Hastings Continued

Faced a Slovak in round four of Hastings Masters this afternoon.
Spanton (1951/168) - Svetlana Sucikova (2160/196)
Caro-Kann Classical
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3
There are more than 19,000 games with this position in ChessBase's 2020 Mega database.
10...e6 11.Bd2
Joe Gallagher in Starting Out: The Caro Kann argues that 11.Bf4 is more precise as it is more active if Black castles short, and he says some players of the black pieces will only castle short if White has played Bd2.
11...Ngf6 12.0-0-0 Be7 13.Ne4
More popular is 13.Kb1, but both moves have been played by lots of strong players.
13...Nxe4 14.Qxe4 Nf6 15.Qe2 Qd5 16.c4 Qe4 17.Qxe4 Nxe4 18.Be3 0-0 19.Ne5 Rfd8 20.g4 Rac8 21.Kb1?!
Decentralising. The main move - we are still in theory - is 21.f3, with the black knight usually retreating to d6. Komodo10 reckons the position is equal, but Stockfish10 gives White a slight edge.
21...Bd6
At last a new move, or at least one not in Mega20. It is Komodo10's choice, but Stockfish10 prefers 21...Bf6.
22.f3 Ng5 23.Bxg5!?
This leaves Black with bishop versus knight with rooks on the board, which usually favours the side with the bishop. However, here the knight is well-placed on the fifth rank, and can only be kicked by weakening the e pawn.
23...hxg5 24.Rhe1 Rc7 25.Kc2 f6 26.Nd3 Kf7 27.c5?!
The engines prefer a move I seriously considered, namely 27.Nc5!? I rejected it because 27...Bxc5 28.dxc5 Rcd7 29.Rxd7+ Rxd7 gives Black the more promising pawn-majority for creating a passed pawn. Komodo10 comes to give Black a slight edge, but Stockfish10 reckons 30.Kc3 equalises.
27...Bg3 28.Re4
Almost certainly better than 28.Rg1?!, which makes it impossible, as far as I can tell, for White to defend the weak d pawn.
28...Rcd7 29.Kc3 Rd5 30.Rd2
Black is clearly better; the only question is whether it is enough for a win.
30...Bc7 31.b4

Can Black get an advantage here?

*****
*****
*****
*****
*****
31...f5?!
The engines give 31...a5 32.a3 b5, when Stockfish10 continues 33.Ne1 e5 34.Rd1 f5 35.gxf5 exd4+ 36.Kc2 Be5 37.Nd3 Bf6 38.bxa5 Ra8 39.Re6 Ra6 40.Rd6 Rxf5 41.Kb3 Rxa5 42.Rxc6 Rxf3 43.Rb6 with advantage to Black. Komodo10 gives 33.f4 f5 34.Ne5+ Bxe5 35.Rxe5 gxf4 36.Rxd5 axb4+ 37.axb4 Rxd5 38.g5 e5 39.dxe5 Rxe5 40.Rd7+ with counterplay for White, but much prefers Black.
32.Ne5+
Now White seems fine.
32...Kf6 33.Ree2 Bb8 34.Kc4 Bc7 35.Kc3 Bb8 36.Kc4 a6 37.Kc3
The engines reckon White is better after 37.Ng6, but I fear I was concentrating so much on avoiding defeat that I rarely looked for ways to get an advantage.
37...Bc7 38.Kc4 Ra8 39.a4 Rad8
39...Bxe5 40.dxe5+ Rxe5 41.Rxe5 Kxe5 temporarily wins a pawn, but White gets strong counterplay with 42.Rd7.
40.a5 Bb8 41.Kc3 Bc7 42.Kc4 Bxe5 43.Rxe5 Kf7 44.Kc3 Ke7!? 45.Rde2 Rxe5 46.Rxe5 Rf8 47.Kd3 fxg4 48.fxg4 Rf3+ 49.Re3 Rf1
49...Rf4 50.Re4 Rxe4 51.Kxe4 Kf6 52.Ke3 e5 53.Ke4 is a draw, but not 53.dxe5+?? Kxe5 etc.
50.Rh3
Active defence.
50...Rb1 51.Kc4 Rc1+ 52.Kd3 Ra1 53.Kc4 Rg1 54.h6 gxh6 55.Rxh6 Kd7
55...Rxg4 56.Rh7+ Kf6 57.Rxb7 is also a draw, according to the engines.
56.Rh7+ Kc8 57.Rh8+ Kc7 58.Rh7+ Kb8 59.Rh8+ Ka7 60.Re8 Rxg4 61.Rxe6 Rg1 62.Re8 g4 63.Rg8 g3 64.Kb3
Not 64.Kd3? g2 65.Ke2 Rb1 66.Rxg2 Rxb4, when only Black can win.
The game finished:
64...Rb1+ 65.Kc3 Rc1+ 66.Kb3 Rg1 67.Kc3 Rc1+ 68.Kb3 Rg1 69.Kc3 g2 70.Kb2 Rd1 ½–½

A Caro-Kann Tabiya

PLAYED last night in the fourth round of the Pimlico Summer open section, which is being held over five consecutive Thursday evenings.
Spanton (171) - Anders Lundback (162)
Caro-Kann Classical
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3 e6
This is overwhelmingly the most-popular move in ChessBase's 2019 Mega database.
Grandmaster Joe Gallagher in Starting Out: The Caro-Kann (Everyman Chess, 2002) says 10...e6, 10...Ngf6 and 10...Qc7 are "interchangeable if Black is going to castle queenside, but if he wants to remain flexible on the future home of his king, then 10...e6 and 10...Ngf6 are the better options."
Put like this, it is hard to understand why any strong player would choose 10...Qc7, but it has been a frequent choice of Epishin and has also been played by Korchnoi and Khalifman.
11.Bd2
The most-popular move, but inaccurate, according to Gallagher. He explains: "This move is considered slightly inferior to Bf4 because Black has an easier time in the lines where he castles kingside. The point is that with the bishop on d2, White will have to waste time with Qe2 if he wants to play Ne5, whereas with the bishop on f4 it is possible to play this at once."
What Gallagher means, I believe, is that with the bishop on d2, White would not want to play Ne5 because, after ...Nxe5, dxe5 Qxd3, the queens come off, White gets a backward d pawn and Black gets the d5 square as an outpost for his king's knight. But with the bishop on f4, …Nxe5 can be met by Bxe5.
11...Ngf6 12.0-0-0 Be7

A popular tabiya in the Caro-Kann Classical

There are more than 3,500 games with this position in Mega19. Most popular is 13.Kb1, although Gallagher suggests it is "too slow to cause Black difficulties." In Starting Out he uses a game that saw 13.Qe2 - "White prepares to play Ne5" - the third-commonest move. Other choices by strong players include 13.c4, 13.Rhe1, 13.Rde1 and 13.Nf1!? My choice is second-most popular with almost 1,300 games.
13.Ne4 Nxe4 14.Qxe4 Nf6 15.Qe2 Qd5
Gallagher gives this an exclamation mark. Black's queen attacks a2 and adds to the pressure on h5.
16.c4 Qe4 17.Rhe1!?
Gallagher says "17.Be3 b5! is very dangerous for White," while 17.Qxe4 Nxe4 18.Be3 "is no longer considered a realistic winning attempt."
17...Qxe2 18.Rxe2 0-0-0
The critical and seemingly good 18...Nxh5! was tried in Tim Dean (-) - Zhichao Li (2028), Edmonton Closed 2005. There followed 19.d5 cxd5 20.cxd5 0-0 21.dxe6 fxe6 22.Rxe6 Rac8+ 23.Kb1 Bc5, with an equal position (but 0-1, 49 moves).
19.Ba5 b6 20.Bc3 Bd6
20...Nxh5?! 21.Ne5.
21.Ne5 Bxe5 22.Rxe5 Ng4 23.Re2 Nf6
A tacit offer of a draw.
24.g4 Rd7
If 24...Nxg4, then 25.Rg1 Nf6 26.Rxg7 Nxh5 27.Rxf7 Rd7 28.Rxd7 Kxd7 29.d5 with advantage. This line is pretty forcing, but Black could try 26...Rh7 27.Rxh7 Nxh7, although White's resulting rook-and-bishop combo has good chances of outplaying Black's rook-and-knight.
25.f3 Rhd8 26.Rde1

How would you assess this position, and how should Black proceed?

*****
*****
*****
*****
*****
White's space advantage and superior minor piece (there is a lot of pawns on the board, but the knight is unlikely to find a good outpost) give an advantage - worth from 0.4 pawns (Komodo9) to about 0.75 pawns (Stockfish10). But much manoeuvring is in prospect and there is no need for Black to panic.
26...c5?
A mistaken bid for freedom.
27.dxc5 bxc5 28.Re5 Rd3
The engines suggest Black's best may be to sac the exchange by 28...Rc7!? 29.Ba5 Rd4 30.Bxc7 Rxc4+ 31.Kb1 Kxc7, but then 32.Rc1 forces an exchange of rooks, leaving Black with a pretty hopeless knight-versus-rook ending.
29.Rxc5+ Kb7 30.Ba5 R8d7 31.Rb5+
The engines want White to play the passive 31.Rf1!? I rejected it because of 31...Re3, missing that Black is hardly in a position to double rooks on White's third rank as White replies Rc7+.
31...Ka6 32.Bc3 Rxf3 33.Ree5 Rf1+ 34.Be1??
Self-pinning is very rarely a good idea, and here it is catastrophic. I rejected the winning 34.Kc2 because, after 34...Rf2+ 35.Kb3 Rb7, White's attack has come to an end. This is true - up to a point. White's queenside attack is over, but White can switch to the kingside with 36.g5, eg 36...hxg5 37.Rxg5 Rxb5 38.cxb5+ Kb7, when the engines continue 39.Bxf6 gxf6 40.Rg7 with a large advantage.
34...Nxg4 35.Re2
A humiliating retreat, but the only way to stay in the game as 35.Re4??, for example, hoping for 35...Nf6?? 37.Re3, fails to 35...Nf2.
35...Rc7?!
The engines prefer an immediate 35...Rb7, with a small edge to Black after 36.Ra5+ Kb6. The drawback of the text will be explained later.
36.b3 Rb7
There is a better chance for a tiny edge, according to the engines, with 36...Nf6 or the pin-breaking 36...Rf5!?
37.Ra5+ Kb6 38.Kc2?!
More accurate is 38.Kb2. The point is that if Black proceeds with 38...Rf3, which the engines recommend after the text, White has 39.Rg2.
38...a6
The engines prefer 38...Rf3, when 39.Rg2?? now loses to 39...Ne3+. This seems to be why the engines did not like Black's 35th - it gave White a tempo to free the b2 square for his king!
39.c5+ Ka7 40.Bc3 e5?!
40...Rb5 41.Ra4 Nf6 42.Rc4 is dead equal, according to the engines, despite Black's pawn-plus.
41.b4?!
The simple 41.Bxe5 Nxe5 (not 41...Re7?? 42.Bb8+) 42.Rxe5 is slightly better for White.
41...Rf5?
The rook-and-pawn ending after 41...Rb5 42.Rxb5 axb5 43.Bxe5 Nxe5 44.Rxe5 is even, according to the engines, but looks easier for White to play.
Also reasonable was 41...f6, although White remains more active thanks to his queenside play.
42.c6?
Better was 42.Rd2, threatening Rd6, eg 42...Rf6 43.a4 (threatening 44.b5) Rc7 44.Kb3, with strong pressure.
42...Rc7
Black may also hold with 42...Rb5!?, the engines giving 43.c7 Kb7 44.Rxa6 Kxc7 45.a4 Rd5 46.Ra7+ Kb8! (better than advancing the king, apparently) 47.Re7, with a position hard to assess (the engines vary from preferring White to reckoning Black has equality).
43.b5 Kb6 44.bxa6 Ka7?
Black is in trouble after this. The engines give best play as 44...Rf6 45.Ra4 Rfxc6 46.a7! Rxa7 (46...Rxc3?! costs Black the exchange, although he does get two pawns for it) 47.Rb4+ Kc7 48.Rxg4 Rxa2+ 49.Kd1 Rxe2 50.Ba5+ Kd7 51.Kxe2, when White has rook, bishop and pawn versus rook and four pawns - equal, according to the engines.
45.Re4 Nf6 46.Rc4 Rxh5
White is also winning after 46...Rf4 47.Rxf4 exf4 48.Bd4+ Ka8 49.Bxf6 gxf6 50.Rc5.
47.Bxe5 Rxe5
Not much better is 47...Rc8, eg 48.Kb3 Rh3+ 49.Ka4 Rd3 50.Rb5 h5 (Black has no useful moves) 51.Bb8+!
48.Rxe5 Kxa6 49.Re3 Kb6 50.Rb3+ Ka5 51.Rb7 Ne8 52.Kb3 1-0