Saturday 1 February 2020

My Best Ever Tournament (part six)

Spanton (2096) - Thorsten Cmiel (2260)
Lausanne Round 6
Sicilian Irregular/English Symmetrical
1.e4 c5 2.c4!?
Sticking with the system I lost with in the previous round, but I felt the loss had little or nothing to do with the opening.
2...g6 3.Ne2 Nc6 4.d4 Bg7!?
4...cxd4 is much more popular.
5.d5 Nd4 6.Nxd4
Vladimir Georgiev (2555) played 6.Nbc3!? in a win over a 2365 in a Danish open eight years ago, but I would be very surprised if the text is not better.
6...cxd4
Vasily Smyslov's choice in a 1962 draw with Ludek Pachman.
7.Bd3 e5?!
Smyslov preferred 5...d6. The problem with the text is it creates a lever for a white kingside attack.
8.0-0 d6 9.f4 exf4?!
From my original notes: "Helping White's development, weakening Black's pawns and making room for White to play e5 cannot be right."
Jan Timman once lost an Olympiad game as White after his Argentine opponent here played 9...f6. The analysis engines Stockfish10 and Komodo10 like 9...Nh6!?, which was played by an unrated in 1986.
10.Bxf4 Nf6 11.e5
Even stronger may be 11.Qa4+, going after the d6 pawn.
11...Nh5 12.exd6?!
Almost certainly better is 12.Qf3, one point being 12...dxe5 can be met by 13.Bxe5. So instead Black should probably play 12...Nxf4 13.Qxf4 0-0, when 14.exd6 transposes to the game, as does 12...0-0 13.exd6 Nxf4 14.Qxf4.
12...0-0?!
Probably better was 12...Nxf4 (however, Komodo10 only comes to prefer this after being shown it on the board) 13.Qe1+ Ne6 14.dxe6 Bxe6 15.Qg3 with a roughly level position.
13.Qf3 Nxf4 14.Qxf4 Qb6!?
The engines slightly prefer the more cautious (read: passive, if Black goes on to lose) 14...Qe8.
How should White proceed?
*****
*****
*****
*****
*****
15.Nd2!
15.b3 Bf5 gives Black a small edge, according to the engines.
15...Qxb2
The engines approve, but now White gets a strong attack.
16.Rab1
16.Rfb1?! hands the initiative to Black after 16...Qc3.
16...Qa3
This is better than 16...Qxa2 17.Ra1 Qb2 18.Ne4, according to the engines. My original notes have Frtiz5's recommendation of 17.d7?! Bxd7 18.Rb7, which Frtiz5 reckoned was slightly better for White, but the modern engines rate it as level. Note that here 16...Qc3 is well-met by 17.Rf3 or 17.Rb3.
17.Rb3 Qa5
If 17...Qxa2, the engines reckon White is much better after 18.c5, eg 18...Qa5 19.Rc1 b6 (best, according to the engines) 20.c6! Qxd5 21.Rb5 Qe6 22.Nf3 (22.Ne4!?) with strong play despite being two pawns down.
18.Rb5?!
My notes show I was hoping for 18...Qd8? 19.c5, which is indeed strong for White. However, the engines come to very much prefer 18.d7!? Bxd7 19.Rxb7 with an initiative.
18...Qa3
18...Qxa2!? 19.Rbb1!? leaves White with more than enough compensation for his pawn deficit, according to Stockfish10, but Komodo10 reckons the position is balanced.
19.Rb3 Qa5
Clearly TC was ready to acquiesce to a draw.
20.Ne4!?
Komodo10's second choice (behind 20.d7!?, as in the note at White's move 18).
20...Qxa2
Correct, according to the modern engines.
Fritz5 reckoned 20...f5 21.Ng5 b6 22.Qh4 h6 23.Ne6 Bxe6 24.dxe6 Qe5 is equal, but Stockfish10 and Komodo10 continue 25.e7 Rfe8 26.c5! bxc5 27.Bc4+ Kh7 28.Qg3! Qxg3 29.Rxg3 Be5 30.Bf7 Bxd6 (or 30...Bxg3 31.Bxe8 Bxd6 32.Bc6) 31.Bxg6+ Kh8 32.Bxe8, when Black has a pawn for the exchange but little hope of saving a half-point. This is a long line, but much of it is forced.
21.Rbb1 b6?!
This stop's White central attack beginning with 22.c5, but now White gets time for a kingside attack. The engines reckon White's advantage is smaller after 21...Bf5!? 22.Rxb7.
22.Nf6+ Bxf6
Not 22...Kh8? 23.d7, eg 23...Bb7 24.Rf3 etc.
23.Qxf6 Qa3?!
This could have worked out well for Black, but objectively better is the engines' 23...Bd7 24.Rbe1 Qd2, although White is well on top after the simple 25.Qxd4.
24.Bxg6?
I originally gave this an exclamation mark, even though my engines at the time thought it loses.
Today's engines reckon the calm 24.Rf3 is much better, eg 24...Qa2 25.h3 (this is more about preventing …Bg4 rather than creating a bolthole for the white king, I believe) Qd2 26.Qe7.
24...hxg6 25.Rf4 Bh3?
Stockfish10 and Komodo10 reckon 25...Bg4! 26.Rxg4 Qe3+ leaves White the equivalent of just half-a-pawn or so better. Probably coming to the same thing is 25...Qe3+ 26.Kh1 Bg4, but then White has the extra option of 27.Rbf1!?
26.Rh4 Qe3+ 27.Kh1 Bxg2+ 28.Kxg2 1-0

No comments:

Post a Comment