Notes in italics are by grandmaster John Fedorowicz.
Elina Groberman (2054) - Olga Sagalchik (2145)
Sicilian Scheveningen
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 a6 5.Nc3 b5 6.Bd3 Bb7 7.0-0 d6 8.f4
Now we've transposed into a sharp Scheveningen.
8...Nf6 9.Qf3 Nbd7 10.Be3
The move 10.a3!? defends e4.
Fedorowicz's suggestion may have been first played by Hugh Alexander in a loss to the Indonesian Hiong Tan at Hastings 1962-3. It is the most popular move, from a small sample size, in ChessBase's 2019 Mega database, but scores just 23 percent. My main analysis engines Stockfish10 and Komodo9 like 10.a4!?, which would apparently be a novelty. EG's choice transposes to a popular line that has seen Tal and Judit Polgar on the white side.
10...b4 11.Nb1
I thought 11.Na4!? was a more aggressive try.
11...Be7 12.Nd2 Nc5 13.Kh1
Another way to go would be 13.a3!? bxa3 14.b4 Nxd3 15.cxd3.
13...0-0
This looks level here.
For what it's worth, the engines very slightly prefer Black.
14.Rae1 Nxd3?! (Fedorowicz's punctuation)
Black makes things easier for White. The pressure on e4 is released. 14...g6!? with ideas of ...e5.
Fedorowicz's notes seem to have been truncated. I guess an additional point is that the knight on c5 is a better piece than White's tied-down light-square bishop.
15.cxd3 Nd7 16.Qg4 Kh8 17.Rf3
 |
| How would you defend as Black? |
*****
*****
*****
*****
17...f5? (Fedorowicz's punctuation)
This panic-button move allows White to sac the exchange with a winning game; 17...g6 doesn't look so bad.
The engines give 17...Nf6 with equality (Stockfish10) or a slight edge for Black (Komodo9).
18.exf5!
It seems mean not to give this move an exclam.
18...Bxf3 19.Qxf3 Qb8 20.Nc6 Qe8 21.fxe6 Ne5 22.Nxe5 dxe5 23.f5! (Fedorowicz's punctuation)
Black is in trouble now.
23...Qb5 24.Qe4 Rad8 25.Nc4 Rd5 26.Bf2 Bf6 27.g4 Rfd8 28.g5 Bxg5 29.Nxe5 Kg8 30.Nf7 Rxf5 31.Nxd8 Rxf2 32.e7 Qe8 33.Qe6+ Kh8 34.Nf7+! (Fedorowicz's punctuation)
Simple and strong.
34...Rxf7 35.Qxf7 Qc6+ 36.Kg1 Qc5+ 37.Kf1 1-0
Olga saw some ghosts and then Elina took advantage of the opportunity.
And the "spooky coincidence"? I recognised the winner's name, as four years earlier I played her at the 1996 US Open in Alexandria, Virginia, when my American rating was exactly the same as hers was in the game above. One for a second edition of grandmaster Jim Plaskett's Coincidences book?
Spanton (2054) - Groberman (1684)
Spanish Classical
1.e4 e5 2.Nf3 Nc6 3.Bb5 Bc5 4.c3 Nf6 5.d3!?
More popular are 5.0-0 and 5.d4, but I decided at the board to play an idea I knew Steinitz was fond of, ie to refuse the centre and castle long.
5...d6 6.Be3!?
Seventh-most popular in Mega19, but it was subsequently played by a 2541.
6...Bb6 7.h3
This obviously protects against ...Ng4, but also prepares the thrust g4.
7...0-0 8.Nbd2
Nesevedski - Bogoljubow, Tallinn 1930 saw a more-conventional plan of 8.0-0 Ne7 9.Nbd2, but that game was decided by a strange blunder by Black (1-0, 46 moves).
 |
| Black to make her eighth move |
*****
*****
*****
*****
8...h6!?
The engines are happy with this move, but it does fall in with White's plans by giving a target for the white g pawn. Spanton (2010) - Hamilton McMillan (2020), Paignton 2008, saw the more circumspect 8...Qe7. The game continued 9.Qe2 Nd8, and now I castled long despite Black not having weakened his kingside (0-1, 53 moves).
9.Qe2 Nh7 10.g4!? Ne7 11.0-0-0 c6 12.Ba4 Re8!?
Stockfish10 does not like this, although Komodo9 is not too bothered. As is often the case, moves in positions with opposite-side castling are hard to evaluate unless they are clear blunders.
13.Rdg1 Ng6 14.Nf1 d5
Central counterplay is the traditional way to meet a flank attack. One of the ideas behind White's 5.d3!? is to reinforce e4 in case of this eventuality.
15.Ng3 Nf6 16.g5 dxe4 17.dxe4 hxg5?
The engines' second-choice, but a mistake. They prefer 17...Nf4, when Stockfish10 calculates 18.Qf1 as giving White a winning advantage, but Komodo9 reckons it leaves Black with a slight edge (see note to Black's 12th move).
18.Nxg5 Bxe3+ 19.fxe3!
Taking the f4 square away from a black knight, and leaving White's queen with access to the h file at h5 or h2.
19...Nh7 20.Nxh7 Kxh7 21.Qh5+ Kg8 22.Nf5 Bxf5 23.exf5 Qh4
The best defence, but White is positionally winning after the queens come off following 24.Qg4 or 24.Qxh4. Instead, concentrating too much on the kingside, I came up with a real lulu:
24.Qf3?? Qxa4 25.fxg6 Qxa2
Possible because the queen covers f7.
26.Rg5?
White remains a pawn down after 26.h4 but has full compensation, according to the engines.
26...Rad8?
Black seems to have a definite edge after the engines' mainline of 26...Re7 27.Rf1 fxg6 28.Rxg6 Rd8.
27.Rhg1?
White gets his pawn back with 27.gxf7+ (27.Qxf7+ comes to the same thing) Qxf7 28.Qxf7+ Kxf7 29.Rhg1.
27...f6 28.R5g2 Rd6 29.Qh5 Qa1+ 30.Kc2 Qa4+
Black offered a draw. Objectively I should accept, but my usual optimism, and the rating difference (significant despite my opponent being a junior and therefore almost certainly underrated), persuaded me to play on.
31.Kb1 Qe4+ 32.Rc2 Rd2?
This looks powerful, but better was 32...Qc4 or 32...Qd5, allowing a check at h8 to be met with ...Qg8.
33.Qh7+ Kf8 34.Qh8+ Ke7 35.Qxg7+ Kd8 36.Qxf6+ Kc8 37.Rgc1 Red8?
Black draws with 37...Rxc2 38.Rxc2 Rd8.
38.Qe6+?
Good enough for a draw, but White is slightly better, according to the engines, after 38.g7, the point being that now ...Rxc2?? fails to Qxd8+.
38...Kb8 39.Qg4 Qxg4 40.hxg4 Rxc2?
An instructive mistake that helps the white king to centralise. The obvious 40...R2d6 seems to be a dead draw.
41.Kxc2 Rg8 42.Kd3 Rxg6 43.Rg1
The rook-and-pawn ending is a little better for White, who has an outside passed pawn and the more-centralised king.
43...Kc7
Possibly better was 43...Rg5 to enable the rook to defend e5 as well as pressurise g4.
44.Ke4
White in turn could probably have played 44.g5 to prevent the above scenario.
44...Rg5 45.Kf3 Kd6 46.Rd1+ Ke7
The engines prefer 46...Ke6!? even though this lets White's rook onto the back rank.
47.Rh1
A neat switchback that forces Black further on the defensive.
47...Rg7 48.Rh5 Ke6 49.e4?!
The engines reckon White keeps an edge after 49.g5 or 49.c4.
49...Rf7+ 50.Rf5
 |
| Black to play and draw |
*****
*****
*****
*****
50...Rh7?
Black is bluffed. Perfectly safe was, for example, 50...a5, as 51.Rxf7?? leads to a lost pawn-ending - White's king will be diverted to the queenside by Black creating a passer there, leaving Black to win White's g and e pawns.
51.g5 Rh1?!
There does not seem to be any way for Black to save this - White's passed pawn is too strong - but 51...Rh2 would have given me something to worry about.
52.Rf6+ Ke7 53.Kg4 Rh2
An admission that Black's 51st lost a tempo.
54.b4 Rc2 55.Rf3 Ke6 56.Kh5 Rh2+ 57.Kg6 Rc2
If 57...Re2, then 58.Rf6+ Ke7 59.Kf5 etc,
58.Kg7 Re2 59.Kf8 Rg2 60.Rf6+ Kd7 61.g6 1-0
No comments:
Post a Comment