4NCL Final Weekend: Game Two (Result)

I WON in 68 moves with black, but it seems to be one of those games that engines have problems assessing correctly.

There was also an incident where perhaps the players should have consulted an arbiter, and there may be a !! move.

Unfortunately I do not have time to look at the game properly tonight, but I hope to do better tomorrow.

4NCL Final Weekend: Game One

Spanton (1944 ECF/1806 Fide) - Bruno Alexandre dos Santos Silva (1772 ECF/1631 Fide)

Division Four: Wessex Some Stars A v MASCeteers

Board Six (of six)

Sicilian Bb5(+)

1.e4 c5 2.Nf3 d6 3.Bb5+ Bd7 4.Bxd7+ Qxd7 5.0-0 Nc6 6.c3 Nf6 7.Qe2 e5!?

The same move as I faced earlier this month at Fagernes, Norway, in round seven.

8.d4!?

Most popular in ChessBase's 2023 Mega database is 8.Rd1, but the aggressive reply 8...Qg4!? seems to produce an unclear position.

8...cxd4 9.cxd4 exd4 10.Rd1

How should Black continue?

*****

*****

*****

*****

*****

10...Be7

My Norwegian game saw the trappy 10...Rc8, when 11.Nxd4?? would have lost to 11...Nxd4 12.Rxd4 Rxc1+. Instead I replied 11.Na3, which may be enough for a tiny edge, but Stockfish15.1 and Komodo14.1 reckon 11.Bg5 gives White much the better game. They suggest meeting 10.Rd1 with 10...Qg4!?, when Sergey Zhurov (2240) - Piotr Dukaczewski (2220), Warsaw 1993, continued 11.Nbd2 0-0-0?! 12.h3 with what the engines reckon is a strong position for White (1-0, 49 moves). However they give 11...Be7 as a major improvement.

11.Nxd4 0-0 12.Nc3

Also interesting is 12.Nf5!?

12...Nxd4

Mustafa Yilmaz (2626) - Aydın Süleymanlı (2541), Chess.com Blitz 2021, went 12...Rfe8 13.Nf5 Bf8, after which the engines reckon 14.f3 would have left White well on top (Yilmaz played 14.Qf3 and won quickly anyway).

13.Rxd4 Qg4!?

This may be a novelty. Black has tried five other moves in Mega23, but none seems to solve the problem of Black's weak d pawn.

14.Qd3 a6??

Rook moves are preferable, but leave White with at least the upper hand, according to the engines.

15.e5 Qe6 16.exf6

There is no back-rank mate (1-0, 60 moves).

Wessex won the match 4-2.

Kingmakers

AM in Warwick for the last weekend of the Four Nations Chess League's main team event.

As per usual for the final weekend, it features three rounds - today at 2pm, tomorrow at 1pm and Monday at 11am.

I am due to play bottom board (of six) for the first team of Wessex Some Stars in division four.

As I understand it, the format of the 4NCL for 2023-4 is somewhat up in the air, with major decisions to be made, but Wessex seem to have a decent chance of crowning the current season by winning promotion of some sort if all goes well this weekend.

Current standings in division four can be seen here.

Central London League

PLAYED on board four (of five) for Battersea against HMC in division one last night.

Metiejus Gudenas (2078) - Spanton (1944)

Spanish Berlin

1.e4 e5 2.Nf3 Nc6 3.Bb5 Nf6 4.d3 Bc5 5.0-0 Nd4 6.Bc4!?

The mainline in ChessBase's 2023 Mega database runs 6.Nxd4 Bxd4 7.c3 Bb6 with an equal position, according to Stockfish15.1 and Komodo14.1.

How should Black respond?

*****

*****

*****

*****

*****

6...Nxf3+!?

This looks strange, certainly at first glance, as Black moves the queen's knight for the third time in six moves, and helps develop the white queen. However the queen can prove somewhat awkwardly placed in front of the f2 pawn.

7.Qxf3 d6 8.Bg5 h6!? 9.Bxf6

Not 9.Bh4?? g5 etc.

9...Qxf6 10.Qxf6 gxf6

How would you assess the position?

*****

*****

*****

*****

*****

Black's kingside is smashed, but there is little chance of a white mating attack. Meanwhile Black has the bishop-pair. Stockfish15.1 gives Black a slight edge; Komodo14.1 calls the game equal.

11.Kh1 Be6!?

Offering up the bishop-pair in return for covering the important d5 and f5 light squares.

12.Bxe6 fxe6 13.f4 exf4 14.Rxf4 Ke7 15.Nc3 Rhg8

Also playable is 15...Bd4 16.Ne2 Bxb2!? 17.Rb1 Ba3 18.Rxb7 Rab8! as 19.Rxc7+? runs into 19...Kd8 with ...Rb1+ to follow, eg 20.Rc4 Rb1+ 21.Ng1 Bc5 22.d4 e5 23.Rxf6 with a sharp position that the engines reckon is much better for Black after both 23...exd4 and 23...Bxd4!?

16.Raf1 Rg6 17.g3 Rag8 18.Kg2 c6 19.Ne2 Bb6 20.R4f3 h5 21.h4 Bc7 22.d4?!

The engines reckon White has a slight edge after 22.c4.

How should Black proceed?

*****

*****

*****

*****

*****

22...R8g7?

I rejected the correct 22...d5 because of 23.e5 fxe5 (24...f5? 25.Nf4 is good for White) 24.Rf7+ Kd8 25.Rh7?, missing the simple 25...R6g7, eg 26.Rxh5 exd4 27.Rf3 e5, after which Black is a pawn up and has an initiative. The engines reckon 25.dxe5 is an improvement, but still prefer Black after 25...Bxe5.

23.Nf4 Rh6 24.d5 cxd5?!

Better, according to the engines, is 24...exd5 25.exd5 Rg8, but they still reckon White is winning.

25.exd5 exd5 26.Nxd5+ Kd8 27.Rxf6 Rh8 28.Rf8+ Rxf8 29.Rxf8+ Kd7 30.Nxc7!? Kxc7 31.Rh8 Re7 32.Rxh5 Re2+ 33.Kf3 Rxc2 34.Rb5 a6 35.Rb3

There is no good response to the threat of Rc3+.

1-0

HMC won the match 3-2.

Beat The ... e6 Sicilian

I WROTE, but did not post, this article during the covid panic.

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 Sicilian with 2...e6: 1.e4 c5 2.Nf3 e6 is sometimes called the safest Sicilian.

White's most popular reply, 3.d4, scores only 50%, as does the second-most popular reply, 3.c3.

Most successful, statistically, is 3.g3!?, which scores 54%.

Position after 3.g3!?

There are four black moves appearing more than 200 times in Mega21.

A) 3...Nc6 (3,607 games)

After 4.Bg2 the line splits.

A1 4...Nf6 5.Qe2, after which the line splits again.

A1.1 5...d5 6.d3 Be7 7.0-0, when A1.1a 7...0-0 8.e5 Nd7 9.c4!? scores 66% for White, A1.1b 7...b6 8.e5 Nd7 9.c4!? scores 63% for White, and A1.1c 7...b5 8.exd5 exd5 9.d4!? scores 63% for White, albeit from a small sample.

A1.2 5...d6 6.c3!? Be7 7.0-0, when A1.2a 7...0-0 8.d4 cxd4 9.cxd4 d5 10. e5 Ne4 11.Be3!? scores 68% for White, and A1.2b 7...e5!? 8.d3 0-0 9.Nbd2 scores 74% for White, albeit from a small sample.

A1.3 5...e5!? 6.d3!? Be7 7.c3 0-0 8.0-0, when A1.3a 8...Re8 9.Na3 h6 10.Nc4 Bf8!? 11.Re1 scores 64% for White, albeit from a small sample, A1.3b 8...h6 9.Na3 Re8 is a transposition to A1.3a, and A1.3c 8...d6 is a transposition to A1.2b.

A1.4 5...Be7 6.d3, when A1.4a 6...0-0 7.0-0 d5 is a transposition to A1.1a, and A1.4b 6...d5 is a transposition to A1.1.

A2 4...d5 5.exd5 exd5 6.0-0 Nf6 7.Re1+ Be7 8.d4 scores 67% for White, albeit from a small sample.

A3 4...d6 5.d4!? cxd4 6.Nxd4, after which the line splits again.

A3.1 6...Bd7 7.0-0 Nf6 8.c4!? scores 59% for White, albeit from a small sample.

A3.2 6...a6 7.0-0 Bd7 8.Nc3, when A3.2a 8...Nf6 9.Nde2!? Be7 10.h3 scores 58% for White, and A3.2b 8...Be7 9.Nb3!? scores 62% for White, albeit from a small sample.

A4 4...g6?! (there are 348 examples of this move in Mega21, but it is probably a mistake) 5.d4 cxd4 6.Nxd4, after which the line splits again.

A4.1 6...Bg7 7.Nb5 scores 87% for White.

A4.2 6...a6 7.Nxc6!?, when A4.2a 7...dxc6 8.Qxd8+ Kxd8 9.e5 scores 80% for White, albeit from a small sample, and A4.2b 7...bxc6 8.0-0 Bg7 9.Na3!? scores 100% for White, albeit from a very small sample.

A4.3 6...Qb6 7.Nb3 Bg7 8.Nc3 scores 100% for White, albeit from a small sample.

B) 3...d5 (828 games)

After 4.exd5 exd5 5.d4 the line splits.

B1 5...Nc6 6.Bg2, after which the line splits again.

B1.1 6...Bg4 7.0-0!?, when B1.1a 7...cxd4 8.Re1+ Be7 9.Nbd2 Nf6 10.Nb3 scores 79% for White, albeit from a small sample, and B1.1b 7...Nxd4 8.Qe1+ Be7 9.Nxd4 cxd4 10.Qe5 Nf6 11.Qxd4 scores 57% for White.

B1.2 6...Nf6 7.0-0, when B1.2a 7...Be7 8.Be3 scores 66% for White, albeit from a small sample, and B1.2b 7...cxd4 8.Nxd4 Be7 9.c4 0-0 10.Nc3 scores 70% for White, albeit from a small sample.

B2 5...Nf6 6.Bg2 Be7 7.dxc5!?, after which the line splits again.

B2.1 7...0-0 8.0-0 Bxc5 9.Nc3, when B2.1a 9...Nc6 10.a3!? scores 65% for White, and B2.1b 9...h6 10.a3!? scores 85% for White, albeit from a small sample.

B2.2 7...Bxc5 8.0-0 0-0 is a transposition to B2.1.

C) 3...b6 (263 games)

After 4.Bg2 Bb7 5.d4!? the line splits.

C1 5...Bxe4 6.Nc3 Bb7 7.d5 scores 83% for White, albeit from a very small sample.

C2 5...cxd4 6.Nxd4 Nf6 7.Nc3 scores 64% for White, albeit from a small sample.

D) 3...a6 (221 games)

After 4.Bg2 the line splits.

D1 4...b5 5.d4!? cxd4 6.Nxd4 Bb7 7.0-0, after which the line splits again.

D1.1 7...Qc7 8.a4 b4 9.c3 scores 80% for White, albeit from a small sample.

D1.2 7...Nf6 8.Re1, when D1.2a 8...Qc7 9.Nc3 d6 10.a4 bxa4 11.Nd5!? exd5 12. exd5+ Kd8 13.Bg5 scores 82% for White, and D1.2b 8...d6 9.a4 b4 10.c3 scores 83% for White, albeit from a small sample.

D2 4...Nc6 5.d4!? cxd4 6.Nxd4, after which the line splits again.

D2.1 6...Qc7 7.Be3!? Nf6 8.0-0 Be7 9.c4!? scores 65% for White, albeit from a small sample.

D2.2 6...Nf6 7.0-0 Qc7 8.Be3!? is a transposition to D2.1.

Corresponding

I PLAY correspondence chess, mainly for various England teams, at the Fide-affiliated International Correspondence Chess Federation.

Engines mean the vast majority of games are drawn, and I do not post them here.

Occasionally a game escapes that fate, or has some other point of special interest.

Ion Rotaru (2279) - Spanton (2278)

Romania v England 2023

Board Nine (of 24)

QGD Exchange

1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 exd5 5.Bg5 c6 6.e3

How should Black proceed?

*****

*****

*****

*****

*****
6...Bd6!?

Overwhelmingly most popular in ChessBase's 2023 Mega database is breaking the pin with 6...Be7, while Nigel Short has been fond of 6...Bf5!? However Magnus Carlsen and Vladimir Kramnik have tried the text.

7.Bd3 0-0 8.Nf3 h6 9.Bh4 Be6 10.Ne5!?

Stockfish15.1 likes this, but Komodo14.1 prefers the more-obvious 10.0-0. Both engines also like 10.Qb3!?

10...c5!?

Taking advantage of the instability of the white king's knight.

11.0-0 Nc6

Alexey Sarana (2688) - Aram Hakobyan (2612), Chess.com Rapid 2022, saw 11...cxd4 12.exd4 Nc6 13.Nxc6 bxc6 14.Na4 with a slight edge for White, according to the engines (1-0, 58 moves).

12.Nxc6 bxc6 13.dxc5 Bxc5 14.Na4 Bd6 15.Rc1 Rc8 16.Nc5 Qb6 17.Nxe6!? fxe6

White has won the bishop-pair and created holes around the black king, but has strengthened Black's centre

*****

*****

*****

*****

*****

18.Qe2 Qc7

The engines, at least at first, prefer 18...Kh8, but the text is also reasonable.

How should White respond?

*****

*****

*****

*****

*****

19.g3??

This would be bad enough over the board, but is horrendous at correspondence. It cannot be blamed on a fingerslip as the procedure at the ICCF is that first you input your move on a board at the website, you then have to scroll down to hit SUBMIT, at which point you are shown the board again and then you have to scroll down to hit COMMIT. In other words there are three chances to avoid a blunder.

Komodo14.1 reckons 19.Bg3 Bxg3 20.hxg3 may give White a slight edge, while Stockfish15.1 suggests 19.Bxf6!? Rxf6 20.g3, claiming White definitely has a slight edge.

19...g5 20.Bxg5 hxg5

White has precious little compensation for a knight.

21.e4 dxe4 22.Bxe4 Nxe4 23.Qxe4 Qf7!? 24.Qe3

The engines prefer this over 24.Rxc6 Rxc6 25.Qxc6 Be5 26.b4 Bd4.

24...Qf5 25.Rc4 c5 26.b3 Rcd8 27.Kg2 Kf7 28.Re1 Rh8 29.Kg1 Be7 30.Ra4 Rd7 31.h4 Bf6 32.Ra6 Re7 33.hxg5 Bd4 34.Qf4 Kg6 35.Qxf5+ Kxf5 36.Kg2 Reh7!? 37.Kf3

Or 37.Raxe6 Rh2+ etc.

37...e5 0-1

Team standings can be seen here.

How Drawish Are Opposite-Coloured Bishops (part four)?

HERE is another opposite-coloured bishops ending I did not cover in Opposite Exceptions.

Black has just captured on d2 in Spanton (164 BCF) - A Trifunovic (147 BCF), Middlesex Teams 1993

*****

*****

*****

*****

*****

*****

*****

*****

*****

*****

Black's extra two pawns are enough to win without any particular problem.

33.Bc2 Bc1 34.b3 Kf8 35.a4

Not 35.Bxh7? g6 etc, which does not change the outcome but makes the win even easier for Black.

35...bxa4 36.bxa4 f5 37.Ke2 Ke7 38.Ba3 Ba5 39.Bd3 Bb4 40.Bxa6 Bxa5

Reducing pawns to one side of the board can give drawing opportunities, but is insufficient here.

The game finished:

41.Kf3 Kf6 42.g4 g6 43.Bd3 Kg5 44.gxf5 gxf5 45.Bc4 e5 46.Be2 Bc7 47.Kg2 h6 48.Bd1 e4 49.Be2 Kf4 50.Kf2 B6+ 51.Kg2 Ke3 52.Bb5 f4 53.Kf1 f3 54.Ke1 Ba5+ 55.Kf1 Kf4 56.Bc6 e3 57.Bb5 Ke5 58.Bd3 Kd4 59.Bb5 Kc3 0-1

How Drawish Are Opposite-Coloured Bishops (part three)?

HERE is another opposite-coloured bishops ending I did not include in my series Opposite Exceptions.

White has just captured on e5 in Spanton (147 BCF) - Hass Ha'eri-Zadeh (-), Highbury (London) Rapid 1990

*****

*****

*****

*****

*****

*****

*****

*****

*****

*****

27...Kg6 28.Bxc7

White is three pawns up, but it will not be a simple matter creating a second passed pawn on the queenside.

28...Bc6 29.Ke3 Kf7 30.g4 g6 31.f5 Bd7 32.fxg6+?!

It was almost certainly better to advance the white king before making an exchange.

32...Kxg6 33.Kf4 Kf6 34.f4 Be6 35.Ba5 Bc4 36.Bd2 Be2 37.g5+ hxg5 38.hxg5+

White has a second passed pawn, but the passers are near to each other, making them less effective in this type of ending as they are more easily restrained by Black.

38...Ke6 39.e5?!

There is no need to rush forward with this pawn and so give up control of the d5 and f5 squares. Probably better is queenside play, starting, for example, with 39.b3.

39...Bd1 40.c3

Now Black's two queenside pawns hold up three white ones.

40...Bh5 41.Be3 Bg6 42.Bd4 Kd5 43.Ke3 Kc4?

It is difficult for White to make progress after a waiting move such as 43...Bh5.

White to play and win

*****

*****

*****

*****

*****

44.Kd2?!

Stockfish15.1 and Komodo14.1 give 44.e6, eg 44...Kd5 45.e7 Ke6 46.Bf6 Kd5 47.Kf4 Ke6 (47...Bf7 48.Kf5) 48.c4! bxc4 49.Ke3 Kd5 50.Kd2 Be8 51.Kc3 Kc5 52.Bg7 Bf7 53.Bf8 (threatens 54.e8=Q+) Kd5 (53...Be8 54.g6 ) 54.Kb4 Be8 55.Ka5 Ke5 56.Kb6 Kf5 57.Kc7 Kxg5 58.Kd8 Bc6 59.e8=Q Bxe8 60.Kxe8 with a simple win.
44...Kd5 45.Ke3 Ke6?

Marking time with the bishop puts up considerably stiffer resistance.

46.Kf4

The engines point out a pretty win by 46.b3! axb3 47.c4! bxc4 48.a4 etc.

46...Kd5 47.Be3 Ke6

HH-Z offered a draw.

Stockfish15.1 reckons 47...Kc4!? draws, but Komodo14.1 disagrees. Best-play may run 48.Bb6 Kd5 49.Bd8 Bd3 50.Bf6 Bg6 51.Kg4 Ke4!? 52.e6!? Kd5 53.e7 Ke4 54.Kg3 Be8, when the engines still strongly disagree, but this may be one of those rare positions where a three-pawn advantage is not enough.

48.Bb6 Kd5 49.Bd8 Ke6?!

Stockfish15.1 reckons marking time with the bishop on the h7-b1 diagonal is best, although Komodo14.1 reckons that still loses.

50.Bf6

50.c4!? bxc4 51.Bf6 seems to win in a similar way to the note at move 44.

50...Kd5 51.Ke3 Bf5 52.Kf3 Bg6 53.Bg7 Ke6?! 54.Ke3

Again sacrificing the c pawn looks promising.

54...Kd5 55.e6!? Kxe6

White is two pawns up, but Stockfish15.1 reckons Black is holding, although Komodo14.1 gives White the upper hand

*****

*****

*****

*****

*****

56.Kd4 Kf7

Better may be 56...Kf5!?

57.Bf6 Bc2 58.c4!?

Ironically, perhaps, now that this pawn push is no longer a sacrifice, the engines do not much like it. Instead they suggest 58.Kc5 Bd3 59.Kb4, although as usual disagreeing as to how much of an advantage White has.

58...bxc4 59.Kxc4 Bb3+ 60.Kd4 Bc2 61.b4!? axb3 62.Kc3 Ke6 63.Kc4 Kd7 64.Kd5 1-0 (Time)

The final position is a tablebase draw.

How Drawish Are Opposite-Coloured Bishops (part two)?

I POSTED the first part of this series on March 29.

HERE is another opposite-coloured bishops ending I did not include in my earlier series Opposite Exceptions.

Black has just captured on g6 in Peter Szabo (141 BCF) - Spanton (147 BCF), Barbican (London) Major 1990

*****

*****

*****

*****

*****

*****

*****

*****

*****

*****

Black is two pawns up and has the possibility of creating a queenside passed pawn to complement the e4 protected passer. Unlike in part one of this series, the advantage is enough for a win to be straightforward.

36.Kg2 Be3 37.Kg3

The king is tied to defending the f4 pawn, making Black's task even easier.

37...Kf6 38.a4 c5 39.Bb3?

This speeds the process of creating a second passed pawn, which is often decisive in this type of ending.

The game finished:

39...Bc1 40.Bd5 Bxb2 41.Bxb7 c4 42.Bd5 c3 43.Bb3 Bc1 44.Bc2 Bd2 45.Bb3 Ke7 46.Bc2 Kd6 47.Bb1 Kc5 48.Bc2 Kb4 49.Bd1 Ka3 50.Kh4 Bxf4 51.Kh5 e3 0-1

Perfect Trinity

RETURNING by tube on the Central Line from a match a few days ago I spotted the following carriage serial number: 91273
As usual with these tube puzzles, each number should be used once, and once only, and must be added, subtracted, multiplied or divided to make a balanced equation.
What makes this special is that there are at least three perfect solutions, ie solutions that use the numbers in the order they appear.
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
My solutions: 9 - 1 ( ÷ 2) = 7 - 3

9 - 1 + 2 = 7 + 3

9 - 1 + 2 - 7 = 3

Lessons From Fagernes IX

MY round-nine game featured an instructive moment in the opening after the moves 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.Bd2 Nc6 6.Nb5 Bxd2+ 7.Qxd2 Nxd4 8.Nxd4 cxd4.

How should White proceed?

*****

*****

*****

*****

*****

*****

*****

*****

The game saw 9.Qxd4, which is OK and has been played by grandmasters, but may not be best. The point is the d4 square is probably best reserved for the knight, which is why the most-popular moves in ChessBase's 2023 Mega database are 9.Nf3 and 9.f4 (to be followed by Nf3).

The game continued 9...Ne7 10.Nf3, after which 10...Nc6 may be most exact, not least because 11.Bb5??, hoping to follow up with Bxc6(+), leaving Black with a bad bishop versus a good white knight, loses to 11...Qa5+ etc.

LESSON: 'obvious' recaptures my not be so obvious if sufficient thought is given to understanding a position.

Lessons from Fagernes VIII

THE following position was reached in my round-eight game.

White has just captured on c4 - what should Black do about the triple-threat to the d6 pawn?

*****

*****

*****

*****

*****

*****

*****

*****

*****

The game saw the passive 24...Ne8?!, which sets a trap, ie 25.Nb6 Rxc1 26.Nxd7?? Rxe1+ 27.Bg1 Rg6, but this is easily avoided by playing 26.Rxc1, after which my main analysis engines, Stockfish15.1 and Komodo14.1, give White at least the upper hand.

Even worse is 24...d5? 25.Nd6! (this is stronger than 25.Ne5, according to the engines) Rxc1 26.Rxc1 Bg6 27.Qb4 with huge pressure against Black's queenside, eg 27...b5 28.Qc5 Kh7 29.Nc8 Bf5 30.Qc7 Qxc7 31.Rxc7 Ne8 32.Rc5 (32.Rxf7?! Rc6) Nf6 33.g4 Bg6 34.Be5 Ne8 35.Rxd5.

Black had a simple solution in 24...Rc6, which maintains a reasonable amount of activity, and gives equal chances, according to the engines.

LESSON: a passive move, however trappy, is usually inferior to an active one.

Championship Chess

PLAYED last night in the Battersea club championship.

Spanton (1944) - Malcolm Dancy (1795)

Siclian Bb5(+)

1.e4 c5 2.Nf3 Nc6 3.Bb5 Nd4!?

This has a dubious reputation but has been played by modern grandmasters.

4.Nxd4 cxd4 5.0-0

There are 43 games in ChessBase's 2023 Mega database with the blunder 5.d3??, although 14 blacks managed to miss the winning 5...Qa5+.

5...a6 6.Bc4!? g6!?

Perhaps 6...e6, blunting the white light-square bishop, is more logical, although it scores only 31% in Mega23, while the less-popular text scores 44%.

7.c3 Bg7 8.cxd4 Bxd4 9.Qb3!?

This more-or-less obliges Black to spend a tempo on ...e6, but it leaves the white light-square bishop without a good retreat-square from c4.

9...e6 10.Nc3 b5

Dmitrij Kollars (2562) - Alban Delorme (2320), Chess,com Rapid 2019, saw 10...Ne7 11.Ne2 Bg7 12.d4 d5 13.Bd3 with a slight edge for White, according to Stockfish15.1 and Komodo14.1 (1-0, 58 moves).

11.Be2 Bb7

The engines are not keen on this natural-looking move, preferring 11...Ne7 or 11...Qb6.

12.d3 Qb6 13.Bg5

Aiming for kingside play, but the plan proves to be slow. The engines reckon White should be playing on the queenside, eg 13.a4 bxa4 14.Qxb6 Bxb6 15.Nxa4, claiming a slight edge for White.

13...Ne7 14.Bh4

Black has the upper hand after 14.Bxe7 Kxe7, according to the engines, but they do not like the text either, suggesting 15.Bf3 (Stockfish15.1) or 15.Bd1!? (Komodo14.1).

14...Bc6 15.Kh1 0-0

How would you assess the position now both sides have castled?

*****

*****

*****

*****

*****

White has (laboriously) prepared an f-pawn push as part of a kingside attack, but most white pieces are not well-placed to support it. Meanwhile Black's pieces are well-placed to control the centre and thus render the white kingside-attack harmless. The engines reckon Black has the upper hand.

16.f4 Bg7 17.f5?!

'Consistent', but missing that Black has a good fork.

17...Nd4 18.Qd1

How should Black proceed?

*****

*****

*****

*****

*****

18...b4?!

Almost certainly better is the simple 18...exf5 19.exf5 Nxf5. In the postmortem White got dangerous play by sacrificing the exchange, but the engines are unimpressed, eg 20.Rxf5!? gxf5 21.Bf3 Rfe8 (21...Bxf3 22.Qxf3 gives White practical chances) 22.Nd5 Bxd5 23.Bxd5 Rac8 24.Qb3 Qh6! 25.Bxf7+ Kh8, after which Black has the deadly threat of ...Rc1+.

19.Na4 Qc6?!

This lets the white queen's rook be developed with tempo. I suggested 19...Qa7?! in the postmortem, but the engines give 20.f6 Bh6 21.Bf2 with sharp play. Perhaps best is my other suggestion of 19...Qb5, when 20.f6 Bh6 21.Bf2 does not produce a pin.

20.Rc1 Qb5??

The engines reckon Black is still better after 20...Qd6.

21.Rc5 Nxe2

Black is sort-of still in the game after the engines' 21...b3! 22.Rxb5 axb5 23.axb3 bxa4, but they give White a winning advantage.

22.Rxb5 axb5 23.Nb6 Rxa2 24.f6 1-0

Lessons From Fagernes VII

MY round-seven game featured a simple but instructive endgame mistake.

White has the better pawn-structure, but Black's activity, including threats from the passed d4  pawn, have forced all the white pieces to the back rank

*****

*****

*****

*****

*****

*****

*****

*****

Stockfish15.1 reckons Black is equal after 27...Nd5 or 27...Nc6. Komodo14.1 agrees those are the best moves, but gives White a slight edge.

The game saw 27...Nc2, presumably expecting an exchange of knights. Indeed Black would probably be winning after 28.Nxc2?? Rxc2.

Instead 28.Nd3 Rc8 29.Bd2 trapped the black knight, which cannot be anchored by the d4 pawn, and no longer has an escape route via b1.

LESSON: what goes up, must come down, or, in chess terms, what goes in, must have a way to come out if it is to survive.

Lessons From Fagernes VI

MY round-six game reached an interesting late-middlegame.

Black has just captured on h6

*****

*****

*****

*****

*****

*****

*****

*****

*****

White is a pawn up and has an obvious plan of advancing the queenside pawn-majority to create a second queen. But the situation is complicated by White having the more-vulnerable king.

A few moves later the following position was reached.

White has just played 43.Qe2 to block the check - how should Black reply?

*****

*****

*****

*****

*****

43...Qf5?!

My opponent may have avoided an exchange of queens because he hoped to get at my king. But after the text Stockfish15.1 and Komodo14.1 immediately flash red, and probably with good reason - actually, with several good reasons:

A) An exchange of queens brings about a rook-and-pawn ending,, and such endings often have strong drawish tendencies.

B) With queens off the board, Black's weak pawns are easier to defend.

C) The text has the specific problem that 44.Rh5 Qg6 45.Re5 allowed White to set up an ideal heavy-piece battery of queen behind rook.

The engines reckon Black should play 43...Rc8, after which White's only hope of a slight edge - Stockfish15.1 is not even sure White has such an edge - lies in 44.Re1, according to the engines. A sample line continues 44...Kf7!? 45.g4 b5 46.Kf3 Qxe2+ 47.Rxe2 Rc4, after which Black's active rook gives enough compensation for being a pawn down, according to the engines.

However, the situation is not as simple as the above may make it seem. For instance, White could meet 43...Rc8 with 44.Rd1, intending Rd2. It might seem natural to reply 44...Qxe2+?!, but the engines reckon this is wrong, presumably at least partly because White gets a more-active version of the previous line, but also because 44...Kf7 45.Rd2 can be met by 45...Qh7, when Black can successfully go back to harassing the white king, eg 46.Kg2 Rh8 with complete equality, according to the engines.

LESSON: simple-looking positions can be deceptively complicated, at least from a positional view. One of the most important things to look for is piece-activity, which can become more crucial as the number of pieces diminishes. The move ...Qf5 allowed White to become active, and that, combined with the extra pawn, decided the game.

Lessons From Fagernes V

MY round-five game reached the following position.

What should White play?

*****

*****

*****

*****

*****

*****

*****

*****

*****

The position arises from the Exchange Variation of the Queen's Gambit Declined. The only slightly unusual feature is that the black dark-square bishop is on d6 rather than performing its more-normal role of breaking the pin on e7.

This apparently minor difference becomes huge if, as 20 times in ChessBase's 2023 Mega database, White continues 11.0-0?

This has been played by two players rated in the 2600s, two rated in the 2400s, three rated in the 2300s, two rated in the 2200s and two rated in the 2100s.

One of the remaining nine games was played at the 1966 Olympiad in Havana, ie before the world chess federation Fide adopted Arpad Elo's rating system, so it seems the mistake is particularly attractive to relatively strong players.

The reason it is a mistake is Black can play the Greek Gift sacrifice, ie 11...Bxh2+!. although five players, including one rated 2491, appear to have failed to spot this opportunity, or at least rejected it for a spurious reason.

After 12.Kxh2 Ng4+ 13.Kg3 Black has 13...g5, which gives a large advantage.

LESSON: apparently minor differences can make major changes. If, in the first diagram, whites, when considering castling, had asked themselves what captures would black then have, only two would have been possible: the clearly bad ...Rxe3 and the alarm-bell ringing ...Bxh2+. It is hard to believe anyone, especially a titled player, would play 11.0-0? if such a basic check had been made.

Lessons From Fagernes IV

MY round-four game saw Black play the Cozio Variation of the Spanish, ie 1.e4 e5 2.Nf3 Nc6 3.Bb5 Nge7, which is only the sixth-most popular Spanish continuation in ChessBase's 2023 Mega database, but nevertheless occurs 9,728 times.

The starting position of the Cozio

*****

*****

*****

*****

*****

Historically the most-popular continuations are 4.0-0, 4.c3 and 4.d4, but 4.Nc3!?, which was an Alexander Alekhine favourite, is arguably the most-critical line.

It might look innocuous, but 4...g6, the normal continuation in the Cozio, runs into 5.d4 exd4 6.Nd5!? Bg7 7.Bg5.

Then Black's most-popular reply, 7...h6, is the only way to keep the game going, according to Stockfish15.1 and Komodo14.1, but 8.Bf6 is pleasant for White.

What else can Black sensibly play?

*****

*****

*****

*****

*****

My game, after 4.Nc3!?, saw 4...Ng6, which dates back to at least 1867 and was played both times by Magnus Carlsen when he reached the position in 2020.

Despite the world champion's endorsement, the move leaves a poor impression, although it may be that the real problem goes back to 3...Nge7.

A) Position after 4...Ng6

*****

*****

*****

*****

*****

Compare the above diagram with the following one, which arises after 1.e4 e5 2.Nf3 Nc6 3.Nf3 Nf6.

B) Position after 3...Nf6

*****

*****

*****

*****

*****

I do not care how strong Carlsen is, there is no doubt diagram A) is more favourable for White than diagram B).

My game at Fagernes continued, from diagram A), 5.d4 exd4 6.Nxd4, which is similar to a Scotch Four Knights, except it is White to move, not Black, and the black king's knight is on g6 rather than f6.

After 6...Nxd4!? (Carlsen has played this) came 7.Qxd4 c6 8.Bc4 Qb6 9.Be3!?, and I went on to win, although the engines reckon Black is OK after the apparent-novelty - it does not appear in Mega23 - 9...Qxb2!?

LESSON: lots of club players, myself included, like to get out of mainlines early to try to set novel problems to their opponents. The downside of this is that sidelines tend to be either inferior, or to require precise play if the opponent knows, or finds over the board, the theory.

Lessons From Fagernes III

MY round-three game featured the Alapin Variation of the Sicilian Defence.

It began 1.e4 c5 2.c3 g6!? (Black is effectively playing a Hyper-Accelerated Dragon) 3.d4 cxd4 4.cxd4 d5.

White has a slight advantage, according to Stockfish15.1 and Komodo14.1

*****

*****

*****

*****

*****

*****

*****

*****

Despite the engines' optimism about the white position, the two main continuations in ChessBase's 2023 Mega database, 5.e5 and 5.exd5, score disappointingly - 46% and 47%, respectively.

My game saw 5.Nc3!?, which scores an almost-as-poor 48%.

After the virtually forced reply 5...dxe4, my opponent played 6.Nxe4, when Stockfish15.1 reckons Black already has a slight edge, although Komodo14.1 calls the game equal.

Black is posed more problems by 6.Bc4!?

Black's difficulties were starkly illustrated in David Howell (2334) - Spanton (2004), Isle of Man 2004, which continued 6...Nf6 7.Qb3 e6 8.d5 exd5 9.Nxd5 Nxd5 10.Bxd5 Bb4+?? 11.Qxb4 Qxd5 12.Qc3 1-0.

Position after 12.Qc3 - Black cannot avoid losing significant material, eg 12...0-0 is met by 13.Bh6

*****

*****

*****

*****

*****

Two years later, again at the Isle of Man, I faced the same line in Ian Marks (1944) - Spanton (2075), which varied with 10...Qe7.

The game continued 11.Bd2!? (IM said afterwards he reckoned this is better than 11.Qc3, although fast-forward 17 years and my main engines both prefer the latter) Nc6 (the engines much-prefer 11...Nd7!?) 12.Bxc6+ bxc6 13.Qc3 e3! (everything else loses on the spot) 14.Qxc6+ Qd7 15.Qxd7+ Bxd7 16.Bxe3 Bb4+ 17.Bd2 Bxd2+ 18.Kxd2 0-0-0.

Black has full compensation for a pawn, according to the engines, and went on to win

*****

*****

*****

*****

*****

Seven years later, at Guernsey, I tried playing the white side in Spanton (2030) - Ad Reijneveld (2135), varying with 11.Qc3.

The game continued 11...Qb4 12.Bd2 Qxc3 13.Bxc3 Rg8, at which point I played the probable-novelty 14.0-0-0!?

After 14...Nc6 15.Bxe4 Be6 16.Nf3 Be7 17.Kb1 Rd8 18.Rxd8+ Kxd8 19.Rd1+ Kc8 20.Nd4 Nxd4 21.Rxd4 Rd8 a draw was agreed.

The position is completely equal, according to the engines

*****

*****

*****

*****

*****

In 2019, again at Guernsey, the game Ton Goris (1913) - Reijneveld (2056) varied with 14.Bxe4.

The game continued 14...Nc6 15.Nf3 Bg7 16.Rc1 Bd7 17.0-0 Bxc3 18.Rxc3, at which point the engines reckon 18...Ke7!? equalises.

Instead the game saw 18...Kf8?! 19.Rd1 Be8, after which White has a large lead in development (1-0, 33 moves).

The somewhat picturesque final position of Goris - Reijneveld

*****

*****

*****

*****

*****

LESSON: a little experience goes a long way in sharp positions, but once (club) players are on their own resources, anything can happen. I do not know how much of this analysis I would have remembered if my opponent at Fagernes had played 6.Bc4!?, but my (admittedly limited) familiarity with the position after 5...dxe4 meant I was reasonably confident, however he had continued.

Lessons From Fagernes II

THE following position could have been reached in my round-two game.

How would you assess the position, bearing in mind it is White to play?

*****

*****

*****

*****

*****

*****

*****

*****

*****

Black is ahead by the exchange and four pawns, and threatens to win the a2 pawn, giving him four connected and passed queenside pawns. But Black has doubled g pawns, and the white d pawn will win back the exchange.

Komodo14.1 reckons the position is completely equal, although Stockfish15.1 gives Black a slight edge. This surprised me - I thought Black must be winning - but concrete analysis seems to support this.

One white move, 24 Ra1?, can be dismissed as horribly passive, while 24.Rf2?? is a blunder thanks to 24...Rf8+. The 'obvious' 24.d8=Q?  Rxd8 25.Bxd8 is also pretty hopeless after 25...Rxa2.

White has to find 24.Re1, the threat being to play 25.Re8+, winning the black queen's rook and following up with promoting the d7 pawn to a queen.

Black can prevent this easily enough with 24...Kf7, but then 25.d8=Q Rxd8 26.Bxd8 gives White good counterplay, thanks to the threat of Re7+.

Both lines lead to unclear play, but the game seems to remain within the drawing margin.

Black could instead try 24...Kf8, but then the engines give 25.Be7+ Kf7 26.Ba3!?, eg 26...Rxa2 27.Re7+ Kf6 28.g5+! Kxg5 29.Re8 Rxa3+ 30.Ke2 Rxe8+ 31.dxe8=Q, after which they reckon the white queen can at least hold its own against the black rook and seven pawns, although again the position is, to say the least, unclear.

LESSON: we all know undue materialism is the enemy of making correct assessments, but it is not always properly remembered in the heat of battle. I, with white, rejected the line leading to the first diagram as I dismissed White's chances without concretely trying to analyse the position.

Lessons From Fagernes

MY round-one game at Fagernes began 1.e4 e6 2.d4 de5 3.e5 c5 4.c3 Bd7 5.Nf3 Qb6!?

How should White respond to the black plan of playing ...Bb5, the idea being to swop the bad black light-square bishop for its good white counterpart?

*****

*****

*****

*****

*****

*****

*****

*****

The game saw 6.Qb3!?, which cuts across Black's plan of exchanging bishops. However, after 6...Qxb3!? 7.axb3 cxd4 8.Nxd4 Bc5 9.Be3 Bxd4!? 10.Bxd4 Nc6 Black has equalised, according to Stockfish15.1 and Komodo14.1.

The engines reckon even more convincing for Black is 6...f6!?, which does not appear in ChessBase's 2023 Mega database, a sample line running 7.exf6 Nxf6 8.Na3!? Qc7!? 10.Nb5 Qd8!?, after which Stockfish15.1 gives Black a slight edge, although Komodo14.1 reckons the position is equal.

Going back to the diagram, the most-popular move in Mega23 is 6.Be2!?, effectively ignoring Black's exchanging-bishops plan, or at least not being worried about it.

The mainline continues 6...Bb5 7.0-0 Bxe2 8.Qxe2.

Position after 8.Qxe2

*****

*****

*****

*****

*****

Black has carried out the plan, but it has cost time, leaving White ahead on development. Komdodo14.1 reckons White is slightly better, although Stockfish15.1 flips between agreeing with that verdict and calling the position equal.

LESSON: just because an opponent's plan is transparently obvious, it does not follow that the plan should be foiled. It may be natural to do so, but most strong players who have reached the position after 5...Qb6!? have decided (rightly, according to the engines) Black should be allowed to get on with it.

Fagernes Game Nine

Spanton (1806) - Jiabao Eivind Ji (1704)

French Winawer

1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.Bd2 Nc6

The mainline in ChessBase's 2023 Mega database runs 5...Ne7 6.Nb5 Bxd2+ 7.Qxd2+ 0-0 with an equal game, according to Stockfish15.1 and Komodo14.1.

6.Nb5 Bxd2+ 7.Qxd2 Nxd4

How should White proceed?

*****

*****

*****

*****

*****

8.Nxd4

This is easily the most-popular choice in Mega23, and is recommended by the engines, but 8.Nd6+!? gives White decent compensation for a pawn. It arguably comes down to a matter of taste.

8...cxd4 9.Qxd4

This 'obvious' recapture is less-popular in Mega23 than 9.Nf3 and 9.f4. Meanwhile Stockfish15.1 suggests 9.h4!?, and both engines come to plump for 9.0-0-0.

9...Ne7 10.Nf3 Bd7

The immediate 10...Nc6 is possible, not least because 11.Bb5??, hoping to swop bishop for knight to produce a minor-piece clash of good knight versus bad bishop, loses to 11...Qa5+.

11.Bd3 Nc6 12.Qg4 g6

Not 12...0-0 13.Bxh7+ Kxh7 14.Qh5+ Kg8 15.Ng5+ - a superior version of the Greek Gift sacrifice.

13.0-0 Qe7 14.Rfe1 0-0-0

Now both sides have castled, how would you assess the position?

*****

*****

*****

*****

*****
Positions with opposite-side castling are among the hardest to judge in chess. Two things are, I believe, clear: White has the better bishop, and the black king looks less secure than its white counterpart. The engines give White the upper hand.

15.a4 Rdg8 16.h4!?

Hoping, believe it or not, to keep lines on the kingside as closed as possible. The engines prefer 16.Nd4.

16...h6 17.c3 g5 18.h5 f5 19.exf6 Qxf6 20.Qg3 Re8?!

Probably better is 20...g4, followed by pressure against h5.

How should White continue?

*****

*****

*****

*****

*****
21.Ne5?!

Occupying the hole at e5 seemed so natural to me, but is strongly disliked by the engines, which give 21.Rad1 or 21.b4!?

21...Rhf8?

After 21...Nxe5 22.Rxe5 Qf4 White is only slightly better, according to the engines.

22.a5?

The engines reckon White is winning after 22.Ng6 Rf7 23,b4 e5 24.b5.

22...Nxe5 23.Rxe5 Qf4 24.Qxf4 Rxf4 25.g3 Rf6 26.Rae1 Rf3

Black's pressure against f2 compensates for White's pressure against e6.

27.Be2 Rf5 28.Bg4 Rxe5 29.Rxe5 Kc7 30.Kf1 Kd6 31.Re3 Rf8 32.Ke1 b6!? 33.axb6 axb6 34.Rf3 Rxf3 35.Bxf3 e5 36.Kd2 Kc5 37.Be2 Be8 38.b4+ Kd6

JJ offered a draw. I played on, but the position is completely equal.

39.Ke3 Ke6 40.Bg4+ Kf6 41.Be2 Kf5 42.f3 Ke6 43.f4 exf4+ 44.gxf4 Kf5 45.Bd3+ Kg4 46.Be2+ Kh4 47.fxg5 Kxg5 48.Bf3 Bxh5 49.Bxd5 Be8 50.c4 h5 51.Kf3 Kf5 52.Ke3 Ke5 53.Bf3 Bf7 54.c5 bxc5 55.b5?!

A last, somewhat-foolish, attempt to win.

55...Bd5

Komodo14.1 claims 55...Kd6 and 55...h4 win, but the Syzygy endgame tablebase shows the position is drawn.

56.Bxh5 Kd6 ½–½

Summing Up Fagernes

MY score in the open of +5=2-2 cost me 2.2 Fide elo. I finished tied for 3rd-8th (third on tiebreak).

Tiebreaks did not count for the main prizes, so I received 50 euros, but I won a further 90 euros for being the best senior (60+).

My third-place finish means I qualify for next year's "GM Group," which is for males over 2000 and females over 1900.

Farewell shot

Fagernes Game Eight

Rafe Rehman (1824) - Spanton (1806)

Jobava-Prié

1.d4 d5 2.Nc3 Nf6 3.Bf4 e6 4.Nb5!?

The main move in ChessBase's 2023 Mega database is 4.e3, but Stockfish15.1 prefers the text.

4...Bd6!?

Offering the bishop-pair ...

5.Nxd6 cxd6

... in return for a preponderance of pawns on the centre two files.

How would you assess the resulting position?

*****

*****

*****

*****

*****

Stockfish15.1 gives White a slight edge; Komodo14.1 calls the game equal.

6.e3 Nc6 7.Nf3 0-0 8.c3 Re8 9.Be2 h6!?

The idea of this move is to prevent ...e5 being met by Bg5.

10.h3!?

This allows the bishop to retreat out of harm's way after ...e5.
10...e5 11.Bh2 Bf5 12.0-0 e4!?

This seems to be a novelty, and possibly not a good one. The move gains space, but makes d6 a target and will allow White to operate on the queenside without fear of the centre being opened. It might be fine if Black could quickly organise kingside play, but that does not prove simple (the black king's rook might be better placed on f8, supporting and advance by the f7 pawn).

13.Nd2 Qd7 14.Qb3 Rac8 15.Kh1 a6 16.Qb6 Re6!?

Defending d6 while at the same time hoping to create kingside threats, but the latter may be unduly optimistic, in which case 16...Red8 is simpler.

17.Rac1 Ne7 18.Qb3 Nc6 19.Qa3 Na7?!

Black's pieces make a bad impression

*****

*****

*****

*****

*****

20.Qb3

White has a positionally won game after 20.c4!?, according to the engines, and if 20...dxc4 21.Nxc4 Nb5, then 22.Qa5 threatens Nb6 and a4.

20...Nc6 21.c4 Na5 22.Qa3 Nxc4 23.Bxc4!?

Forcing opposite-coloured bishops, which, contrary to the usual situation in an ending, gives the middlegame player with the initiative increased attacking chances.

23....dxc4 24.Nxc4 Ne8?!

Passive play. Black equalises (Stockfish15.1), or at least comes close to equalising (Komodo14.1), after 24...Rc6.

What should White play?

*****

*****

*****

*****

*****

25.Rg1?

Correct is 25.Nb6, meeting 25...Rxc1 with 26.Rxc1, but not 26.Nxd7?? Rxf1+ 27.Bg1 Rg6 etc.

25...Qd8?!

Probably better is 26.d5, and if 26...Ne5 then 27.Qd8, after which the white knight can be evicted at will.

26.Na5 Rc7 27.Qb4 Ree7 28.Nc4 Re6?!

Still hoping for kingside play, but it was probably better to concentrate on defending the queenside and centre with 28...Red7.

29.d5 Rf6 30.Na5 Bc8 31.Bg3 Qe7?

Necessary was 31...Rf5, with just a slight edge for White, according to the engines.

32.Rxc7 Nxc7?

It is better to give up the e4 pawn with 32...Qxc7.

33.Rc1

What should Black play?

*****

*****

*****

*****

*****

33...Bxh3!?

This is best, according to the engines, but clearly losing.

The game finished:

34.Qxb7 Nxd5 35.Qxd5 Be6 36.Qxe4 Rf5 37.Kg1!? Rxa5? 38.Rc8+ 1-0

Fagernes Round Nine

DREW with white in 56 moves against junior (born 2012) Jiabao Eivind Ji (1704).

View from the venue-hotel bar

Fagernes Round Eight

LOST with black in 38 moves after being downfloated against junior (born 2009) Rafe Rehman (1824).

The lake is used as an airstrip by light aircraft

Fagernes Game Seven

Spanton (1806) - Sigurd Lauvås Gaure (1616)

Sicilian Bb5(+)

1.e4 c5 2.Nf3 d6 3.Bb5+ Bd7 4.Bxd7+ Qxd7 5.0-0 Nc6 6.c3 Nf6 7.Qe2 e5!?

I faced 7...e6 in round six.

How should White respond?

*****

*****

*****

*****

*****

8.d4!?

The mainline in ChessBase's 2023 Mega database runs 8.Rd1 Be7 9.d4 cxd4 10.cxd4 exd4 11.Nxd4 with at least a slight edge for White, according to Stockfish15.1 and Komodo14.1. However Black can put a spoke in this with the engine-approved 8...Qg4!?, after which the mainline runs 9.d3 Be7 with an unclear position - Black has more space, but White has the better bishop.

8...cxd4 9.cxd4 exd4 10.Rd1

Probably not 10.e5?! dxe5 11.Nxe5 Qe6.

10...Rc8

This sets a trap, but the engines prefer 10...Qg4, when Komodo14.1 continues 11.Nxd4 Qxe2 12.Nxe2 Nxe4 13.Nbc3 Nxc3 14.Nxc3, claiming good compensation for a pawn. Stockfish15.1 gives 11.Nbd2 (it also likes 11.Qb5) Be7 12.h3 Qe6 13.b4!?, claiming compensation for a pawn, but no more.

How should White proceed?

*****

*****

*****

*****

*****

11.Na3

Not 11.Nxd4?? Nxd4 12.Rxd4 Rxc1+, but the engines prefer 11.Bg5, and if, as in the game, 11...Qg4, then 12.Bxf6 gxf6 13.Nxd4 regains the pawn and leaves Black with multiple weaknesses.

11...Qg4!?

This move has, to an extent, dominated the notes, and it is also interesting here.

How should White reply?

*****

*****

*****

*****

*****

12.Bg5

Komodo14.1 gives 12.Nxd4!? Qxe2 13.Nxe2 Nxe4 14.Be3, awarding White a slight edge, presumably rating White's lead in development as more significant than Black's extra pawn. Stockfish15.1 gives 12.Nb5, and if 12...Qxe4 then 13.Qf1!? Be7 14.Nxd6+ Bxd6 15.Re1 0-0 16.Rxe4 Nxe4, when White has queen for rook, knight and pawn, and, according to the engine, a slight edge. However, neither engine is very sure of the position, for example for a long time it is Stockfish15.1 that recommends 12.Nxd4!? and Komodo14.1 that gives 12.Nb5.

12...Qxe4?!

Black at least equalises with 12...Nxe4 13.h3 Qe6, according to the engines.

13.Qb5

Black is two pawns up, but the queen-and-king line-up on the e file means Black is worse.

What should Black play?

*****

*****

*****

*****

*****

13...a6?

Best, according to the engines, is 13...Be7 14.Bxf6 gxf6, although White immediately regains one pawn and has more than enough compensation for the other one.

14.Qxb7 Rb8?

The engines recommend 14...Ne7, but Black's position is horrible.

15.Qxb8+??

Plenty good enough is 15.Qxa6 as 15...Rxb2? fails to 16.Re1, the point being White has two pieces covering e2. But even stronger seems to be the engines' 15.Qc7, eg 15...Be7 16.Re1 Qd5 17.Rxe7+.

15...Nxb8 16.Re1 Kd7 17.Rxe4 Nxe4 18.Bf4 Nc6

After the flurry of (misplayed) tactics, how would you assess the position?

*****

*****

*****

*****

*****

Equal, according to the engines, but not easy for either side to play.

19.Rd1 g5!?

This positional pawn sacrifice is the engines' top choice for a while, although they later come to also like 19...d5 and 19...Nc5.

20.Nxg5 Nxg5 21.Bxg5 Bg7 22.Nc2

Material is level in this ending, but is the position?

*****

*****

*****

*****

*****

The engines reckon the game is equal, but clearly it is one of those positions in which there are imbalances, a concept popularised by international master Jeremy Silman. All of Black's pawns are isolated, and two of them are doubled, but Black has more space. and the White pieces are not well-coordinated. Black's king is also more active.  I suspect the position is easier for Black to play, but both sides have things to worry about.

22...Re8 23.Kf1 Re5 24.Bf4 Rc5 25.Ne1 Be5 26.Bc1 Nb4

Not 26...Bxh2? 27.g3 etc.

27.a3 Nc2?

The knight has no way out from this square.

28.Nd3 Rc8 29.Bd2

This takes away the knight's escape via a1.

29...Ke6?!

Unprotecting the rook does not help matters.

30.Rc1 Kf5 31.Nb4 d3!? 32.Nxd3 Rc4?!

A better try is 32...Rc6, but 33.Bc3 Nd4 34.Re1, threatening 35.g4+, wins, according to the engines.

The game finished:

33.Ne1 Ke4 34.f3+ Kd4 35.Bc3+ Kc5 36.Rxc2 Rh4? 37.Bxe5+ 1-0