|How should White proceed?|
Monday, 2 August 2021
MY streak of facing juniors over the board ended at seven when I was paired against a grown-up - albeit one born 39 years later than myself - in the Adolf Anderssen Memorial.
Spanton (1808) - Aleksander Kędzierski (1540)
1.e4 c6 2.d4 d5 3.f3 e6 4.Nc3 Bb4 5.Bd2!?
Until now the game had followed the main line of the Fantasy Variation of the Caro-Kann in ChessBase's 2021 Mega database, but here there are more-popular moves in the shape of 5.Bf4, 5.Ne2 and 5.a3.
And here the most-popular move is 5...Ne7, although the text has also been played by grandmasters.
The analysis engine Stockfish14 much prefers 6...Be7. The text achieves a certain amount of simplification, and Komodo12.1.1 is OK with it, but White gets a handy lead in development.
7.Qxd2 Nf6 8.0-0-0 0-0 9.Nh3
The pawn-sac 9.Bd3?! Qxd4 is playable but is not a winner as 10.Nxf6+ is simply met by 10...Qxf6.
9...Nbd7 10.Bd3 e5?!
Opening the position when behind in development is rarely wise.
11.Nxf6+ Qxf6 12.Ng5 g6 13.h4 h6 14.Ne4 Qg7
The engines prefer getting queens off the board with 14...Qf4.
It was almost certainly better to offer up the d pawn and get on with attacking on the kingside with 15.g4 (Stockfish14), or preparing such an attack with 15.Kb1 (Komodo12.1.1).
15...Nxe5 16.Qf4 Nxd3+ 17.Rxd3 Bf5 18.Rhd1 Rae8 19.Rd4 Bxe4 20.Rxe4 Rxe4 21.Qxe4
Komodo12.1.1 prefers, at least for a while, the strange-looking 21.fxe4. Neither move promises White as much as I thought (hoped?) during the game.
I felt (correctly) that 22.Qd4 Qxd4 23.Rxd4 promised White very little, but the text gives up an open file.
The position is more critical, at least for White, than I realised. The engines like 23.Qb4, 23.c3 or 23.a3, but reckon Black has a slight edge.
Missing 23...Qd6, whose strength I only spotted after making my 23rd move. I was more-or-less reconciled to meeting it with 24.Qf4!? Qxf4 25.gxf4, but Black is clearly better.
24.Qe3 Kg7 25.Qxa7?!
Almost certainly a fundamental misunderstanding of the position. I felt clearing a line for my a pawn would help me more than Black's simultaneous clearing of a line for his f pawn, but that is a static evaluation that ignores the dynamics of the position,
Better, according to the engines, is the defensive 26.Qg1, but in my mind I still thought I had an edge.
I had planned 27.Qe7, but got cold feet on seeing 27...Qf4+ 28.Kb1 Qd2?!, but White is probably OK then. However the engines find the almost-certainly superior 28...Qf6 29.Qxf6+ Kxf6, when the black king is much more active than its white counterpart.
Of course the pawn cannot be taken, but also bad is 28.Qa5? as Black has 28...Ra8!
28...Rd4 29.Qe5+ Qxe5 30.Rxe5 Rxh4 31.Rxc5 g5
Black is ahead in the race to queen.
The engines much prefer 32...Re4, but this is a very tricky ending to evaluate.
33.Ke3 Kg6 34.b3?!
The engines do not like this, suggesting instead 34.Rc6+ f6 35.Rc8, but Black seems quite a lot better after the engines' 34...h5.
34...h5 35.a4 h4 36.Rc4 Rf6
Certainly not 36...Rxc4?? 37.bxc4, when the a pawn queens.
Less clear immediately is 36...h3? 37.Rxf4 gxf4+ 38.Kf3 Kf5 39.a5 h2 40.Kg2 Ke4 41.Kxh2 Ke3, but the continuation 42.a6 f3 43.a7 f3 48.a8=Q f1=Q 49.Qa7+ seems strong for White.
This is not liked by the engines, but White seems to be lost whatever is played.
37...h3 38.Ra4 h2 39.Ra1 g4 40.a6 g3 41.a7 g2 42.a8=Q g1=Q+ 43.Kd2
White is getting mated.
43...Qd4+ 44.Kc1 Qxa1+
Missing 44...Rf1#, but it hardly matters.
Sunday, 1 August 2021
IN the second of today's two games in the Adolf Anderssen Memorial, I met a boy registered with Belgium and aged 14 or 15, making my seventh junior in a row in over-the-board competition.
The text is much less well-known than 4...Nc6 and 4...Bg7, but it is popular with some grandmasters. Often the main idea is to prevent a Maróczy Bind, but it also sets a rather obvious trap.
Igor Vanduyfhuys (1605) - Spanton (1808)
Sicilian Accelerated Dragon
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nf6!?
|Position after 4...Nf6!?|
But apparently the trap is not so obvious. However, all is not lost for White for after ...
5...Qa5+ 6.Nc3 Qxe5+ 7.Be2
... White gets an initiative.
7...Nc6 8.Be3 Bg7 9.0-0 0-0 10.Re1 d6?!
The engines prefer withdrawing the queen with 10...Qc7.
11...Bd7 leaves Black slightly better, according to the engines.
Now White gets his pawn back, and still has an initiative.
12...Qg5 13.Rxe7 Nh5?
The analysis engines Stockfish14 and Komodo12.1.1 reckon best is the passive 13...Rd8, but White has a large advantage.
14.Bxg7 Nxg7 15.Re1 Bf5 16.Nd5
Probably even stronger is 16.Qxd6 Bxc2 17.Nd5.
Missing how vulnerable f6 is. The engines give 16...Be6, but with a large advantage for White.
This really was a bad (tactical) day at the office for me. Best, according to the engines, is 17...Rxe1+ 18.Rxe1 Re8 19.Re3 Rxe3, but after 20.Nxe3 Black is losing material.
Or 18...Kh8 19.Rxe6 fxe6 (19...Rxe6? 20.Ne4+ etc) 20.Nxe8+ e5 21.Nxd6! exd4 22.Nf7+ and 23.Nxg5.
TODAY is the double-round day in the Adolf Anderssen Memorial, and this morning I faced a Polish girl aged 15 or 16.
Spanton (1808) - Justyna Łochina (1526)
1.e4 c5 2.Nf3 e6 3.g3 Nc6 4.Bg2 Nf6
This is Black's usual response in ChessBase's 2021 Mega database.
But more popular for White are 5.Qe2 and 5.d3.
5...d6 6.0-0 Be7 7.d3 0-0 8.h3
Preparing Be3 without the possibility of harassment by ...Ng4.
One of the rules-of-thumb put forward by first world correspondence champion Cecil Purdy is to never place a rook behind an unmoved pawn, even if you intend moving the pawn. I suspect there are many cases when this rule should be ignored, and perhaps it suffices to point out in this instance that the text has been played by Alexander Morozevich, Loek van Wely, other strong grandmasters ... and a 13-year-old Garry Kasparov.
9.a4 a6 10.Be3 Qc7
There seems no good reason to avoid the consistent 10...b5.
Preparing like this the pawn-thrust f4 seems to be a novelty, and it is not liked by the analysis engines Stockfish14 and Komodo12.1.1. Known moves are 11.Nh2!?, 11.Qd2 and 11.Kh2.
11...Rd8 12.f4 d5
Countering in the centre is the classic way to meet a flank attack. Black is better, according to the engines.
The engines give 13...d4!?, which I thought was impossible because of 14.exf6, overlooking that Black can simply reply 14...Bxf6, when 15.Bd2 dxc3 16.Bxc3 Nd4 is better for Black, according to the engines.
Not 14.d4?? cxd4 15.Bxd4 as 15...Nxd4 cannot be met with 16.Qxd4? as Black wins the white queen with 16...Bc5.
14...Bxh4 15.gxh4 seems nothing special for Black, who is left with a bad bishop; White's kingside weaknesses are not easy for Black to get at, and White may develop decent kingside attacking chances.
15.exf6 Nxf6 16.Re1 Nd4 17.Nf3
The engines like 17.a5!? Nc6 18.Nf3, and if 18...Nxa5 then 19.Ng5, when White wins back his pawn and can claim an edge.
17...Nxf3+ 18.Qxf3 b6
The engines prefer this to the more loosening 18...b5 19.axb5 axb5.
Weakening control over the e4 square but fixing the e6 weakness.
|How should Black proceed?|
19...cxd4?!JŁ goes for liquidation, which is probably the wrong thing to do as White will be left with the superior pawn-structure and the better bishop.
20.Bxd4 Bc5 21.Qe3 Bxd4 22.Qxd4 Qc5 23.Rad1 Kf7!?
Queens are going to come off, so JŁ decides her king should come nearer the centre and, while doing so, it can support the backward e pawn.
24.Kh2 Qxd4 25.Rxd4
This queenless middlegame looks good for White, who has the better pawn-structure, the better bishop and the generally more-active pieces. Even so, the engines reckon it is only a slight edge.
Heading for e5, or g5.
26...Rdc8 27.c3 Bc6 28.Nd2 b5?!
Consistent with her last move, but ...
... leaves the black bishop looking very bad.
The engines suggest the surprising 29...b4!?, when 30.Rxb4 Rxb4 31.cxb4 Bd7!? 32.Nf3 is good for White, but Black has counterchances thanks to her protected passed pawn and active rook. Indeed the engines reckon White should reply 30.c4!? with continuing pressure on the black centre.
30.Nf3 Nd7 31.h4
Freeing the h3 square for the white bishop.
31...Re8 looks natural, but 32.Bh3 is very good for White, according to the engines.
Necessary is 32...Re8, but White remains on top, eg 33.g4 Nc5 34.g5 Nb3 35.Rdd1 Nxa5? (the engines give 35...b4 but believe White still has a big advantage) runs into 36.Ra1 and 37.Rxa6.
More natural is 33...Kf6.
This looks like a blunder, but it is Komodo12.1.1's choice for a while. The fact is JŁ's position has gone downhill since she took the decision to liquidate at move 19.
35.Nc6+ Kf6 36.Nxb8 Rxb8 37.Rf4?!
Almost certainly better is 37.Rf1, not worrying about a later ...Nb3, and if, as in the game, 37...g6, then 38.h5.
Not 38.h5?? Nd3.
I rejected the engines' choice, 38.Ref1, because I did not want to give up the open file, but 38...Re8?! 39.h5 Re2+ 40.R4f2 Rxf2+ 41.Rxf2 keeps White well on top.
Better is 38...h5 to stabilise the kingside, when White is better but the game very much continues.
39.g4 Nxa5 40.gxf5 gxf5
Or 40...g5 41.Re6+ Kf7 42.Rf2 etc.
41.Rxf5+ Kg6 42.Re6+
The engines point out two mates-in-four, one starting with 42.Re2+ and the other with 42.Re7+, but the text wins easily enough.
42...Kg7 43.Rxa6 Nc4
The game finished:
44.Ra7+ Kg8 45.Rh5 Bb7 46.Rxh6 Kg7 47.Rh5 Kg6 48.Rg5+ Kf6 49.Rf5+ Kg6 50.h5+ Kh6 51.Rf7 Bc8 52.Rf8 Nb6 53.Bxc8 Rxc8 54.Rf6+ Kxh5 55.Rxb6 1-0
|Black has just played 30...Bb7-e4 and offered a draw in Spanton (1814) - Maximilian Ponomarev (2017), Basel Round 4. Was I right to accept his offer?|
The white knight is hanging, so 31.fxe4?! Qxe5 32.exf5 Qxe3+ 33.Rxe3 exf5 is nothing for White after, say, 34.Rc3 Rd5. But note Black must avoid 31...Rxe4?? 32.Qc3.
Saving the knight with 31.Nd3!? allows 31...Bxd3 32.Qxe6+ Kg7 33.Qe8! (better than immediately capturing on d3) Be4! 34.fxe4 f4! 35.Rf1 Rd8 36.Qe6 fxg3 37.Qf6+ Kg8 38.h4, a complicated line that the engines reckon is equal. Note that in this variation 33...Qd8, which the engines give at first, runs into 34.Re7+ Kh6 35.Qf7 Qh8 36.cxd3 Rxd3 37.Kf1, when the better-coordinated white pieces are enough for an advantage, according to the engines.
However, White seems to be much better after 31.c3. I am fairly sure I rejected it because of 31...Ra4?, but the engines point out that 32.Rd1 allows a winning exploitation of Black's weak king (33.Rd7 cannot be prevented without large material loss). The engines reckon the best reply to 31.c3 is 31...Rd5, but then 32.Nxg6 hxg6 33.fxe4 leaves White much better, eg 33...Re5 34.Qh6 Qg7 35.Qxg7+ Kxg7 36.Kf2 fxe4 37.Ke3.
Saturday, 31 July 2021
FLEW to Poland today for the Adolf Anderssen Memorial, which is being held in his birth-place of Wrocław (more commonly known as Breslau in the 19th century).
Ryanair flies direct to the Silesian city, which is always something of a good-news/bad-news situation.
The bad news today was that Ryanair told me to be at Stansted airport three-and-a-half hours before my flight was due to take off at 11:40, but checkin opened at gone 08:40.
The airline also gave me wrong information about which printed documents I needed, but otherwise checkin was fairly smooth, and passing through security even smoother.
The chess festival consists of a rapid (10 minutes with a five-second increment) today, which does not interest me, a blitz later in the week, and four "Open" tournaments - actually an over-2000, an under-2001, an under-1601 and a beginners', for children aged under 12.
I am seeded eighth of 68 in the under-2001, officially called Wrocław B, which, according to the tournament website, has an average rating of 1571, an average age of 27, nine females, 63 rated players and five federations. A slight oddity is that two of the unrated players have estimated ratings of 2000 which, if even approximately correct, would push my real ranking down to 10th.
Two players, including myself, were born in 1957 - everyone else in 1959 or later. Since I was born in February, I am probably the oldest player in the tournament.
Thirty-two players were born in 2000 or later, including two in 2010, so there is a good chance I may extend the junior fest I 'enjoyed' earlier this month in Basel.
The over-2000 and under 2001 have two rounds tomorrow, followed by one round a day for seven days. The time limit is 40 moves in 90 minutes, a 15-minute 'windback' and a 30-second increment from move one.
Polish customs took about 45 minutes to get through, but standing in the queue made it seem longer - perhaps I have been spoilt by recent smooth border-crossings.
|Wrocław is situated on the Oder and its tributary the Oława (above)|
|The rapid in progress this afternoon at the Sports Club AZS Wratislavia|
|The refreshment tent|
|Paddlesteamer on the Oder at the bottom of the sports club's grounds|
|White has just played 14.Bf4-d2 in Jeremias Stark (1599) - Spanton (1814), Basel Round 3|
Black has to move the queen, but to where? The candidate moves, in my opinion, are 14...Qh5, 14...Qe5, 14...Qc7 and 14...Qa6, and here I want to deal with them in turn.
This comes to be, at least for a while, the second choice of the analysis engines Stockfish14 and Komodo12.1.1. The queen is actively placed on h5. But the activity of the black pieces, including a rook that will soon come to the half-open b file, is easier to direct against the white queenside, so it makes sense to place the black queen where it can add to this pressure.
A queen in the middle of the board in the middlegame is powerful if it cannot be molested, but is often subject to harassment. The latter looks more likely here, but it is not clear such harassment is beneficial to White. For example, 15.f4?? is a gross blunder thanks to 15...Qd4+, while 15.Bc3 Qf4 (or 15...Qh5) 16.Bxg7 Kxg7 leaves Black without his fianchettoed bishop but with the better remaining-bishop and with exploitable dark-square weaknesses in the white position. The engines reckon the best answer to 14...Qe5 is 15.Kh1, with what they rate as approximate equality.
This is what I played. The queen remains on the queenside but is not very active. The engines reckon Black has, at best, a slight edge.
General principles in chess are often a useful guide to finding good moves, but many times two general principles clash. Here 14...Qa6 runs contrary to the general principle of not placing the queen in the firing line of a weaker piece, but conforms with the general principle of pinning enemy men (in this case the c4 pawn). I rejected 14...Qa6 because of 15.Qc2?!, missing that 15...Nb6 costs White a pawn, eg 16.Nxb6 Qxb6+ and 17...Qxb2. Note that in this line 16.c5? loses to 16...Qxa4 17.cxb6 Qxc2 18.Bxc2 Bxb2 19.Rab1 Bd4+. The engines reckon the best answer to 14...Qa6 is 15.b3!?, when Black can probably accept the exchange, but is also much better, according to the engines, after the simple 15...Nb6.
In conclusion, it seems 14...Qa6 is best.
Friday, 30 July 2021
|White is two pawns down in Spanton (1814) - Anuar Tureshbayev (-), Basel Round 2, but can conjure up a strong attack by finding the right move|
Black gets a large advantage after this with the game continuation 24...Bf5.
White needed to find 24.Bd3!, according to the analysis engines Stockfish14 and Komodo12.1.1. The first point is 24...exd4? allows a draw by 25.Bh7+ Kh8 26.Bg6+ Kg8 27.Bh7+ etc. Also drawing is 24...g6? 25.Qh6 exd4 (only move) 26.Bxg6 fxg6 27.Qxg6+ etc. That leaves 24...Rfe8, when 25.Bc4! Be6 26.Bxe6 fxe6 27.R4d3 is only slightly better for Black, according to the engines.
|White is still two pawns down, but can equalise by finding the right move|
The game saw the hopeless retreat 28.Nf3?, but the engines point out 28.Bxf7+! Bxf7 28.Qh8+ Kxh8 29.Nxf7+ and 30.Nxd8.
Thursday, 29 July 2021
|White has just played 15.Bg6-e4 in Cyrill Speiser (1320) - Spanton (1814), Basel Round 1|
White is a pawn up and winning, according to Stockfish14; has a slight edge, according to Komodo12.1.1.
The fact that two strong engines can disagree to such a large extent suggests this is not an easy middlegame to judge.
However, given enough time, Stockfish14 comes to reduce White's advantage to 'only' having the upper hand (+1.28 pawns, compared with Komodo12.1.1's +0.52 pawns).
Grandmaster Nigel Davies states that the No1 factor in chess is king safety, and here White is much better situated. White's pieces are also more active.
The number of pawn-islands is equal, but clearly Black's pawn-structure is superior, which could be a significant factor in any ending.
So what should Black play?
This is Komodo12.1.1's choice, but Stockfish14 may well be right in preferring 15...Qg7!? After 16.Qxg7+ Kxg7 both engines reckon White has the upper hand, but Black has time to complete development and try to start manoeuvring against the white queenside pawns. Of course swopping queens reduces the chance of a tactical coup, but, for reasons explained above, the tactical situation in the diagram favours White (there was also the psychological consideration that CS is a junior, and it rarely hurts to get queens off against juniors). However, White is not forced to exchange queens - 16.Qe3 keeps an edge.
Trying for a kingside attack, but the attempt is unrealistic. Again ...Qg7 is sensible.
Forced, thanks to the threat of Ng6+, so 16...Kh8? has wasted two tempi in what was already a bad position.
|Can White avoid an exchange of queens by retreating to h5?|
Yes, as 19...Bg4? fails to 20.Bxh7+!
However, Black can play 19...Qf7, when either queens come off or the white queen has to retreat from the vicinity of the black king.
|Can Black grab the undefended a2 pawn?|
20...Bxa2?! runs into 21.Nf5+, when 21...Kg8? 22.Rd7 is the start of a massacre, so Black has to play 21...Kh8, when 22.Nd6 is unpleasant, the engines continuing 22...Na5!? 23.Ra1 Nc4 24.fxe5 fxe5 25.Nf7+!? Kg7 26.Bxb7 - a sharp line, but all variations seem good for White.
The engines reckon the best move in the diagram is 20...Rad8, when they agree White is better but disagree as to how to proceed.
Wednesday, 28 July 2021
ONLY one of my five games at Basel reached an ending, but it is quite instructive.
|Position after a black capture on c4 in Jeremias Stark (1599) - Spanton (1814)|
Black is a pawn up and has two protected passed pawns. But this is a rook-and-pawn ending, and White has compensation in the shape of an outside passed pawn, a juicy square at e6 and targets at e7 and h6. Nevertheless the analysis engines Stockfish14 and Komodo12.1.1 reckon Black is clearly winning.
It makes sense to get the passed pawn off the same rank as the king, and the move gives White the option of following up with Ra2 or Re6. Note that 45.Kg3?! Rc3+ 46.Kg2 Ra3 helps Black.
A radical and unnecessarily complicating way of dealing with the weakness at h6. Black is still much better after the text, but simpler is 45...d5, not least because 46.Re6 d4 starts a race White cannot win.
White gets drawing chances after this. The engines reckon 46...Rh4 leaves Black in control.
Much better is 47.h6 Rh4 48.Re6, although after 48...c4 Black has the upper hand, according to Stockfish14; just a slight edge, according to Komodo12.1.1. Note that 47...Rxf5? runs into 48.h7 Kg7 49.Rxe6+, although Black has a draw after 49...Kh8, according to the engines.
More clear-cut is 47...Rh4.
Stockfish14 reckons this is best, but it seems pointless in that the pawn-ending is winning for Black despite White having passed pawns on the a and h files. Komodo12.1.1 wavers between the text and 48.h6, although the latter is easily met by 48...Rf6.
Perhaps the best practical chance is 48.a4!?, when the engines reckon only 48...d5 maintains Black's advantage.
The pawn-ending is lost, so White should have tried pushing the a pawn or relocating the rook.
This comes to be Komodo12.1.1's choice, at least for a while, but it makes Black's task easy. A better practical try is 50.a4, but 50...c4 wins after 51.a5 c3 52.a6 c2 53.a7 c1=Q 54.a8=Q Qg1+ 55.Qg2 (55.Kf3 Qh1+) Qxg2 56.Kxg2 and after 51.Kf2 c3 52.Ke2 f4 53.a5 (what else?) f3+ etc.
50...gxh4+ 51.Kxh4 c4 (1-0, 58 moves).
Tuesday, 27 July 2021
IN round five I had black against Steve Heron from Liechtenstein, who is 11 or 12 and has a Fide rating of 1580.
It might be thought White is doing well as Black's fianchettoed bishop has been exchanged, but White has few prospects of exploiting the holes around the black king. Meanwhile Black has more space in the centre and the more-active bishop. The engines agree Black is better.
Sicilian Accelerated Dragon
1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Nxd4 Nc6 5.Nb3!?
There are eight more-popular moves in ChessBase's 2021 Mega database, but the text has been played by grandmasters. The idea, normally, is to inhibit Black's thematic Accelerated pawn-thrust ...d5.
But this puts ...d5 back on the agenda. Normal is 6.Be2, while 6.Nc3 allows the controversial 6...Bxc3+!?
6...Nf6 7.Nc3 d5
There are 50 examples of 7...0-0 in Mega21 and just five of 7...d5, but Kjetil A Lie (2493), the strongest player to reach the position after 7.Nc3, chose the thematic text.
Komodo12.1.1 narrowly prefers 8...d4, but Stockfish13 likes castling.
Both engines like 9...d4.
10.Bg5 d4 11.Nd5 Be6 12.Nxf6+ Bxf6 13.Bxf6 Qxf6
|Position after Black has connected rooks|