Monday 31 May 2021

Beat The ... Bird

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

Bird's Opening: 1.f4 is rare among the elite but popular with club players who hope to get their opponents out of book as soon as possible.
Black has many possible replies, but of those moves in Mega21 that occur at least 200 times, the most successful statistically by a long way is 1...g6, which scores 58%.
One significant point in its favour is that discourages White from following up with an early b3 (some whites would love to play 1.b3 followed by a quick f4 to clamp down on the e5 square, but are put off by the possibility of 1.b3 being met by 1...e5, which is indeed Black's commonest response).
Position after 1...g6
White has three reasonably-common continuations.

A) 2.Nf3 (2,972 games)
After 2...d5!? (Black often avoids playing 1...d5 because it can be met by 2.b3, but with 1...g6 discouraging b3, then the move ...d5 becomes desirable) the line splits.
A1 3.e3 Bg7, after which the line splits again.
A1.1 4.Be2 c5 5.0-0 Nc6, when A1.1a 6.d3 Nf6 (this and 6...Nh6 are equally successful, but Stockfish13 and Komodo12.1.1 much prefer ...Nf6) 7.Qe1 0-0 8.Qh4 b6!? 9.Nbd2 Ba6 scores 79% for Black, albeit from a small sample, and A1.1b 6.d4 Nf6 7.c3 0-0 8.Ne5 Qc7 scores 70% for Black, albeit from a small sample.
A1.2 4.d4 Nh6!?, when A1.2a 5.Bd3 0-0 6.0-0 c5 7.c3 Qc7 scores 87% for Black and A1.2b 5.c4 dxc4!? 6.Bxc4 0-0 also scores 87% for Black, albeit from a small sample.
A1.3 4.Nc3 Nh6!? 5.b3 0-0 6.Bb2 d5 scores 75% for Black, albeit from a small sample.
A1.4 4.c4 Nf6, when A1.4a 5.Nc3 0-0 6.d4 c6 7.Bd3 c5!? scores 80% for Black, albeit from a small sample, and A1.4b 5.cxd5 Nxd5 6.Nc3 0-0 7.Bc4 Nb6 scores 71% for Black, albeit from a small sample.
A2 3.g3 Bg7 4.Bg2 Nh6!?, after which the line splits again.
A2.1 5.0-0 0-0 6.d3 d4!? 7.c3 c5 8.e4 dxe3 9.Bxe3 Qc7 scores 58% for Black.
A2.2 5.d3 d4!? 6.c3 c5 7.e4 dxe3 8.Bxe3 Nd7 scores 79% for Black, albeit from a small sample.

B) 2.e4 (403 games)
After 2...d5!? the line splits.
B1 3.e5 c5 4.Nf3 Nc6 5.c3 Nh6!? scores 80% for Black, albeit from a small sample.
B2 3.exd5 Qxd5 4.Nc3 reaches a position in Mega21 in which Black scores 75% with both 4...Qa5 and 4...Qd6, albeit from small samples.

C) 2.g3 (95 games)
After 2...d5!? the line splits.
C1 3.Bg2 Bg7 4.Nf3 is a transposition to A2.
C2 3.Nf3 Bg7 4.Bg2 is also a transposition to A2.

Sunday 30 May 2021

Calvià (conclusion)

MY final score at Calvià of +4=2-3 saw me gain 37.6 Fide elo.
With Guernsey strongly rumoured to be cancelled again this year, and the 21st Calvià chess festival announced for October 9-17, there is a strong chance I will play in Mallorca again.
Details of the tournament are at https://www.calviafestival.com/
Early entries, including ones from England and Ireland, are at https://chess-results.com/tnr560234%20.aspx?lan=1&turdet=YES+&flag=30

Saturday 29 May 2021

Calvià (part nine)

Adrian Ponce Cano (2169) - Spanton (1793)
Calvià U2350 Round 9
Sicilian Alapin/Hyper-Accelerated Dragon
1.e4 c5 2.Nf3 g6 3.c3 Bg7 4.d4 cxd4 5.cxd4 d5 6.e5 Bg4 7.Bb5+ Nd7!?
The point of avoiding 7...Nc6 is to avoid contracting a backward pawn on a half-open file after 8.Bxc6+!?
8.0-0 Nh6?!
Russian grandmaster Boris Savchenko has played both 8...e6 and 8...a6, the two moves suggested by Stockfish13 and Komodo12.1.1. The text is strongly disliked by the engines.
9.Nbd2 0-0 10.h3 Bxf3 11.Nxf3 a6 12.Bd3
Position after 12.Bd3
12...e6
Miroslav Zacharias (1727) - Spanton (1923), Trebic Highlands (Czechia) 2018, saw 12...f6 13.exf6 Nxf6 14.Ng5 Qd6 15.Re1 with a large advantage for White, according to the engines (but 0-1, 58 moves).
13.Re1 Qb6 14.a4 Rae8?!
Over-optimistically expecting the game to be decided on the kingside. The engines prefer 14...Nf5 or 14...Nb8!?, while giving White a large advantage. 
15.a5 Qa7 16.Qa4 Nb8 17.b4
White's queenside initiative is winning, according to the engines.
17...Nc6 18.b5 axb5 19.Bxb5 Rc8 20.Ba3 Rfd8 21.Bc5 Qa8 22.Bb6 Rf8 23.Rec1 Nf5 24.a6 Nfe7 25.Nc5 Rfe8
Or 25...Bh6 26.axb7 Qxb7 27.Rcb1 with a large advantage.
26.Bxe7 Rxe7 27.axb7 Qxb7 28.Ba6 Qb2 29.Bxc8 Ra7 30.Qd1 1-0

Friday 28 May 2021

Calvià (part eight)

Spanton (1793) - Enrique Fernando Asensio Ramos (2070)
Calvià U2350 Round 8
Hippopotamus Defence
1.e4 g6 2.Nc3 Bg7 3.Nf3 a6
Eight games reach this position in ChessBase's 2021 Mega database, but none saw:
4.a4
However we return to theory after ...
4...b6 5.d4 Bb7 6.Bc4
... when this move reaches a position my analysis engines Stockfish13 and Komodo12.1.1 reckon is very favourable for White, although 6.Bc4 scores just 33% in Mega21.
6...h6 7.h3
Again the game leaves known theory, although not for long.
My idea with the text was to rule out Black playing ...Nf6-g4. Stockfish13 gives White the upper hand; Komodo12.1.1 has White slightly better.
6...e6 8.Bf4 d6 9.Qd2 Ne7 10.h4!?
With g4 unlikely to be visited by a black piece any time soon, or so I thought, it seemed fine to use the h pawn to gain kingside space.
10...Nd7
A picturesque position after 10...Nd7
11.b4!?
Once again leaving known theory, and this time for good. Known moves are 11.0-0, 11.0-0-0 and 11.Kf1!?
11...Nf6 12.Bd3 Ng4
So a black piece occupies g4 after all. The position is dead-equal, according to the engines.
13.a5!?
The engines dislike this, preferring 13.0-0.
13...b5 14.Rd1 0-0 15.0-0 Qc8 16.Ne2 Rd8 17.h5!? e5 18.Be3
I wanted to play 18.hxg6!? exf4 19.gxf7+ Kxf7 20.Nxf4, but did not believe it. However the engines reckon Black's exposed king gives White more than enough compensation. Indeed they reckon Black should reject the piece sacrifice and instead play 18...fxg6, 18...Nxg6 or 18...f5.
18...f5 19.exf5?
The engines give 19.d5, and if 19...f4 they reckon 20.Bxf4 exf4 21.Nxf4 again gives White good compensation. But in this line they like 19...c6 or 19...Rf8.
19...e4 20.Bxe4 Bxe4 21.fxg6
White has three pawns for a bishop, whereas in previous lines he only had two, but the engines reckon here Black is clearly winning.
21...Rf8 22.Nf4?!
Better, according to the engines, is 22.Nh4, 22.Bf4 or even 22.Ng5!?
22...Nd5 23.Ne1 Qf5 24.Nxd5 Bxd5 25.f3 Rae8 26.fxg4!?
It is hard to come up with anything significantly better.
26...Qxf1+ 27.Kh2 Rxe3!? 28.Qxe3 Rf2 29.Rd2 Rxd2 30.Qxd2 Bf6 31.Qxh6?
This speeds the end, which came after:
31...Qxe1 32.Qh7+ Kf8 33.g7+ Bxg7 34.Qf5+ Bf7 35.Kh3 Qe3+ 36.Kh4 Qe1+ 37.Kh3 Qh1+ 38.Kg3 Bxd4 39.Qe4 Be5+ 40.Kf2 Qh4+ 0-1

Thursday 27 May 2021

Calvià (part seven)

Guillermo Simo Bordoy (1903) - Spanton (1793)
Calvià U2350 Round 7
Sicilian Accelerated Dragon
1.e4 c5 2.Ne2 g6 3.d4 cxd4 4.Nxd4 Nc6 5.Nxc6!?
Not a popular choice, but there are more than 2,300 examples in ChessBase's 2021 Mega database.
5...bxc6 6.Qd4
This seems to have been first played by Emanuel Lasker in 1892 in a match-game win against Henry Bird.
6...Nf6
Bird played 6...f6, which is today generally regarded as too passive.
7.e5 Ng8!?
My turn for passivity. More popular is 7...Nd5.
8.Bc4 Bg7 9.0-0 d5 10.Bb3 Nh6 11.Bf4!?
This is not much liked by the analysis engines Stockfish13 and Komodo12.1.1. Known moves are 11.f4 and 11.c4, but these are not liked by the engines either. Stockfish13 suggests 11.Re1 0-0 12.Qc3, while Komodo12.1.1 gives 11.Ba4 Bb7 12.Bxh6, in each case with approximate equality.
11...Nf5 12.Qd2 0-0
The engines give 12...g5!?
13.Nc3 e6 14.Bg5 f6 15.exf6 Bxf6 16.Bxf6 Qxf6 17.Na4 Nh4?!
Probably better is 17...e5 to anticipate White's plan to restrain the black centre.
18.f4! Nf5 19.Rae1 Nd6
Planning to occupy the e4 outpost and so mask the weak e pawn.
Position after 19...Nd6
*****
*****
*****
*****
*****
20.Nc3
The engines prefer 20.c3!?, not worrying about 20...Ne4. Perhaps the point is other black pieces are not well-placed to support the knight.
20...Ba6 21.Rf2 Rae8 22.Rf3?!
Admitting that his last move lost a tempo - or perhaps this move loses a tempo, as the engines basically see no difference in value between 21.Rf2 and 21.Rf3. Instead, after 21.Rf2 Rae8, they want White to play 22.Qe3!? d4 23.Qd2 c5 24.Ne4.
22...c5 23.Qf2 d4
*****
*****
*****
*****
*****
24.Nb1?
Black has a winning advantage after this retreat, according to the engines. They give 24.Na4! c4 25.Nc5 cxb3 26.Nxa6 bxa2 27.Nc5, reckoning Black has a slight edge.
24...Bb5
Enough for a large advantage, but even stronger, according to the engines, is 24...Bb7.
25.f5!?
Best may be 25.Nd2 c4 26.a4 Bc6 27.Nxc4 Nxc4 28.Bxc4 Bxf3 29.Qxf3 Qxf4 30.Qxf4 Rxf4, although Black has a large advantage as 31.Bxe6+? loses to the simple 31...Kg7, eg 32.Re2 d3! 33.cxd3 Kf6 34.Bg4 Rxe2 35.Bxe2 Rxa4 etc.
25...Nxf5 26.Qd2!?
Again the engines prefer Nd2.
26...c4 27.g4 cxb3 28.gxf5 bxa2
Black has emerged from the tactics three pawns up. The game finished:
29.Na3 Bc6 30.Rg3 exf5 31.Ra1 Bd5 32.Nb5 Rb8 33.Nc7 Bc4 34.b3 Rfc8 35.Rxa2 Rxc7 36.bxc4 Rb1+ 37.Kf2 Re7 38.Qa5 Rbe1 39.Qd5+ (or 39.Qxe1 f4!) Kg7 40.Rd3 Qh4+ 0-1

Wednesday 26 May 2021

Morphing The French XIX

ANOTHER attempt to play against the French Defence in the style of Paul Morphy.
Spanton (1793) - Juan Ramon Galiana Salom (2248)
Calvià U2350 Round 6
French Exchange
1.e4 e6 2.d4 d5 3.exd5 exd5 4.Nf3 Bd6
This is Black's most-popular reply in ChessBase's 2021 Mega database.
5.Bd3 Bg4 6.0-0 Nc6 7.Re1+ Nge7 8.c3!?
The type of move Morphy seemed to avoid, but 8.Be3 is no improvement, according to the analysis engines Stockfish13 and Komodo12.1.1. One point about the text is that as well as protecting d4 it prepares the thrust b4 in the likely case of Black castling long.
8...Qd7 9.Nbd2 0-0-0!?
Easily the most-popular move in the position, but the engines do no like it, preferring short castling.
10.b4 Ng6
Spanton (1855) - Leopold Adrian (2061), Biel Corona (Switzerland) 2020, went 10...Rde8 11.Nb3 Nf5 12.Be3 with an unclear but roughly equal position (0-1, 81 moves): https://beauchess.blogspot.com/2020/07/morphing-french-xv.html
11.b5 Nce7 12.Qa4 Kb8 13.Ba3
A possible improvement is the more-popular 13.Ne5.
13...Bxf3!?
An apparent-novelty that is quite liked by Komodo12.1.1 but not by Stockfish13, which much prefers 13...Bxa3. Commonest in Mega21 is 13..Nf4.
14.Nxf3 Nf4 15.Bf1
The engines give 15.Bxd6 cxd6!? 16.Bf1 h5, preferring White.
15...Nc8 16.Ne5 Bxe5 17.dxe5
The game has transposed into a known position in which a 2420 played 17...g5, but the engines prefer:
17...Qg4!?
I replied:
18.Kh1
Probably better is the engines' 18.Qc2!? The text overlooks a tactical shot.
Black to play and win a pawn
18...Nd3!? 19.Qxg4 Nxf2+ 20.Kg1 Nxg4
Black has won a pawn but, according to the engines, the bishop-pair gives White more than enough compensation after 21.h3 Nh6 22.Bd3.
Going back to the diagram, the engines reckon best-play runs 18...Nb6 19.Qd1 Qxd1 20.Raxd1 with a slight edge for White.
21.Rad1? Nb6?
Black consolidates his advantage, according to the engines, after 21...Rhe8 22.c4 Nxe5 23.cxd5 Nd6 24.Bb2 f6.
22.Rd4 h5 23.Bc1?
Still the way to go was h3.
23...Rhe8 24.Bf4 f6 25.e6 g5 26.h3
At last, but the knight will not be so awkwardly placed on h6 as it would have been earlier.
26...Nh6 27.Bh2 Nf5 28.Rd2 Ng7
The engines prefer 28...Nd6!?, the idea being that if, as in the game, White replies 29.Rf2, Black has 29...Ne4.
29.Rf2 f5 30.Bd3 f4?!
Better, according to the engines, is 30...Rxe6 31.Rxe6 Nxe6 although White still has good compensation for a pawn after 32.Rxf5.
*****
*****
*****
*****
*****
31.g3?
This loses a second pawn, whereas after 31.e7! Re7 32.Re5 Rdxe7 33.Rxg5 Black cannot defend all his pawn weaknesses.
31...fxg3 32.Bxg3 Rxe6 33.Rxe6 Nxe6 34.Rf6 Nc5 35.Rf7 Nbd7
White's bishop-pair gives a lot of compensation, but not enough for two pawns, according to the engines.
36.Bf5?!
This is probably wrong because it lets Black activate his passive rook..
36...Rf8! 37.Rxf8 Nxf8 38.Be5 Nce6 39.Kf2 b6 40.Ke3 Kb7 41.Kd3 a6 42.c4 dxc4+ 43.Kxc4 axb5+ 44.Kd5!?
44.Kxb5 may be objectively better, but I felt it made Black's task easier.
44...Nc5 45.Bf6 g4 46.hxg4 hxg4 47.Bxg4 Nfd7?!
Probably better is 47...c6+ 48.Kd4 (48.Kd6? Ne4+) Nfe6+ 49.Ke5 Nc7.
48.Bd8?!
Probably better is 48.Bd4, and if 48...c6+ then 49.Kd6 gives White full compensation for being two pawns down, according to the engines.
48...Nb8 49.Be7 Nba6 50.a3 Na4 51.Bd7 Nb8 52.Be8
52.Bxb5!? c6+ 53.Kc4 cxb5+ 54.Kxb5 Nc3+ gives a position in which, according to the endgame tablebase Syzygy, Black has a win but cannot achieve it in under 50 moves. However Black would have good practical chances.
52...c6+ 53.Ke5 Nb2 54.Kd4 Kc7 55.Ke5 Nd7+ 56.Kd4 Na4 57.Bg6 Nac5 58.Bf7 Nb7 59.Bb3 Nd6 60.Kc3?!
This eases Black's task, according to the engines, who prefer 60.Be6 or 60.Bg5.
60...Nc5 61.Bc2 Nde4+ 62.Kd4 Nd6 63.Kc3 Kd7?!
Simpler seems to be the engines' 63...Nc4.
64.Bh4 Ne6 65.a4?!
The engines do not like this exchange, preferring a non-committal bishop move such as 65.Bg3.
65...bxa4 66.Bxa4
Position after 66.Bxa4
White's situation is not quite as desperate as it might appear, in that the game is drawn if White can capture both black pawn,s even if it costs both bishops.
66...b5 67.Bc2 c5 68.Bf2 Kc6 69.Bd3 c4 70.Bc2 Nc5 71.Bg1 Nde4+ 72.Kd4 Nd2
This is Komodo12.1.1's choice at first, but both engines come to prefer 72...Ng5!?, after which 73.Be3 Nge6+ 74.Kc3 Nc7 75.Kd4 leaves the engines disagreeing as to whether Black's advantage is winning (Komodo12.1.1) or merely gives the upper hand (Stockfish13).
73.Be3 Nf3+ 74.Kc3 Ne1 75.Bg6 Ned3 76.Be8+ Kb6 77.Bg6 Ne5 78.Be8 Ned3 79.Bg6?!
Simpler is 79.Bf7 Ne5 (prevents 80...Bxc4) 80.Bd5, when it is hard to see how Black can prevent a draw.
79...Ka5 80.Bf7?!
But now Bf7 makes life difficult for White. Best probably is 80.Bf5, with good drawing chances.
*****
*****
*****
*****
*****
80...Na4+
The engines are very keen on 80...Ne5!?, which is winning (Komodo12.1.) or at least gives Black the upper hand (Stockfish13). But the Syzygy tablebase shows the position to be drawn after 81.Bxc5, although Black would retain practical chances.
81.Kd4 Nac5?!
81.Nab2 keeps the game going.
82.Bxc4 bxc4 83.Kxc4 Ka4 84.Bxc5 Nxc5 85.Kxc5 ½–½

Tuesday 25 May 2021

Calvià (part five)

Santiago Juan Mas (1975) - Spanton (1793)
Calvià U2350 Round 5
Morra Gambit Declined
1.e4 c5 2.d4 cxd4 3.c3 g6!?
Not a popular move, but it scores a very healthy 58% in ChessBase's 2021 Mega database.
4.Nf3!?
Much more popular is 4.cxd4, but the game soon transposes to similar lines.
4...Bg7 5.cxd4 d5
It is generally considered important for Black to challenge the white centre, by 5...d5 or 5...Nf6, before White has time to consolidate. However grandmasters have also played 5...d6.
6.exd5
Kasparov is among those who played the more popular 6.e5, which scores 50% in Mega21 compared with 49% for the text.
6...Nf6 7.Bb5+ Nbd7 8.Nc3 0-0 9.0-0 Nb6 10.Re1 Nbxd5 11.Bg5 Be6 12.Qd2 Rc8 13.Rac1 Nxc3!?
A committal move, but it was IM Andrew Martin's choice in a 1990 game in Guernsey and GM Maxim Matlakov's in a 2016 rapid game in Talinn, Estonia.
14.bxc3 Bd5
Position after 14...Bd5
15.Bd3!?
This apparent-novelty is quite liked by my main analysis engines Stockfish13 and Komodo12.1.1.
Gerhard Reink - Martin (2415), Guernsey 1990, continued 15.Bxf6 Bxf6 16.Ne5 Qa5, when Black is already better, according to the engines (0-1, 32 moves).
Margeir Petursson (2509) - Matlakov (2682), Keres Memorial Rapid (Talinn) 2016, saw 15.Qe3 Qa5 16.Bf1 Bxf3!? 17.Bxf6!? Bxf6 18.Qxf3 Rc7, when again Black is better, according to the engines (0-1, 26 moves).
15...Bxf3 16.gxf3 Qd7 17.Kg2 e6 18.h4 h5 19.Bh6?!
Manoeuvring to give up the bishop-pair seems strange as White will not be in much of a position to exploit Black's weak kingside dark squares.
19...Nd5 20.Bxg7 Kxg7 21.Be4 Rfd8 22.c4 Nb6 23.c5!? Nd5 24.Qb2
White's queenside pressure does not quite make up for his shattered pawn-structure, according to the engines.
24...Qe7 25.Bxd5 Rxd5 26.Re4 Rcd8
Probably not 26...f5?!, when the engines reckon 27.Re5 equalises.
27.Rce1 Kh7 28.Qa3?
An oversight the drops a pawn. The engines give 28.a4, with slightly the better game for Black.
28...b6 29.Qb3 bxc5 30.dxc5 Rxc5 31.Qb4 Rc7 32.R1e3 Qxb4 33.Rxb4 Rd2 34.Ra4 Kg7 35.a3?!
Komodo12.1.1 likes this for a while, but it is probably a mistake. Better seems to be 35.Kg3 or 35.f4.
35...Kf6?!
The engines much prefer 35...Rcc2 36.Rxa7 Rxf2+ 37.Kg3 Kf6.
36.Ree4
The engines prefer 36.Re1!?, the idea being to meet 36...Rcc2 with 37.Rf1, when it is not clear how Black makes progress except by painstaking baby steps on the kingside.
36...Rd5
Again ...Rcc2 may be better.
37.Rf4+ Ke7 38.Rfb4 Rdc5 39.f4?
This gains space but weakens the f pawn and the g4 square. The engines suggest 39.Ra6 or 39.Rb8, while still giving Black a winning advantage.
39...a5 40.Rd4 Rd7 41.Re4 Rd3 42.Re2 Kf6 43.Rb2 Rdc3 44.Rd4 Kf5 45.Rb7 f6 46.a4 R3c4 47.Rxc4 Rxc4 48.Rb5+
Or 48.Kg3 Rxf4 49.Rb5+ Ke4.
The game finished:
48...Kg4 49.Rxa5 Rxf4 50.Ra8 Kxh4 51.a5 Kg5 52.a6 Ra4 53.a7 Ra3 54.Kh2 e5 55.Kg2 Kf5 56.Kh2 e4 57.Kg2 Ra4 58.Kh2 g5 59.Kg2 Kg4 60.Re8
Barely better, according to the engines, is 60.Rf8.
60...Rxa7 61.Rxe4+ Kf5 62.Rb4 Re7 63.Ra4 Re4 64.Ra8 Kf4 65.Rh8 h4 66.Ra8 f5 67.Rh8 Re2 68.Rh7 Rb2 69.Rc7 Rb4 70.Rh7 Rb3 71.Ra7 h3+ 72.Kg1 Kf3 0-1 (Time)

Monday 24 May 2021

Calvià (part four)

Spanton (1793) - José Luis Moreno Mateos (1812)
Calvià U2350 Round 4
Caro-Kann Bronstein-Larsen
1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.Nxf6+ gxf6 6.c3!?
This quiet continuation is the commonest in ChessBase's 2021 Mega database.
6...Bf5 7.Bc4!?
Usual is 7.Nf3.
7...e6 8.Ne2 Bd6!?
This is the most-popular move in Mega21, but the analysis engines Stockfish13 and Komodo12.1.1 much prefer 8...h5 or 8...Nd7.
9.Ng3 Bg6 10.Qf3 Nd7 11.Bh6
Position after 11.Bh6
*****
*****
*****
*****
*****
11...f5?!
An apparent-novelty, and probably not a good one. Normal is 11...Rg8, while the engines like 11...Bf8!?
12.h4 f4!?
Giving up a pawn is Black's best chance, according to the engines.
13.Bxf4 Bxf4 14.Qxf4 Nb6 15.Bb3 Nd5 16.Qe5 Qf6 17.Qxf6 Nxf6 18.0-0-0 h6 19.Rhe1 0-0-0 20.Bc2 Bxc2 21.Kxc2 Rhg8 22.c4?
This gives the pawn back. The engines reckon 22.Rd3 keeps the upper hand for White.
22...Rg4 23.Kc3 Rxh4 24.Rh1 Rxh1 25.Rxh1 Rh8
Possible is 25...Rg8!? as 26.Rxh6 runs into 26...Ng4 27.Rh7 Nxf2 with at least equality.
26.Kd3 Ng4 27.Ke2 Kd7 28.Ne4 Ke7 29.f3 Nf6 30.Nc5 b6
Probably better is the engines' 30...h5!?, freeing the rook so 31.Nxb7 can be met by 31...Rb8.
31.Nd3 Rg8?
The engines suggest 31...Nd7 or 31...h5.
32.Rh2?
Strong is 32.g4 as the h pawn must fall.
32...Rd8?
The engines give 32...Rg5 or 32...h5.
33.Ne5 Kf8
Not 33...Rxd4?? 34.Nxc6+.
34.Rxh6 Ng8?
Less passive is 34...Kg7.
35.Rh4 Rc8 36.Rh7 Rc7 37.Ng6+ Ke8 38.Rh8 fxg6 39.Rxg8+ Kf7 40.Rh8
White is only one pawn up in a rook-and-pawn-ending, but the white rook is more active and Black has two isolanis.
40...Kf6 41.Ke3 c5 42.Rh4 a6 43.Kd3 b5 44.dxc5 Rxc5 45.b4 bxc4+ 46.Rxc4 Rd5+ 47.Rd4 Rh5 48.Kc4 Ke5 49.a4 Rh1 50.Kc5 Rb1 51.Re4+ Kf5 52.Kb6 Rb2 53.g4+ 1-0
JM tried to play 53...Ra2, and then resigned.

Sunday 23 May 2021

Calvià (part three)

Jordi Baldo Company (1565) - Spanton (1793)
Calvià U2350 Round 3
Sicilian Accelerated Dragon
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Bg7 6.Be3 Nf6 7.f3!?
There are more than 6,000 games with this move in ChessBase's 2021 Mega database, but much more popular are 7.Be2 and especially 7.Bc4.
7...0-0 8.Qd2 d5
The thematic point behind the Accelerated Dragon - playing ...d5 without spending a tempo on ...d6.
9.Nxc6 bxc6 10.exd5
10.0-0-0 and 10.e5 are major alternatives.
10...cxd5 11.0-0-0 Qa5 12.a3!?
The main line in Mega21 runs 12.Nxd5?! Qxa2 13.Nxe7+ Kh8 14.Qb4, when White has won a pawn but has a lost game, according to the analysis engines Stockfish13 and Komodo12.1.1.
12...Bb7
The engines prefer 12...Rb8 or 12...Qc7!?
Position after 12...Bb7
13.Nb5!?
This apparent-novelty is liked by Komodo12.1.1 but not by Stockfish13, which prefers 13.Bd4. The only game in Mega21 saw 13.Be2?!, which lets Black take the initiative with 13...e5.
13...Qxd2+ 14.Rxd2 a5
Saving the pawn, but the engines prefer 14...e5!?
15.Nc7 Rac8 16.Bb6!? Bh6 17.Bxa5 e5
There is no rush to take the exchange.
18.Nb5 Bxd2+ 19.Kxd2
The bishop-pair and a pawn give White full compensation for the exchange, according to Stockfish13, but Komodo12.1.1 gives Black a slight edge.
19...Ra8 20.Bb4 Rfd8?? 21.Be7 Kg7 22.Bxd8 Rxd8
White has regained the exchange, remains a pawn up and is winning, according to Stockfish13, although Komodo12.1.1 'merely' gives White the upper hand.
23.Kc1 d4
Creating a knight outpost at e3.
24.Bd3 Nd5 25.Re1!?
The engines' choice. A major alternative is 25.Be4 f5 26.Bxd5 Bxd5, although then Black has the better minor piece for cooperating with a rook.
25...Ne3 26.Re2 f5 27.c3 e4??
Sensible is 27...Kf6, when White is still better but Black has counterplay.
28.Nxd4 exd3 29.Rxe3 Kf6 30.h4
There seems no good reason not to play 30.Rxd3.
30...f4 31.Re6+
Again Rxd3 is possible but the text is preferred by the engines.
31...Kf7 32.Kd2? Rxd4! 33.cxd4??
White is still better, according to the engines, after 33.Rb6, but the text loses.
33...Kxe6 34.Kxd3 Kd5
The game finished:
35.Kc3 Ba6 36.b4 Bb5 37.Kb3 Kxd4 38.a4 Bf1 39.b5 Kc5 40.Kc3 Bxg2 41.Kd3 Bxf3 42.Kd2 Bd5 43.Ke2 Kb6 44.Kf2 Ka5 45.Ke2 Kxa4 46.b6 Kb5 47.b7 Bxb7 48.Kd3 Kc5 0-1

Saturday 22 May 2021

Morphing The French XVIII

ANOTHER attempt by me to play against the French Defence in the style of Paul Morphy.
Spanton (1793) - Jordi Salvador Ramis Sebastia (1519)
Calvià U2350 Round 2
French Exchange
1.e4 e6 2.d4 d5 3.exd5 exd5 4.Nf3 Nf6 5.Bd3 Bg4 6.0-0 Be7 7.Re1 0-0 8.h3 Bh5 9.g4!?
Not a popular choice. The analysis engines Stockfish13 and Komodo12.1.1 reckon White gets a tiny pull with 9.Bg5, which has been played by Mark Hebden.
9...Bg6 10.Ne5 Nbd7 11.Bg5!?
The engines strongly dislike this, reckoning the game is roughly equal after 11.Nxg6 (Stockfish13) or 11.Nc3 (Komodo12.1.1).
11...Nxe5 12.dxe5 Ne4 13.Bxe7?!
Black is only slightly better, according to the engines, after 13.Bxe4 Bxe4 14.Be3.
13...Qxe7 14.Bxe4
This is best, according to the engines, but after ...
14...dxe4
... White has big problems.
15.Nc3
Even worse seems to be 15.f4?! exf3 16.Qxf3 Bxc2 (or 16...f5!?, as suggested by the engines). Also very good for Black is 15.Qd4 Rad8!? 16.Qc3 Qg5.
15...Qxe5
Even better, according to the engines, is 15...Rad8 16.Qe2 Qxe5, and if 17.Nxe4 then simply 17...Qxb2.
16.f4??
I cannot recall, but I guess I thought half-opening the f file might give me an initiative. Instead I lose a second pawn and weaken my king. The engines like 16.Qd5!, which gives White a surprising amount of counterplay, eg 16...Qxd5 17.Nxd5 Rae8 18.Nxc7 Rc8 19.Nd5 Rxc2 20.Ne7+, or 16...Qf4 17.Qxb7.
16...Qxf4 17.Qe2 Rfe8 18.Nd5 Qg3+ 19.Qg2 Qxg2+?!
Exchanging queens brings an ending closer, but the white king goes from cowering in the corner to being more active than its black counterpart.
20.Kxg2 Rac8?!
Passive rook-play is rarely correct. Here 20...Rad8 is possible as 21.Nxc7? Rd2+ is good for Black.
21.Rad1 h5 22.Nf4 hxg4 23.hxg4 Bh7 24.Nd5 c6 25.Ne3 Re5?
Black should contest the open file.
26.Rd7 Rb8
Or 26...Rb5 27.b3, when the black pieces are badly uncoordinated.
27.Kg3
The engines prefer 27.Red1 or 27.Rc7.
27...Kf8
And here the engines prefer 27...g5.
28.Red1 Re7
Again the engines like ...g5.
29.Rd8+ Rxd8 30.Rxd8+ Re8 31.Rd7 Rb8
Black remains two pawns up but all of the white pieces are better than the black ones.
32.Kf4 a5 33.Nf5?!
This lets Black swop off his bad bishop for the good white knight.
33...Bxf5 34.gxf5 Re8 35.Ke3 b5 36.Rc7 Re5 37.Rxc6 Rxf5 38.Kxe4 g6 39.Ra6 b4 40.Ra7?!
Almost certainly better is creating a passed pawn by 40.c4 to give counterplay.
40...Kg7?!
Here 40...Rc5 41.Kd3 Kg7 is very strong.
41.c3 bxc3 42.bxc3 Kf6 43.c4 Re5+?!
The engines reckon Black still has an edge after 43...Ke6.
44.Kd4 g5
Black has nothing better, according to the engines. eg 44...Rf5 45.c5.
45.c5
A simpler draw is to be had by 45.Rxf7+!?
45...g4 46.c6!?
The engines reckon 46.Ra6+ Kf5 47.Ra8 is dead-equal.
46...g3 47.Rxf7+??
This trick no longer draws. Instead 47.c7 Re8! (47...g2 48.Ra6+ Re6 49.Rxe6+ fxe6 is a tablebase draw) 48.Rb7 g2 49.Rb1 Rc8 50.Rg1 Rxc7 51.Rxg2 leaves Black needing to find 51...Rb7! for a tablebase win.
47...Kxf7
It was only now that I saw how Black wins.
48.Kxe5
Black to play and win
*****
*****
*****
*****
*****
48...g2??
Missing a simple win by 48...Ke7 or 48...Ke8.
49.c7 g1=Q 50.c8=Q Qg3+ 51.Kd4 Qf2+ 52.Ke5 Qe3+ 53.Kd5 Qd3+ 54.Ke5 Qe3+ 55.Kd5 ½–½

Friday 21 May 2021

Calvià

LAST October I flew to Mallorca to play at Calvià, a seaside resort next-door to Magaluf.
I had considered the tournament previously but been put off because the dates clashed with Guernsey and the rounds started at 8.30pm.
But beggars cannot be choosers, so I was happy to give it a try even though it meant wearing a facemask most of the time in Mallorca, including outdoors.
I stayed at a hotel in Magaluf - the only one in the area that was open as far as I could discover.
Magaluf resembled a ghost town, with almost every cafe, restaurant and bar closed.
But I was happy having nearly all of my meals in my hotel room, and I found a cafe run by a Spaniard who turned out to have been born in Hammersmith, West London, so most mornings I enjoyed an espresso or two on his premises' veranda while soaking up Mediterranean sunshine.
I did not post about my participation for various reasons, including the fact that several of my games did not finish till past midnight.
But with unofficial news that Guernsey will be cancelled again this year (so playing at Calvià this coming October is a distinct possibility), now is as good a time as any for going through my games from last October.
Nicolo Napoli (2309) - Spanton (1793)
Calvià U2350 Round 1
Sicilian Alapin
1.e4 c5 2.c3 g6 3.d4 cxd4 4.cxd4 d5 5.exd5 Nf6 6.Nc3 Nxd5 7.Qb3!?
More popular in ChessBase's 2021 Mega database is 7.Bc4, but the text scores a better percentage.
7...e6!?
Overwhelmingly the main move is 7...Nb6.
8.Nf3 Bg7 9.Bg5 Qa5 10.Bb5+ Bd7 11.Bxd7+ Nxd7 12.0-0 N7b6 13.Bd2
Eric Lobron (2528) - Tal Baron (2544), Amsterdam Batavia Grolsch 2017, saw 13.Ne4 0-0 14.Nc5 Qb4 15.Qxb4 Nxb4 (½–½, 24 moves).
13...Qa6 14.a4!?
The analysis engines Stockfish13 and Komodo12.1.1 prefer 14.Ne4 with what they reckon is a roughly equal game.
14...0-0 15.Ne5
If 15.a5, then 15...Nxc3 followed by 16...Nd5 or 16...Nc4.
15...Nxc3 16.Bxc3 Rac8 17.Rfe1
Position after 17.Rfe1
*****
*****
*****
*****
*****
17...Bxe5!?
Leaving White with a slightly-bad bishop and Black with an unchallengeable knight-outpost at d5.
18.dxe5 Rfd8 19.Qa3?!
It is not clear what this achieves as White is unlikely to be able to push his b pawn, but perhaps the idea is to try to attack the weak dark squares around the black king with moves such as Qe7 and Re3-f3/h3.
19...Rd3 20.g3?!
This is also puzzling.
20...Nd5 21.Rac1
This natural-looking move is not liked by the engines, who want White to give up a pawn, eg 21.Qb3!? Nxc3 22.bxc3 Rdxc3 23.Qb4, although the resulting position is clearly very good for Black.
21...Nxc3?!
Black is still better after this, but giving up a good knight for a very-limited bishop is not optimal.
22.bxc3 Qc4?!
Black probably has better than this, although the engines cannot agree on what should be played.
23.Red1! Rcd8
Not 23...Rxd1+?! 24.Rxd1 Qxc3?? 25.Rd8+ etc.
24.Rxd3 Qxd3 25.Qb4 Rd7 26.a5 Qd5 27.Re1 Rc7 28.Re4?
The engines reckon White is only very slightly worse after 28.Qd6!? Rxc3 29.Qxd5 exd5 30.Rb1 Rc7 31.f4 despite being a pawn down.
28...Qc5?
Missing the simple 28...Rxc3.
29.Rd4 Kg7 30.f4?
The engines reckon 30.Rd8 effectively equalises.
30...Qxc3 31.Qxc3 Rxc3 32.Rd7 Ra3 33.Rxb7 Rxa5
Black is a pawn up but White's active rook gives good drawing chances.
34.Kg2 Ra2+ 35.Kh3 h5 36.Rb8 a5 37.Ra8!?
The engines strongly dislike this natural-looking move, preferring 37.Rc8, 37.Rd8 or 37.Re8 
37...a4 38.Ra7 a3 39.Ra8 Ra1?
Black is much better, according to the engines, after 39...g5!?, eg 40.fxg5 Kg6 41.g4 Kxg5 42.Rg8+ Kh6 43.gxh5 Re2.
40.Ra5?
White apparently equalises with 40.h4, and if 40...Ra2 then 41.h3.
40...a2?
40...Ra2 would have corrected my previous move.
41.Kh4 Kh6
I offered a draw.
42.Ra7 Rc1 43.Rxa2 Rc7 44.Ra8 Rb7 45.Rd8 Rc7 46.Rd2 Rb7 47.Kh3 g5 48.Rf2 g4+ 49.Kh4 Kg6 50.Ra2 Rb1 51.f5+!? Kxf5 52.Kxh5 Rb4 53.Re2
The game has been dead-equal for many moves, according to the engines.
53...Re4?
Allowing a transposition into a pawn-ending that is easier for White to play. The engines reckon White is winning this ending (Stockfish13) or is at least much better (Komodo12.1.1), but the position is still a tablebase draw.
54.Rxe4 Kxe4 55.Kxg4 Kxe5 56.Kg5
The idea I missed at move 53. It came a s shock to me (but it is not enough for a win).
56...f6+ 57.Kg6
Position after 57.Kg6 - Black has three drawing moves and two that lose
*****
*****
*****
*****
*****
57...f5?
Also losing is 57...Ke4, but the other three king moves, all of which let the e pawn rush forward to queen, draw, according to tablebases.
The game finished:
58.h4 Ke4 59.h5 Kf3 60.h6 Kxg3 61.h7 f4 62.h8=Q f3 63.Qh1 Kf2 64.Kg5 Ke2 65.Kf4 f2 66.Qf3+ Ke1 67.Qe3+ 1-0

Thursday 20 May 2021

Beat The ... Sicilian Rossolimo

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

The Rossolimo Variation of the Sicilian Defence: 1.e4 c5 2.Nf3 Nc6 3.Bb5 is one of the most popular anti-Sicilians.
Black's commonest reply is 3...g6, which scores 43%, but that is comfortably eclipsed by the 46% of 3...e6.
Position after 3...e6
White has five moves that appear at least 100 times in Mega21.

A) 4.0-0 (7,832 games)
After 4...Nge7 the line splits.
A1 5.Re1 Nd4!?, after which the line splits again.
A1.1 6.Nxd4 cxd4, when A1.1a 7.c3 Nc6!? 8.d3 Bc5 scores 50% for Black and A1.1b 7.d3 a6 8.Ba4 Nc6 9.Nd2 b5 10.Bb3 Bb7 scores 67% for Black, albeit from a small sample.
A1.2 6.Bf1 Nxf3+ 7.Qxf3 Nc6 scores 79% for Black, albeit from a small sample.
A2 5.c3 a6, after which the line splits again.
A2.1 6.Ba4 c4!?, when A2.1a 7.d4 cxd3 8.Qxd3 reaches a position in which Black scores 61% with both 8...b5 and 8...Ng6, and A2.1b 7.Bc2 Ng6 8.b3 reaches a position in which Black scores 50% with both 8...cxb3 and 8...d5, albeit from very small samples.
A2.2 6.Bxc6 Nxc6 7.d4 d5, when A2.2a 8.exd5 Qxd5 9.Be3 c4 scores 66% for Black and A2.2b 8.e5 Bd7 scores 80% for Black, albeit from a small sample.
A2.3 6.Be2 d5 7.exd5 exd5 8.d4 cxd4 scores 83% for Black, albeit from a very small sample.
A3 5.b3 Ng6 6.Bb2 Qc7 7.Re1 f6 8.c3 Be7 scores 100% for Black, albeit from a small sample.
A4 5.Nc3 Ng6, after which the line splits again.
A4.1 6.d4 cxd4 7.Nxd4 Qc7 8.Be3 Be7 scores 70% for Black, albeit from a small sample.
A4.2 6.Re1 Be7, when A4.2a 7.d4 cxd4 8.Nxd4 0-0 scores 54% for Black, albeit from a small sample, and A4.2b 7.Bxc6 bxc6 scores 95% for Black, albeit from a small sample.
A4.3 6.Bxc6 bxc6, when A4.3a 7.d3 e5!? 8.Ne2 d5 scores 75% for Black, albeit from a very small sample, and A4.3b 7.e5 f6 scores 68% for Black, albeit from a small sample.
A5. 5.d4 cxd4 6.Nxd4 Qb6, after which the line splits again.
A5.1 7.Nxc6 bxc6 (this and 7...Nxc6 score equally well, but Stockfish13 and Komodo12.1.1 prefer the pawn capture), when A5.1a 8.Bd3 Ng6 9.c4 Bc5 10.Nc3 Qc7!? scores 62% for Black, albeit from a small sample, and A5.1b 8.Be2 d5 scores 50% for Black, albeit from a very small sample.
A5.2 7.Bxc6 Nxc6 8.Nb3 Be7 scores 55% for Black, albeit from a small sample.

B) 4.Bxc6 (4,191 games)
After 4...bxc6 the line splits.
B1 5.d3 Ne7, after which the line splits again.
B1.1 6.Qe2 d5!?, when B1.1a 7.c4 Ng6 8.0-0 Bd6 scores 66% for Black, albeit from a small sample, and B1.1b 7.0-0 Ng6 8.c4 is a transposition to B1.1a.
B1.2 6.0-0 Ng6, when B1.2a 7.Ng5 f6 8.Nh3 Bd6 scores 100% for Black, albeit from a very small sample, B1.2b 7.Nc3 is a transposition to A4.3a, and B1.2c 7.e5 f6 8.exf6 gxf6 scores 70% for Black, albeit from a small sample.
B1.3 6.h4 d6 7.h5 e5 scores 63% for Black, albeit from a small sample.
B2 5.0-0 Ne7, after which the line splits again.
B2.1 6.d3 is a transposition to B1.2.
B2.2 6.b3 Ng6 7.Bb2 f6, when B2.2a 8.e5 Be7 9.d3 0-0 10.Nbd2 Qc7!? scores 83% for Black, albeit from a very small sample, and B2.2b 8.d4 cxd4 9.Nxd4 Qc7 scores 79% for Black, albeit from a small sample.
B3 5.b3 Ne7 6.Bb2 f6!? scores 62% for Black, albeit from a small sample.
B4 5.e5 f6, after which the line splits again.
B4.1 6.0-0 fxe5 7.Nxe5 Nf6 scores 53% for Black.
B4.2 6.d3 fxe5 7.Nxe5 Nf6 scores 68% for Black, albeit from a small sample.

C) 4.c3 (364 games)
After 4...Qb6!? the line splits.
C1 5.Qe2 Nf6 scores 67% for Black, albeit from a small sample.
C2 5.Ba4 Nge7 also scores 67% for Black, albeit from a very small sample.

D) 4.Nc3 (249 games)
After 4...a6!? 5.Bxc6 bxc6 the line splits.
D1 6.0-0 Ne7 scores 100% for Black, albeit from a very small sample.
D2 6.d3 Qc7 also scores 100% for Black, but again from a very small sample.

E) 4.b3 (191 games)
After 4...Nge7 the line splits.
E1 5.Bb2 Ng6!? 6.0-0 is a transposition to A3.
E2 5.0-0!? is also a transposition to A3.

Wednesday 19 May 2021

Beat The ... Tarrasch

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

The Tarrasch Variaton of the Queen's Gambit: 1.d4 d5 2.c4 e6 3.Nc3 c5 is intermittently popular among the elite but always has a dedicated following at club level.
One of its attractions is that Black hopes to avoid defending a long grind, as is often the case in the Exchange Variation of the Queen's Gambit Declined.
White's best response, statistically, is the traditional favourite 4.cxd5, which scores 57%.
Position after 4.cxd5
Black has two major replies.

A) 4...exd5 (8,653 games)
After 5.Nf3 the line splits.
A1 5...Nc6 6.dxc5!? after this surprising capture, which has been played by Magnus Carlsen and is also the choice of Stockfish13 and Komodo12.1.1 (it scores 64%, easily beating the 59% of the well-known 6.g3), the line splits again.
A1.1 6...d4 7.Na4 Bxc5!? 8.Nxc5 Qa5+ 9.Bd2 Qxc5 10.b4!?, when A1.1a 10...Nxb4 11.Rc1 Qd6 12.e3 scores 71% for White and A1.1b 10...Qb6 11.Qa4 scores 72% for White, albeit from a small sample.
A1.2 6...Nf6 7.Be3!? scores 89% for White, albeit from a small sample.
A1.3 6...Bxc5 7.Nxd5!? scores 90% for White, albeit from a small sample.
A2 5...Nf6 6.Bg5!? after this slightly surprising move, which has been played by Karpov and scores 69%, ahead of the 67% of 6.g3, the line splits again.
A2.1 6...Be6 7.e4!?, when A2.1a 7...dxe4 8.Bb5+ scores 100% for White, albeit from a small sample, and A2.1b 7...Be7 8.e5 scores 67% for White, albeit from a very small sample.
A2.2 6...Be7 7.dxc5, when A2.2a 7...Be6 8.e3 0-0 9.Rc1 scores 93% for White, albeit from a small sample, A2.2b 7...0-0 8.e3 Be6 is a transposition to A2.2a, and A2.2c 7...Nc6 8.e3 scores 74% for White.
A2.3 6...cxd4 7.Nxd4, when A2.3a 7...Be7 8.e3 0-0 9.Bd3 Nc6 reaches a position in which White scores 75% with both 10.0-0 and 10.Nxc6, albeit from small samples, and A2.3b 7...Nc6 8.e3 Be7 9.Be2 0-0 10.0-0 Be6 11.Rc1 Rc8 12.Nxc6 scores 70% for White, albeit from a small sample.
A2.4 6...Nc6!? 7.Bxf6, when A2.4a 7...gxf6 8.g3!? scores 83% for White, albeit from a small sample, and A2.4b 7...Qxf6!? 8.Nxd5 Qd8 9.e4 scores 60% for White, albeit from a small sample.
A3 5...Be6 6.e3!?, after which the line splits again.
A3.1 6...Nf6 7.Bb5+ scores 75% for White, albeit from a small sample.
A3.2 6...Nc6 7.Be2 scores 75% for White, albeit from a small sample.

B) 4...cxd4 (3,538 games)
After 5.Qa4+ Bd7 6.Qxd4 exd5 7.Qxd5 the line splits.
B1 7...Nc6 8.e3!? Nf6 9.Qd1 Bc5 10.Nf3 Qe7 11.Be2 0-0-0 12.0-0 g5 13.Qc2 scores 67% for White, albeit from a small sample.
B2 7...Nf6 8.Qb3 Nc6 9.Nf3, after which the line splits again.
B2.1 9...Bc5 10.Bg5, when B2.1a 10...h6 reaches a position in which in which White scores 100% with both 11.Bh4 and 11.Bf4, albeit from very small samples, and B2.1b 10...Qa5?! 11.e3 scores 100% for White, albeit from a very small sample.

Tuesday 18 May 2021

Second-Shot Update

I HAVE had no discernible side-effects, apart from tenderness around the injection site, since my second AstraZeneca shot on Sunday.
This compares with a mildly sore arm after my first shot on February 28.
Perhaps the improvement is not very surprising in that I have read people are more likely to suffer a noticeable reaction after a first AstraZeneca shot, while Pfizer recipients are more likely to suffer after a second shot.
My NHS app - not to be confused with the NHS trace & test app - was updated yesterday to show I have received two doses.
The question is - how soon will countries accept this 'vaccine passport'?

Monday 17 May 2021

Another Devon Bites The Dust

I have just received this:

The DCCA would like to formally announce that there will be no DCCA "Paignton" Congress in Autumn 2021. We are hopeful that the next Congress - which will be the 70th - will be able to take place in Autumn 2022, with further details (including date and venue) being circulated later this year.

The viability of running the 70th Congress in 2022 - which we hope to make a special event, in keeping with the landmark year - is contingent on us forming a full organising committee in the coming months to support the event. 

Chess Scalps (part five)

Spanton (2018) - Petr Marusenko (IM 2342)
Hastings 2005-6
Sicilian Bb5(+)
1.e4 c5 2.Nf3 d6 3.Bb5+ Nd7 4.d4 Nf6 5.Nc3 cxd4 6.Qxd4 h6!?
An unusual move, but one that has been played by grandmasters. Much more popular are 6...e5 and 6...a6.
7.Be3
I strongly considered 7.e5!?, but felt Black is fine after 7...dxe5 8.Nxe5 a6, although Stockfish13 and Komodo12.1.1 like White.
7...e5 8.Qd3 Be7 9.Bc4!? 0-0 10.Nh4?!
Misremembering analysis by GM Murray Chandler. Almost certainly better is 10.0-0.
10...d5! 11.exd5 e4 12.Nxe4??
Tomáš Oral (2510) - Jiří Štoček (2470), Ostrava Cup  (Czechia) 1998, saw 12.Qd4 Bc5 13.Qd2 Bxe3 14.fxe3 Ng4, when Black has good compensation for a pawn (½–½, 49 moves).
12...Nxe4 13.Ng6 Nd6?
Simple and strong is 13...fxg6, although the text is also good enough for a large black advantage.
14.Nxf8
Black to play and lose
*****
*****
*****
*****
*****
14...Ne5?? 15.Qh7+ 1-0
So I have chess scalps in five categories: a draw with an FM, a win against an FM, a draw with an IM, a win against an IM and a draw against a GM. But I still await a scalp in the sixth category - a win against a GM.

Sunday 16 May 2021

Chess Centre Meet-Up

ADAM Raoof has alerted me to the return of something approaching chess normality in at least one part of the country, namely Ilkley, West Yorkshire.
If you do not receive Adam's newsletter, here is the relevant section:

“Dear Friends and Fellow Chess Players,

We are delighted to confirm that, pursuant to the relaxation of COVID-19 restrictions from Monday 17th May, we can finally open our doors to adults players.

Therefore, our weekly Thursday night social chess evenings will commence from this Thursday (20th May), and will be every Thursday 7pm onwards.

We would love to invite anyone wanting to come play some chess in a friendly and relaxed environment to visit us on a Thursday and see the Chess Centre. This is open to players of all ages, so whether you just want to come have a few games, learn from some of our more experienced players or just generally see what's going on then you are most welcome. Due to COVID rules we are limited in numbers at present so please let us know in advance if you can make it.

Also, please note that we are awaiting confirmation of our premises liquor licence from Bradford Council (expected June 2021). Therefore, at the present time we are not serving alcohol, but we do have a range of coffees, hot chocolates and soft drinks available, with full table service.

Full details can be found on our newly launches website: https://www.chesscentre.online

On the website you will also see further details of our upcoming events, including:

  • Our first ever "Mini-Congress" - 22nd - 23rd May

  • Our first ever Open RapidPlay Tournament - 29th May

  • Our first ever Junior RapidPlay Tournament (16 and under) - 5th June

We have a wide range of upcoming events throughout the summer, and we would love to see as many of you as possible at some of these events.

All events registrations are online, but please don't hesitate to get in touch if you have any questions.

Kind Regards

Andy & Matt

Double Shot

HAD my second AstraZeneca shot this morning.
I got to the vax centre almost 30 minutes early, hoping to find somewhere to have an espresso while waiting for my appointment.
But noticing the waiting area was empty, I went up to one of the volunteers and was ushered straight in.
My vaccination card
As a minor aside, I wonder if people's confidence in the AstraZeneca jabs would increase if the staff administering them could write the name correctly?

Chess Scalps (part four)

Spanton (2096) - Viesturs Meijers (IM 2418*)
Lausanne (Switzerland) 1999
Philidor Defence
1.e4 e5 2.d4 exd4 3.Nf3
I remember I wanted to play some kind of Scotch, while avoiding a Latvian or Elephant Gambit, but instead VM transposes into a Philidor.
3...d6 4.Nxd4 Nf6 5.Nc3 Be7 6.Bc4 0-0 7.0-0 Nc6 8.Bf4!?
Not a popular choice but liked by the analysis engines Stockfish13 and Komodo12.1.1.
8...Nxd4 9.Qxd4 Nd7!?
Another far-from-popular choice. One idea, presumably, is to hit the white queen with 10...Bf6, although this is easily countered.
10.Nd5 Bf6 11.Nxf6+ Qxf6 12.Qxf6 Nxf6
The exchanges have taken pressure off the black position, but the white bishop-pair should be worth something.
13.Rfe1 Be6 14.Bb5
Anatoly Ozgibcev (2368) - Alexandr Ponomarenko (2231), Russian Team Blitz Championship 2018, saw 14.Bd3 (Stockfish13's choice) Nd7!? 15.b3 Ne5 16.Bf1 with a slight edge for White (1-0, 61 moves). Komodo12.1.1 prefers the text.
14...a6 15.Ba4 Ng4 16.e5?!
I was fixated on drawing.
16...Nxe5 17.Bxe5 dxe5 18.Rxe5 Rad8 19.Re2!?
This allows a little combination, but worse, according to the engines, is 19.a3?! Rd2.
Position after 19.Re2!?
*****
*****
*****
*****
*****
19...Bxa2
The engines prefer 19...c5 20.c3 g6, giving Black a slight edge.
20.b3 b5 21.f3 bxa4 22.Rxa2 axb3 23.cxb3 Rd6
Black is a pawn up but his weaknesses mean White has equality, according to the engines.
24.Re7 Rc6 25.Rd2 Rc3 26.Rdd7 Rxb3 27.Rxc7 Rb6 28.Ra7 h6
Stockfish13 reckons the position is dead-equal, but Komodo12.1.1's verdict of a slight edge for Black makes more sense to me.
29.Kf2 Rf6 30.h4 Rb8 31.g4 Rb2+ 32.Re2!?
As a general rule, two rooks are better than one when it comes to converting an extra pawn, but that is not so clear here as the white rooks were well-coordinated.
32...Rxe2+ 33.Kxe2 Rb6 34.h5 Kf8 35.Kd3 g6 36.hxg6 fxg6 37.Rh7 g5 38.Kc4 a5 39.Ra7 Rf6 40.Rxa5 Rxf3
The position is a tablebase draw, although older engines, eg Fritz5, tend to reckon Black's advantage is almost winning.
41.Ra6 Rf4+ 42.Kd3 Kg7 43.Ra7+
The only move to maintain the draw, but pretty obvious.
43...Kg6 44.Ra6+ Rf6 45.Rxf6+!?
VM's body language suggested he was surprised at this exchange, but I had correctly calculated the pawn-ending as drawn. According to the Syzygy tablebase, White also draws with 45.Ra1 and 45.Ra8.
45...Kxf6 46.Ke4 Kg6
*****
*****
*****
*****
*****
47.Kf3!
The only drawing move.
47...Kf7
Or 47...h5 48.gxh5+ Kxh5 49.Kg3=
48.Ke3 Ke7 49.Ke4 Ke6 50.Kd4
Again an only-move, but not a difficult one.
50...Kf7 51.Ke3 Kg7 52.Ke4 Kf6
*****
*****
*****
*****
*****
53.Kf3
Komodo12.1.1 incorrectly gives 53.Ke3? as also drawing, but White is lost after 53...Ke5.
53...Ke5 54.Ke3 Kd5 55.Kd3 Ke5 ½–½
More detailed notes to this ending can be found at https://beauchess.blogspot.com/2020/01/when-it-paid-to-trade-down-to-pawn.html
*VM, who is Latvian, became a grandmaster in 2004.

Saturday 15 May 2021

Chess Scalps (part three)

Arnold Denker (GM 2376) - Spanton (2026)
US Open (Kissimmee, Florida) 1997
Colle System
1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 b6 5.0-0 Ba6!?
I had been impressed by this idea after seeing it in a book by Alex Dunne, the correspondence columnist for US magazine Chess Life.
6.Bxa6
Played very quickly by AD.
6...Nxa6 7.c4
Attacking Black's queenside light-square weaknesses.
7...c6!?
Normal is 7...Nb8. The text does not appear in ChessBase's 2021 Mega database, but is quite liked by the analysis engine Stockfish13.
8.Ne5 Qc7 9.Qa4 Nb8 10.Nc3 Bd6 11.f4 0-0 12.Bd2 Ne4 13.Nxe4 dxe4 14.Qc2 f5 15.Bc3 Bxe5 16.fxe5 Nd7 17.Rad1 Rad8
Both sides have mobilised, and now the grandmaster seizes the initiative.
Position after 17...Rad8
*****
*****
*****
*****
*****
18.g4! g6 19.gxf5 gxf5 20.Kh1 Kh8 21.Qf2 Rf7!
A far-from-obvious defensive move that allows the rook to cover three important dark squares: f6, e7 and g7.
22.Rg1 Rg8 23.Rxg8+ Kxg8 24.Rg1+
A logical move and enough for an edge, but probably stronger, as suggested by Stockfish13 and my other main analysis engine, Komodo12.1.1, is 24.d5!?
24...Rg7 25.Rxg7+ Kxg7 26.Qh4 Nf8 27.Qf6+ Kg8
The black position looks vulnerable, but seems to hold in all lines.
28.Kg2 Qf7 29.Qxf7+ Kxf7 30.Bb4 Ng6 31.Kg3 Kg7 32.Bd6 h5 33.h4 Kf7
I offered a draw.
34.a4 a6 35.b3 Ke8 36.Bb4 Kd7 37.Bd6 Kc8 ½–½

Friday 14 May 2021

Chess Scalps (part two)

HERE is my first win against a Fide master.
William Paschall (FM 2330*) - Spanton (2060)
Hastings Challengers 1995-6
QGD Semi-Tarrasch
1.d4 Nf6 2.c4 e6 3.Nf3 d5 4.Nc3 c5 5.cxd5 cxd4!?
Considerably more popular are 5...Nxd5 and 5...exd5, but the text has been played by many top players including Lasker, Gligorić, Keres, Tal, Benko, Kramnik and So.
6.Qxd4 exd5 7.Bg5
Theory considers 7.e4!? to be critical, although the bishop development scores four percentage-points more in ChessBase's 2021 Mega database.
7...Be7 8.e3 Nc6 9.Qd3!?
This somewhat unusual move is the choice of the analysis engines Komodo12.1.1 and Stockfish13. Most popular in Mega21 are 9.Bb5 and 9.Qa4.
9...Be6 10.Be2 0-0 11.0-0 h6 12.Bh4 Qa5!?
This was a novelty back in 1996 (the game was played on New Year's Day), and still does not appear in Mega21. The engines reckon 12...Qb6 equalises.
13.Rfd1 Rfd8 14.Rac1 Nb4!?
More logical-looking is 14...Rac8.
15.Qb1 a6 16.Nd4 Rac8?
The engines reckon 16...Nc6 keeps Black's disadvantage to manageable proportions.
17.Nf5?!
Even stronger seems to be 17.Nxe6 fxe6 18.Qg6, eg 18...Nc6 19.Bg4 or 18...d4!? (the engines' choice) 19.exd4 Qf5 20.Qxf5 exf5 21.a3.
17...Bxf5 18.Qxf5 g6 19.Qb1
White has more chance of a slight edge, according to the engines, by keeping the queen active on the kingside with 19.Qf3 or 19.Qh3.
19...Nc6 20.Bf1 d4!? 21.exd4 Nxd4 22.Re1 Nf5 23.Rxe7 Nxh4 24.Rxb7?
Grabbing a pawn, but now White comes under a strong attack. The engines prefer 24.Ne2!? with equality.
Position after 24.Rxb7?
24...Qg5 25.h3
The engines come to prefer 25.g3 Nf3+ 26.Kg2 but reckon Black has a big advantage after 26...Qc5.
25...Rd2 26.Rd1
White seems to be busted. The best the engines come up with is 36.Nd5 when, for example, 36...Rd8 37.Nxf6+ Qxf6 38.f3 loses to 38...Nxf3+! 39.gxf3 40.Qxf3 etc.
26...Rxc3!?
Not the only way to win.
27.bxc3 Nf3+ 28.Kh1 Qf4 29.g3 Qxg3! 0-1
*Paschall, who is American, became an IM in 1998.

Thursday 13 May 2021

Chess Scalps

MOST players will never gain a master title, whether grand, international or Fide.
But most players can aspire to collecting the scalps of such masters in classical over-the-board games.
I like to think of there being six categories of such scalps - one for drawing with an FM, one for beating an FM, one for drawing with an IM, one for beating an IM, one for drawing with a GM and one for beating a GM.
You can think of yourself as having a full set if you have a scalp in all six categories.
My collection is not quite complete - I have five of the six categories - but I hope to get there one day.
In this series I will be showing how I completed each of the five categories I have.

Spanton (*)  - Claude Adrian (FM 2275)
Lloyds Bank (London) 1993
French Winawer
1.e4 e5 2.d4 d5 3.Nc3 Bb4 4.e5 Ne7
A popular alternative to the most-common continuation, 4...c5, but it was the first time I had faced it.
5.Bd2 c5 6.a3 Bxc3 7.Bxc3
White has won the bishop-pair, but White's unopposed dark-square bishop is not, at least at present, a great piece, and Black with his next move prepares to deprive White of the bishop-pairing.
7...b6 8.Qg4!?
More popular are 8.Bb5+ and 8.b4, but the text is a typical white idea in the Winawer.
8...Ba6?!
Consistent but probably wrong, as is evidenced by the fact that ...Ba6 does not appear in ChessBase's 2021 Mega database. Of the 33 games that reach the position after 8.Qg4!? in Mega21, more than half see 8...0-0.
9.Qxg7 Rg8 10.Qxh7 Bxf1 11.Kxf1
Position after 11.Kxf1
It is common for Black to give up the kingside in the Winawer, but here it is far from clear there is enough compensation.
11...Nbc6 12.Nf3 Qc7 13.Re1 0-0-0!?
Black offers a third pawn to speedily get a position with opposite-side castling, when dynamic factors can easily trump purely material considerations.
14.Qxf7 Qd7 15.h4 Rdf8 16.Qh7 Rg4 17.Qd3 c4 18.Qd1 Nf5
Black's open lines on the kingside mean his attack on the white king is much more advanced than the white attack on Black's king. But three pawns are three pawns.
19.b3?!
This is strongly disliked by my main analysis engines, Stockfish13 and Komodo12.1.1, which reckon 19.Bd2!? gives White a winning advantage. One point is that the reply 19...Ncxd4 can be met by 20.Ng5 (threatening the rook at g4) Rxh4 21.Rxh4 Nxh4 22.Bb4.
19...b5?!
The engines reckon 19...Qg7 20.Rh2 Rg8 21.Ng5 Rxh4 22.Rxh4 Nxh4 gives Black a lot of compensation, and Black would certainly have practical chances.
20.bxc4 bxc4 21.Bb2?
Too defensive. I guess I was wanting to strengthen my centre by playing c3, but it is slow. The engines like occupying the newly opened file with 21.Qb1.
21...Qg7 22.Rh2 Qh6?
Better is 22...Rh8. The text also puts pressure on h4, but gives White a tempo for redeploying the bishop.
23.Bc1 Qh5 24.c3 Kd7
Not 24...Nxh4? 25.Nxh4 Rxh4? 26.Qxh5.
25.Re2?
White had many better moves than this, including the solidifying 25.Bg5.
25...Re4?
Missing 25...Rxg2!, when 26.Kxg2?! Qg4+ 27.Kf1 Qxf3 is very dangerous for White, but even 26.Rxg2 Qxf3 is problematic.
26.Rxe4 dxe4 27.d5!
Perhaps CA missed this.
27...exf3
CA offered a draw.
28.dxc6+ Kxc6 29.Qxf3+
Even stronger, according to the engines, is 29.Qa4+.
29...Qxf3 30.gxf3 Kd5
White remains three pawns up but has many weaknesses, and Black's pieces are much more active.
31.f4 Rb8 32.Ke2 Rb1 33.Kd2 Ke4 34.h5 Rb7
The passed pawn must be stopped.
35.h6?
Now the pawn is doomed. Much better is 35.Ke2, allowing the bishop to protect f4.
35...Rh7 36.Ke2 Rxh6 37.Rxh6 Nxh6 38.f3+ Kd5 39.Be3 a6
White is two pawns up but it is hard to see how progress can be made.
40.Bf2 Nf5 41.Kd2 Kc6 42.a4 Kd5 43.Kc2 Kc6 44.Kb2 Ne7 45.Bg3 ½–½
*I did not have a Fide rating but my BCF grade was 161. My opponent, who is French, became an IM in 2001.

Wednesday 12 May 2021

Beat The ... Panov Attack

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

The Caro-Kann was once thought to be practically refuted by the Panov Attack: 1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4, and this system, aka the Panov-Botvinnik Attack, is still popular, especially at club level.
Easily the most-popular reply, 4...Nf6, is also the most successful statistically, scoring 46%, and after 5.Nc3 we reach a position with almost 35,000 games in Mega21.
Here Black's commonest continuations are 5...e6, which scores 44%, 5...Nc6, which scores 47%, and 5...g6, which scores 49%.
But all those moves are eclipsed, at least statistically by Evgeny Agrest's speciality 5...Be6!?, which scores 53%.
Position after 5...Be6!?
This move surprises many whites, which partly explains why no reply has become dominant. I will look at the six most-frequent continuations, although none is especially popular.

A) 6.c5 (100 games)
After 6...g6 the line splits.
A1 7.Nf3 Bg7, after which the line splits again.
A1.1 8.Bb5+ Bd7!? scores 70% for Black, albeit from a small sample.
A1.2 8.Bd3 0-0 9.0-0 scores 80% for Black, albeit from a small sample.
A2 7.Qb3 Qc8 scores 62% for Black, albeit from a small sample.

B) 6.Nge2 (54 games)
After 6...dxc4 7.Nf4 Bg4 8.f3 Bd7 9.Bxc4 e6 the line splits.
B1 10.d5 e5 11.Nd3 Bd6 scores 50% for Black.
B2 10.0-0 Bd6 scores 60% for Black.

C) 6.Nf3 (32 games)
After 6...dxc4 the line splits.
C1 7.Ne5 Nc6 8.Nxc6 bxc6 9.Qa4 Qd7 10.Bxc4 Bxc4 11.Qxc4 e6, after which the line splits again.
C1.1 12.0-0 Bd6 scores 70% for Black, albeit from a small sample.
C1.2 12.Bf4 Be7 13.0-0 0-0 14.Rac1 Rac8 scores 75% for Black, albeit from a very small sample.
C2 7.Bg5 Nd5 8.Qa4 Nc6 9.Ne5?! (played in eight of the 14 games to reach this position - Stockfish13 and Komodo12.1.1 give 9.Qxc4, but much prefer Black) Nxc3 10.bxc3 Qa5 11.Qxa5 Nxa5 scores 92% for Black, albeit from a small sample.
C3 7.Ng5 Bd5, after which the line splits again.
C3.1 8.Nxd5 Qxd5 9.Be2 e6 10.0-0 Nc6 scores 87% for Black, albeit from a small sample.
C3.2 8.Qa4 Nc6 scores 80% for Black, albeit from a small sample.
C4 7.Qa4+ Nc6 8.Bxc4 Bxc4 9.Qxc4 e6 10.0-0 Be7 scores 64% for Black, albeit from a small sample.
C5 7.Be2 Nc6 8.0-0 g6 scores 80% for Black, albeit from a small sample.
C6 7.Bxc4?? (there are six games with this blunder in Mega21) Bxc4 8.Qa4+ b5 scores 100% for Black, albeit from a small sample.

D) 6.cxd5 (29 games)
After 6...Bxd5!? the line splits.
D1 7.Nf3 Nc6, after which the line splits again.
D1.1 8.Be2 e6 9.0-0 Be7 10.Ne5 0-0 scores 62% for Black, albeit from a small sample.
D1.2 8.Nxd5, when Black scores 83% with both 8...Nxd5 and 8.Qxd5, albeit from small samples.
D1.3 8.Bb5 e6 scores 64% for Black, albeit from a small sample.
D1.4 8.Be3 e6 9.Be2 Be7 10.0-0 0-0 scores 90% for Black, albeit from a small sample.
D2 7.Nxd5 Nxd5 scores 75% for Black, albeit from a very small sample.

E) 6.Bg5 (27 games)
After 6...dxc4 7.Bxf6!? exf6 8.d5 Bf5 9.Bxc4 Bd6 Black scores 75%, albeit from a small sample.

F) 6.Qb3 (18 games)
After 6...Nc6 Black scores 61%, albeit from a small sample (note that 7.Qxb7?! Nxd4 is good for Black).

Tuesday 11 May 2021

Beat The ... Elephant Gambit

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

The Elephant Gambit: 1.e4 e5 2.Nf3 d5!?, also known as the Queen's Pawn Countergambit, can come as a shock to whites who have not seen it before.
The replies 3.exd5 and 3.Nxe5 both score 58%, but the latter is less than half as popular as the former, at least in Mega21, so that is the move I will cover.
Position after 3.Nxe5
Black has two popular replies and two other moves that, while not common in Mega21, could easily be played at club level.

A) 3...Bd6 (619 games)
After 4.d4 dxe4 5.Bf4!? Nf6 6.Bc4 0-0 7.Nc3 White scores 80%, albeit from a small sample (it is unusual for a line not to split long before this but Black has no significant alternatives along the way).

B) 3...dxe4 (200 games)
After 4.Bc4 the line splits.
B1 4...Qg5!? 5.Bxf7+ Ke7 6.d4 Qxg2 7.Rf1, after which the line splits again.
B1.1 7...Bh3 8.Bc4 Nf6 9.Bf4 scores 83% for White, albeit from a small sample.
B1.2 7...Nd7 8.Bc4 scores 75% for White, albeit from a small sample.
B2 4...Nh6 5.d4 scores 100% for White, albeit from a small sample.

C) 3...Qe7 (60 games)
After 4.d4 the line splits.
C1 4...f6 5.Nc3!? fxe5 6.Nxd5 Qf7 (6...Qd6?! 7.Qh5+ g6 8.Qxe5+ Qxe5 9.dxe5 gives White a large advantage, according to Stockfish13 and Komodo12.1.1) 7.Bc4 scores 100% for White, albeit from a small sample.
C2 4...dxe4 is rarely seen, but White scores 100% after both 5.Bc4 and 5...Nc3.

D) 3...Nf6 (19 games)
After 4.d4 the line splits.
D1 4...Nxe4 5.Nd2 Nxd2 6.Bxd2 scores 79% for White, albeit from a small sample.
D2 4...dxe4?! 5.Bc4 Be6 (5...Nd5?! 6.Nxf7! Kxf7 7.Qh5+ etc) 6.Bxe6 fxe6 7.Bg5 scores 83% for White, albeit from a small sample.

Monday 10 May 2021

Beat The ... French Winawer

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

The Winawer Variation of the French Defence: 1.e4 e6 2.d4 d5 3.Nc3 Bb4 is very popular, especially at club level.
White's most-popular reply is 4.e5, which scores 56%. However the lesser-known, but increasingly popular, 4.Qd3 scores 58%.
Position after 4.Qd3
Black has three main replies.

A) 4...dxe4 (541 games)
After 5.Qxe4 Nf6 6.Qh4 the line splits.
A1 6...c5 7.dxc5 Bxc3+ 8.bxc3 Qa5 9.Qb4 scores 60% for White.
A2 6...Nc6 7.Nf3 Nd5 8.Bg5 scores 64% for White.

B) 4...Ne7 (180 games)
5.Ne2 0-0 6.a3 only occurs eight times in Mega21 but scores 63% for White.

C) 4...Nf6 (66 games)
This line is more important than the number of times 4...Nf6 has been played as the position can also arise from the move-order 1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Qd3 Bb4, although in that move-order Black would more-normally play 4...dxe4.
After 5.e5 the line splits.
C1 5...Ne4 6.a3 Bxc3+ 7.bxc3 scores 62% for White.
C2 5...Nfd7 6.Qg3 scores 70% for White.

The 4.Qd3 system is relatively new - only two games in Mega21 pre-date 1985 - and could easily catch an opponent unprepared.
There is little established theory and the main ideas seem less complicated than those of more-mainline Winawers.

Sunday 9 May 2021

Beat The ... New London (part two)

IN this series I am looking at the statistically best way to play against popular opening lines.
The numbers are drawn from the 2021 edition of ChessBase's Mega database, ignoring, where possible, those results that include very few games and so are statistically insignificant.

After 1.d4 Nf6 2.Bf4 Black has five moves each that appears more than 3,000 times in Mega21.
The most successful, statistically, is 2...c5!?, which scores 60%.
Position after 2...c5!?
White has four popular replies.

A) 3.e3 (2,207 games)
After 3...Nd5!? the line splits.
A1 4.Bg3 Qb6, after which the line splits again.
A1.1 5.b3 cxd4 when A1.1a 6.exd4 Nc6 scores 83% for Black, albeit from a small sample, and A1.1b 6.Qxd4 Qxd4 scores 71% for Black, albeit from a small sample (the moves 6...Nb4, 6...e6 and 6...Qa5+ each scores 100% for Black, but each occurs just twice in Mega21).
A1.2 5.Nc3 Nxc3 6.bxc3 e6 scores 94% for Black, albeit from a small sample.
A2 4.Bxb8 Rxb8, after which the line splits again.
A2.1 5.c4 Qa5+ scores 75% for Black, albeit from a small sample.
A2.2 5.c3 reaches a position with just a small number of games - Black scores 75% with 5...e6 and 5...Qb6.
A2.3 5.dxc5 e6 scores 67% for Black, albeit from a very small sample.

B) 3.d5 (572 games)
After 3...g6!? the line splits.
B1 4.Nc3 Bg7 5.e4 0-0 scores 71% for Black, albeit from a small sample.
B2 4.c4 d6 5.Nc3 Bg7 6.e4 0-0, after which the line splits again.
B2.1 7.Be2 reaches a position in which Black scores 56% with 7...e6 and 7...Qa5, but with each move there are just nine examples in Mega21.
B2.2 7.Nf3 Qa5!? scores 58% for Black.

C) 3.c3 (505 games)
After 3...Qb6 the line splits.
C1 4.Qb3 Nc6 5.e3 d5, after which the line splits again.
C1.1 6.Nd2 c4, when C1.1a 7.Qc2 e6 8.Nf3 Be7 9.Be2 0-0 10.0-0 Bd7 scores 87% for Black, albeit from a small sample, and C1.1b 7.Qxb6 axb6 8.a3 b5 9.Rc1 Bf5 scores 63% for Black.
C1.2 6.Qxb6 axb6 7.Bc7!? Bf5 8.Bxb6 cxd4 scores 50% for Black, albeit from a small sample.
C2 4.Qc2 Nc6!? 5.e3 d5, after which the line splits again.
C2.1 6.Nd2 cxd4!? 7.exd4 Bg4 scores 70% for Black, albeit from a small sample.
C2.2 6.Nf3 Bg4 7.Nbd2 e6, when C2.2a 8.h3 Bf5 9.Qb3 c4 10.Qxb6 axb6 scores 81% for Black, C2.2b 8.Bd3 Be7!? scores 100% for Black, albeit from a small sample, and C2.2c 8.Be2 Bf5 scores 83% for Black, albeit from a very small sample.
C3 4.b3 has only been played 16 times in Mega21. Black scores 79% from seven games with 4...e6 but the engines much prefer 4...cxd4!?, which scores 75% from four games.

D) 3.dxc5 (237 games)
After 3...e6!?, which scores 69%, White has tried six moves in Mega21, but none more than five times.