.Rf1 Qg6 12.Qe2! (only move, according to the engines) Ne5+ 13.Kf2 Qxc6, claiming Black has an edge despite being a pawn down.
C) 10.Ke2!? does not appear in Mega22, but is the choice of Komodo12.1.1, at least for a time. The engine continues 10...Ne5 11.Nd2 Nxc6, giving Black a slight edge. Stockfish14.1 prefers 10...Qf6!? 11.Qa4 Qf2+ 12.Kd1 Qxg2 13.Re1 Ne5 14.Nd2 Nxd3 15.Rf1 Bg4+ 16.Kc2 Be2, claiming Black is winning at the end of this largely forced line, eg 17.Bxa8 Rxa8 18.b3 Bxf1, although Komodo12.1.1 is less sure of Black's advantage.
D) My game saw 10.Kg1, which comes to be both engines' top choice. After 10...Qf6 my opponent played 11.Qf3?!, which also featured in a Latvian Championship game (see https://beauchess.blogspot.com/2021/12/alicante-round-one.html for details). The engines agree 11.Qf3?! is not good, but have trouble deciding what White should play.
D1) Stockfish14.1 settles on 11.Qe1, giving best-play as 11.Qxc6 12.h3 Nf6 - it also likes 12...Ne5 - 13.e5 Nd7 14.Qg3 f6 with an unclear position but one which both engines reckon slightly favours Black.
D2) Komodo12.1.1 likes 11.Bf4!?, continuing 11.Qxc6 12.h3 Qb6+ 13.d4 Nf6 14.Nd2 dxe4 15.Nc4 Qb7, when the engines agree Black is better.
CONCLUSION: There is a lot of analysis here but it seems 5...d5!? stands up to the attempted refutation by 6.Nxe5 - Black is better in most lines and winning in many. However the variation is very sharp. It is easy for either player to go wrong, but easier, I would say, for White, especially if caught by surprise.
No comments:
Post a Comment