Bad Neuenahr Round Nine

I WAS upfloated against the top seed, a Fide master, on board one.

Stephan Buchal (2202) - Spanton (1911)

King's Indian Attack/English Symmetrical/Réti

1.Nf3 d5 2.g3 Nf6 3.Bg2 c5 4.0-0 g6 5.c4

What's in a name? At various points in the opening, ChessBase classifies it as "A08: King's Indian Attack," "A34: Symmetrical English: 2 Nc3, lines with ...d5" and "A38: Symmetrical English vs ...g6 4 Bg2 Bg7 5 Nf3 Nf6," and I suspect some people would regard it as a Réti

*****

*****

*****

*****

*****

5...Bg7 6.cxd5 Nxd5 7.Nc3 0-0

With both sides castling early, how would you assess the game?

*****

*****

*****

*****

*****

The position occurs 2,330 times in ChessBase's 2026 Mega database, with White scoring a below-par 51%. Stockfish18 calls the game completely equal, but Dragon1 gives White at least the better part of equality.

8.Nxd5!? Qxd5 8.d3 Nc6

This is the point at which, according to ChessBase, the opening switches from A34 to A38.

9.d3 Nc6 10.Be3!?

How should Black proceed?

*****

*****

*****

*****

*****

10...Qd6

This is the commonest continuation in Mega26, followed by 10...Bd7, but the engines reckon Black should grab the proffered pawn with 10...Bxb2!? After 11.Rb1 Bf6!?, White has various moves, including offering a second pawn, temporarily, with 12.Nd4!?, when 12...Qxa2!? 13.Ra1!? Qb2 14.Nxc6 bxc6 15.Rb1!? Qa2 16.Bxc6 Bh3!? 17.Bxa8 Bxf1 18.Bf3 Bh3 19.Bxc5 is completely equal, according to the engines. But there are lots of alternatives along the way, including 14.Rb1 Qxb1!? 15.Qxb1 Nxd4, when Black has rook, knight and two pawns for a queen - imbalanced, but equal, according to the engines. Give the engines enough time, and Stockfish18, at least, rates 10...Bd7 on a par with 10...Bxb2!?, and it is not as though they reckon the text, which has been played by many grandmasters, is a mistake.

11.Rc1 Nd4

Grischuk and others have tried 11...b6?!, but 12.d4 is a strong answer.

How should White proceed?

*****

*****

*****

*****

*****

12.Bxd4!?

The engines prefer the main move, 12.Nxd4, when 12...cxd4 13.Bd2 is slightly better for White, according to the engines, which value White's pressure against Black's queenside more highly than Black's central space.

12...cxd4 13.Nd2

What should Black play?

*****

*****

*****

*****

*****

13...Be6!?

This is the main move in Mega26, albeit from a small sample, but the engines come to more-or-less plump for another known move, 13...Qb4!? (the choice of an 1831, the lowest-rated player to reach the position in Mega26), but they also suggest apparent-novelties in 13...a5!? and 13...Bh6!?

Now White has the chance to capture the opponent's QNP, but should he?

*****

*****

*****

*****

*****

14.Qa4!?

After 14.Bxb7 (there are no examples of this move in Mega26) Rab8 15.Bf3!? Rxb2 16.Nc4 Bxc4 17.Rxc4 Rxa2, White has enough for a pawn, but no more, according to the engines. They reckon the text is best.

14...Bd7

The engines prefer two apparent-novelties in 14...Bh6!? and 14...Rfb8!?

15.Qb3 Be6?!

The engines suggest another apparent-novelty in 15...Rab8.

Now can White grab the b pawn?

*****

*****

*****

*****

*****

Yes, according to the engines, which give best-play as 16.Qxb7 Rab8 17.Qxa7 Rxb2 18.Ne4! Qb8 19.Qxe7, all of which occurred in Vlastimil Sejkora (2221) - Daniel Koval (2093), Tatranské Zruby (Slovakia) 2011, after which the engines reckon White had the upper hand (1-0, 36 moves).

My game went ...

16.Qa4 Bd7 17.Qb3 Be6?!

... at which point SB wrote down the move 18.Qa4 and announced he was claiming a draw by repetition.

½–½

My third draw in my last 30 games.

Bad Neuenahr Round Eight

Siegmund Kolthoff (2074) - Spanton (1911)

Maróczy Bind

1.Nf3 c5 2.e4 g6 3.d4 cxd4 4.Nxd4 Nc6 5.c4 Bg7 6.Be3 Bf6 7.Nc3 d6 8.Be2 0-0 9.0-0

This position occurs 20,723 times in ChessBase's 2026 Mega database

*****

*****

*****

*****

*****

*****

White's space advantage gives at least a slight edge, according to Stockfish18 and Dragon1.

9...Bd7 10.h3!?

Dragon1 prefers the commonest continuation, 10.Qd2, when the main line in Mega26 runs 10...Nxd4 11.Bxd4 Bc6 12.f3 a5!? 13.b3 Nd7 14.Be3!? Nc5 15.Rab1!? Qb6!? 16.Rfc1, with a slight edge for White, according to both engines.

Stockfish18 likes keeping pieces on the board with 10.Nc2!?, when the main line in Mega26 goes 10...Rc8 11.f3 a6 12.Qd2 Ne5!? 13.b3 (13.Na3!? is a popular alternative) b5!? (the almost equally popular 13...Qc7?! is a mistake, according to the engines, which reckon several moves, including 14.a4!?, give White the upper hand) 14.cxb5 Qc7 15.Bd4, leaving White with at least a slight edge after 15...axb5, according to the engines. But they reckon 15...Bxb5!?, which does not appear in Mega26, more-or-less equalises.

10...Nxd4 11.Bxd4 Bc6 12.Qc2!?

In 1904, in what is generally regarded as the first example of what came to be called the Maróczy Bind, the move 12.Qd3 was played against Geza Maróczy. Two years later, when Maróczy had the same position as White, he preferred 12.Bf3!? Today the text is most popular, perhaps not surprisingly - it scores 70% in Mega26.

How should Black respond?

*****

*****

*****

*****

*****

12...Nd7!?

The engines prefer the mainline 12...a5!?

13.Bxg7

Tiviakov reckons an exchange of dark-square bishops in the Maróczy Bind nearly always favours Black, but this position may be an exception - at least, the engines reckon it is. Note that 13.Be3 does not lose a pawn, as 13...Bxc3?! 14.Qxc3 Bxe4? runs into 15.Bh6, when 15...Nf6 16.Bxf8 gives White a large advantage.

13...Kxg7

How should White proceed?

*****

*****

*****

*****

*****

14.Rad1

SK was upset with himself in the postmortem for not having played 14.b4!?, which is indeed the engines' top choice. Black's best reply is perhaps 14...b6 (a 2331 lost a blitz game with 14...a6?? 15.b5), which is not the sort of move Black normally plays in the Maróczy Bind. However, the text is also popular, and gives White at least a slight edge, according to the engines.

14...a5 15.f4 Qb6+ 16.Kh1!? a4!? 17.Bg4 Nc5 18.Bf3

Possibly a novelty. Olga Badelka (2427) - Kamil Szadkowski (2278) Chess.com Blitz 2020, went 18.e5!? dxe5 19.fxe5 Rfd8?! 20.Nd5 Bxd5 21.cxd5, with a positionally won game for White, according to the engines (1-0, 30 moves). However, they reckon Black had several equalising moves in reply to19.fxe5, including 19...e6 and 19...Qb4.

18...Qb4 19.Ne2

SK was highly critical of this in the postmortem, but the engines are fine with it. although their top choice is probably 19.Be2 or 19.Nd5.

19...Ra6!? 20.Nd4 Rb6 21.Rf2 e5!?

Seeking to win the d4 square for Black's knight.

22.Nxc6

Also possible is 22.a3!? Qa5 23.Ne2, but the engines prefer the text.

22...Rxc6 23.f5 Nd7 24.a3!?

The engines suggest 24.b3 or 24.Be2.

What should Black play?

*****

*****

*****

*****

*****

24...Qb6

The engines reckon 24...Qxc4 25.Qxc4 Rxc4 26.Rxd6 Nc5 gives Black at least a slight edge.

25.Rfd2

Not 25.Qxa4?? Qxf2.

25...Nc5 26.Rd5

SK offered a draw.

Objectively, should Black play on?

*****

*****

*****

*****

*****

Black has the better minor piece and pressure against b2 and c4. White has pressure against d6 and, arguably, the less-threatened king, but that still leaves Black with at least a slight edge, according to the engines.

26...f6!?

If 26...Rfc8, White has 27.Qd2, and if 27...Nb3. then 28.f6+!, when 28...Kxf6?! leaves Black in big trouble after 29.Qh6. So Black should almost certainly play 28...Kh8, when 29.Qh6 Rg8 is completely equal, according to the engines.

However, instead of the text, Black can try 26...Nb3, when best-play may run 27.Qc3 f6 28.fxg6 hxg6 29.Bg4 Qc7 30.Qg3!? Nd4 31.Bh5!? g5 32.Qd3, with an imbalanced position, although the engines reckon Black is slightly better.

27.Qf2 Qa6 28.h4

Dragon1 reckons the position after 27...Qa6 is completely equal, while Stockfish18 awards Black the better part of equality, but both engines find it difficult to settle on White's best continuation, candidates including Kh2, Qg3, fxg6 and Qe2. The text is also probably all right, if followed up correctly.

28...Qxc4 29.Rxd6?!

Probably better is 29.fxg6 hxg6, and now 30.Rxd6 Rxd6 31.Rxd6 Nxe4 32.Rd7+ Rf7 33.Rxf7+ Kxf7 34.Qb6!, when the engines agree Black has nothing more than a draw.

29...Rxd6 30.Rxd6 Nxe4 31.Bxe4?!

Possibly better - for a long time Dragon1 prefers the text - is 31.Rd7+ Rf7 32.Rxf7+ Kxf7 33.fxg6+, when 33...hxg6 34.Qb6! transposes to the drawn line given at move 29. But the engines point out Black can improve with 33...Kxg6!, when 34.h5+ Kg7 35.Qh4 Qc1+ 36.Kh2 Nc5 gives Black the upper hand (Dragon1) or at least a slight edge (Stockfish18).

31...Qxe4 32.fxg6 hxg6

Here 33...Kxg6? is a mistake that allows a draw after, for example, 34.Qg3+ Kf7 35.Rd7+ Ke6 36.Qg7 etc.

After 32...hxg6, what is White's best hope for a draw?

*****

*****

*****

*****

*****

33.Rd7+?!

Heading for a queen-and-pawn ending, but it is completely lost, so better chances lay in, for example, 33.Rb6, although the engines agree that after 33...Qd5!?, Black is winning, eg 34.Rb4 Re8 35.Qe3 (35.Rxa4?? Qd1+) e4 36.Rd4 Qc6, when Black's extra pawn should tell, albeit after a slow and somewhat painstaking process.

33...Rf7 34.Rxf7+?

White should definitely keep rooks on.

34...Kxf7

At first glance it might seem White has hopes of a perpetual, but it turns out every White continuation either drops material or allows transposition into a trivially won pawn ending

*****

*****

*****

*****

*****

35.Kg1 Qd4 0-1

Chess Tip Of The Day 423

In the opening the pawn-majority is not felt as a major factor, but its significance grows rapidly through the middlegame, and often becomes decisive in the endgame.
Dražen Marović, Dynamic Pawn Play In Chess

Bad Neuenahr Round Seven

Spanton (1911) - Michael Negele (2149)

French Winawer

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

This is second in popularity to the mainline 5.a3.

5...Ne7

This is a normal way to the develop the black king's knight in the Winawer Variation of the French Defence, but here it has the added point that, in the event of Nb5, Black can castle before White continues with Nd6+

6.a3

Even so, the commonest continuation in ChessBase's 2026 Mega database is 6.Nb5, when 6...Bxd2+ 7.Qxd2 0-0 8.f4 a6!? (8...cxd4 is also popular) 9.Nd6 cxd4 10.Nd4, gives a game with equal chances, according to Stockfish18 and Dragon1.

6...Bxc3

How should White recapture?

*****

*****

*****

*****

*****

7.Bxc3

This is overwhelmingly most popular, but the engines fluctuate between it and 7.bxc3!?, which was played by Lajos Steiner in a win against Aron Nimzowitsch  in the 1928 stem game.

7...cxd4

The engines suggest 7...Nbc6 or 7...b6.

8.Bxd4

The engines prefer the much more successful 8.Qxd4, meeting 8...Nbc6 with 9.Qg4!?, claiming the upper hand for White, one point being 9...d4?! can be met by 10.0-0-0, and, if 10...Nf5, then 11.Nf3.

8...Nbc6 9.Nf3 0-0 10.Bd3

Preserving the bishop-pair with 10.Bc3 is recommended by the engines.

10...Nxd4 11.Nxd4 Ng6

The engines suggest 11...Nc6, or David Howell's 11...Qb6.

12.Bxg6!

Apparently a novelty, and much liked by the engines. Carles Díaz Camallonga (2467) - George Stoleriu (2349), Chess.com Blitz 2020, went 12.Nf3 Nf4!? 13.0-0 Bd7 14.Re1, with a slight edge for White, according to the engines (but 0-1, 80 moves).

How should Black recapture?

*****

*****

*****

*****

*****

12...fxg6!

The engines agree this is one of those rare occasions when capturing away from the centre is better, even though here it weakens e6.

13.0-0

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

*****

*****

*****

*****

*****

White has an unchallengeable knight on d4, and fewer pawn-islands, which combine to give the upper hand, according to Stockfish18, although Dragon1 only awards White a slight edge.

13...Bd7 14.Qd2 Rc8 15.c3 g5!?

The engines like this.

16.h3 h6 17.Rfe1 Qe8 18.Re3 Qg6 19.Qd3 Qxd3 20.Rxd3 Rf4 21.Re1 Rcf8 22.Rf3 Rxf3 23.Nxf3

How would you assess this ending?

*****

*****

*****

*****

*****

White has the better pawn-structure, and has good-knight-v-bad-bishop. But rook-and-bishop is nearly always a stronger partnership than rook-and-knight, especially with play over the whole board, so any opening up of the game could work in Black's favour, at least as long as rooks stay on the board. However, both engines give White the upper hand.

23...Kf7 24.Nd4 Ke7 25.Re3 a6!? 26.Kf1 Rf4 27.g3

Not 27.Rf3?, as both 27...Re4 28.Re3 Rxe3 29.fxe3, and 27...Bb5+!? 28.Ke1 Re4+ 29.Re3 Rxe3+ 30.fxe3, leave the game equal, according to the engines.

27...Rf8 28.Rf3 Rc8

The engines reckon White is only slightly better after 28...Rxf3!? 29.Nxf3, but it looks an unpleasant position for a human to defend.

29.Ke2 Be8 30.Re3 Bh5+ 31.f3 Rf8 32.Nb3 Be8 33.Nd4 h5!?

This is the engines' top choice.

34.Kf2 Bg6

On 34...g4 I intended 35.h4, with what the engines agree is a slight edge.

35.Ke2 Bb1 36.Nb3

The engines suggest 36.g4!?

36...b6 37.Nd4 Kd7 38.Kf2 a5!?

This is not liked by the engines. They prefer 38...g4 or 38...Bf5.

39.Ke2

White is much better after 39.Re1, according to the engines, one of their lines running 39...Bf5 40.g4 hxg4 41.hxg4 Bg6 42.Ne2!? Rh8 43.Kg3 Kc6 44.Rd1 Kc5 45.Nd4 Re8, although there is a lot of play left in the position, and there were many reasonable alternatives for both sides along the way.

39...a4!?

The engines are happy with this, even though it puts another pawn on a light square. I suppose the point is it restrains White's queenside, and anyway the pawn is hard for Black to get at.

40.Kf2 Ba2?!

MN strongly criticised this afterwards, saying the bishop was "so powerful" on the b1-h7 diagonal; the engines agree.

How should White proceed?

*****

*****

*****

*****

*****

41.Re1

Keeping the bishop off b1-h7, and freeing e3 for the white king.

41...Bc4 42.Ke3 Rh8 43.h4!?

With Black unable to create counterplay on the queenside (or in the centre), it is a good time to try to make gains on the kingside, especially with the bishop somewhat out of play.

43...gxh4

Not 43...g4? 44.fxg4 hxg4 45.Kf4!, eg 45...Rf8+ 46.Kxg4 Rf2 47.Re3 Rxb2 48.Kg5 Ra2 49.Rf3, when White's kingside play is much quicker than Black's on the queenside.

44.gxh4 Ke7 45.Rg1 Rg8?

Black had to play 45...Kf7, when the engines give White the upper hand, or at least a slight edge (Stockfish18 is less enthused than Dragon1). However, it is far from clear White can make progress, although White can keep probing, without much risk of losing.

46.Rg6

Winning a pawn, and the game.

46...Kf7 47.Rxe6 g5 48.Rf6+ Ke7 49.hxg5 Rxg5 50.Rf5 Rg2 51.Rxh5 Rxb2

Black is 'only' one pawn down, but White's e pawn wins easily

*****

*****

*****

*****

*****

52.Rh7+ Ke8 53.e6 Rb1 54.Kf4 Re1 55.Nf5!? 1-0

Chess Tip Of The Day 422

A lead in development is at its most useful in open positions or in situations where the active side can achieve a central breakthrough.

Luděk Pachman, Complete Chess Strategy 3: Play On The Wings

Olden Castling

RED MK II is 9.9 kilometres (6.2 miles), mostly on the flat, with a steepish ascent of 182 metres (597 feet) at the end, which proved of surprising interest.

The route starts by following the Ahr downstream, the river, bolstered by overnight thunderstorms, shown here flowing past part of Bad Neuenahr kurhaus, with the associated Steigenberger Hotel in the background

You eventually leave the river and head for a conical hill that used to be the site of a Roman watchtower 

The area looks prosperous, especially the village of Heppingen

Nearby Gimmigen is not quite so swanky, but claims seniority by virtue of age

Fine views on the climb

Barley(?)field

Approaching the summit of the hill, whose modern name, Landskrone, is taken from the ruins of the imperial castle of Landskron 

How the castle may have looked in its medieval heyday

View from the top

Taking a more direct route downhill allows for a small detour to a 13th century chapel whose name, Maria Hilf, literally translates as Mary Help

I could not go inside as I did not have a 50-cent coin for the front door (it would not accept my one-euro coin)

The path down - this photo does not begin to do it justice -reminded me a little of the zig-zag at naturalist Gilbert White's Selborne Hill in Hampshire

Bad Neuenahr Round Six

Udo Schneider (1780) - Spanton (1911)

QGD Normal Variation

1.d4 d5 2.Nf3 e6 3.c4 Nf6 4.e3!?

There are 14,082 examples of this move in ChessBase's 2026 Mega database, which puts it massively ahead of 4.cxd5 in popularity, but behind 4.Bg5, 4.g3 and especially 4.Nc3.

4...c5 5.a3!?

Magnus Carlsen is among those who have played this move, which scores better percentagewise in Mega26 than the more popular 5.cxd5 and 5.Nc3. The move is not defensive, or. at least, not purely defensive, as will soon become clear.

5...Nc6

The top choice of Stockfish18 and Dragon1 is 5...dxc4!?, which seems strange when White has not spent a tempo on moving the king's bishop. However, after 6.Bxc4 a6 7.0-0 b5 the engines reckon the game is equal.

6.dxc5!?

Also strange, but it is the engines' top choice, and again has been played by Carlsen, although Dragon1 for quite some time marginally prefers 6.Nc3.

6...Bxc5 7.b4 Be6 8.Bb2 0-0 9.Be2!?

The engines prefer 9.Nbd2 or 9.cxd5.

9...dxc4 10.Qxd8+ Rxd8 11.Bxc4 Bd7

Possibly a novelty. Known moves are 11...a6 and Murray Chandler's 11...Nd5.

12.0-0

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

*****

*****

*****

*****

*****

Black is ahead on development, quantitively, in having developed five pieces against White's three (and it is Black to move). But White's quality of development is better, in that White's bishops are much more active than their black counterparts. The engines reckon the position is equal, but if White can catch up in the number of pieces developed, then White may well have a decent advantage, unless Black in the meantime can catch up on quality of development.

12...Rac8 13.Bb3

The engines suggest 13.Nbd2.

13...Ne4!?

This sets a little trap.

How should White proceed?

*****

*****

*****

*****

*****

14.Nc3?

The engines suggest 14.Nbd2, but reckon Black has at least the better part of equality after both 14...Nxd2 and 14...Nd6.

Black to play and get what the engines reckon is a winning advantage

*****

*****

*****

*****

*****

14...Nxc3 15.Bxc3

US offered a draw, shortly after pressing the clock.

15...Na5! 16.bxa5

There is nothing better.

16...Rxc3 17.Rfb1

Black's activity, bishop-pair and superior pawn-structure add up to an advantage worth almost a minor piece, according to the engines

*****

*****

*****

*****

*****

17...Bf6 18.Ra2 Bc6

What should White play?

*****

*****

*****

*****

*****

19.h3?

The engines suggest 19.Rd2 Be4 20.Rbd1! Rdd3 21.Rxd3 Bxd3 22.Nd2!?, but agree Black is winning. Their alternative is 19.Nd2, but they reckon Black is again winning after, for example, 19...a6 20.h3 Kf8 21.a4 h5, although, in both cases, White has not fallen behind on material. Note that 19.Kf1?, which, like the text, tries to solve White's back-rank-mate problems, runs into 19...Bb5+ (19...Be4 is also good) 20.Ke1 Bd3.

19...Be4 20.Rbb2 Rc1+ 21.Kh2 Bxb2 22.Rxb2 Bxf3!? 23.gxf3 Rc5 0-1

Chess Tip Of The Day 421

Back towards the start of May, Blogger's feeder gadget started failing to update properly. I reported the issue and was warned it might take up a a week for "the team" to fix it. We are now in June and, although there are days when the gadget functions normally, the issue is still not resolved - the latest chess tip on the website version of Beau Chess is stuck at 419, which was posted two days ago.

If a backward pawn sits on a half-open file, it might be weak. If it isn't on a half-open file, the pawn probably won't become a significant weakness.

Jeremy Silman, The Reassess Your Chess Workbook

The Ascent Of A Man

ARGUABLY the most strenuous of the walks pushed by the local tourist board is blue MK III, which is only 7.6 kilometres (4.7 miles) in length, assuming you follow the correct route (not a given in my case), but involves 279 metres (915 feet) of ascent.

Waymarks start, somewhat belatedly, near a station of the cross

Much of the route is steadily uphill

Another station of the cross, it would seem, although maybe not part of the same series

Nearly all the walk is under trees, whose foliage provided shade on what proved a sultry, if not particularly sunny, day, and later provided protection from a series of heavy-sounding, but short in duration, showers, none of which was accompanied by forecast thunder and lightning

Lots of evidence of soil erosion

Bad Neuenahr Round Five

Spanton (1911) - Harald Hiby (1857)

French Alekhine-Chatard

1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.h4!?

How should Black meet this gambit?

*****

*****

*****

*****

*****

6...c5

Most popular in ChessBase's 2026 Mega database is 6...a6, the main idea of which is to prepare ...c5 without allowing the reply Nb5. Stockfish18 and Dragon1 recommend accepting the gambit, their main line running 6...Bxg5 7.hxg5 Qxg5 8.Nh3 (Albin, in the stem game, played 8.Nf3) Qh4!? (a rare continuation) 9.Nb5!? (9.g3 is more common)  Kd8 10.Qd2 a6 11.Nc3 c5 12.dxc5!? Nc6 13.0-0-0, and now they suggest an apparent novelty in 13...Kc7!? (the known move is 13...h6), albeit awarding White a slight edge.

7.Bxe7

The immediate 7.Nb5?!. although quite popular (300 examples of the move in Mega26), can be met by 7...f6!, eg 8.exf6 Nxf6, when the knight on b5 looks somewhat foolish. Alternatively, Dragon1's apparent-novelty, 8.Qh5+!?, continuing 8...g6 9.Qg7 fxg5 10.h5!?, can be met by 10...Bf8!? 11.Nd6+ Bxd6 12.exd6 Rg8!?, after which the engines agree White does not have enough for a piece.

7...Kxe7!?

Korchnoi and Short are among many who have played the obvious 7...Qxe7, but 8.Nb5 leaves Black with a difficult and unpleasant choice.

How should White proceed?

*****

*****

*****

*****

*****

8.Qg4

The main line in Mega26 runs 8.f4 cxd4 9.Qxd4 Nc6 10.Qd2, with the upper hand for White, according to the engines. However, even in this short variation, there are major alternatives, with the engines agreeing one improvement for Black is 9...Qb6!? Stockfish18 likes 8.Qd2 or 8.dxc5, while Dragon1 prefers the popular 8.f4 or 8.Bb5 - testament to just how unclear these lines are.

8...Kf8

This is the main move, but the engines also suggest 8...Nc6, when they reckon White's best reply is 9.dxc5!?

9.Nf3 Nc6

Also interesting is 9...cxd4 10.Nb5!? Qa5+!? (another apparent-novelty) 11.Kd1!?

10.0-0-0!?

Possibly better is 10.Qf4 or 10.dxc5!?

10...cxd4 11.Nxd4!?

This is the only move in Mega26, but the engines suggest 11.Nb5 or 11.Rxd4!?, eg 11.Nb5 Ndxe5 12.Nxe5 Nxe5 13.Qxd4 a6!? 14.Qxe5 axb5 15.Kb1, when White may have enough for a pawn, or 11.Rxd4!? Ndxe5!? (11...Nxd4 12.Qxd4 gives White decent compensation for the exchange, according to the engines) 12.Nxe5 Nxe5 13.Qg3 Nc6 14.Rd2, when the engines reckon Black has the upper hand, but the position is far from easy for Black to play, one example of this being that after 14.Rd2 the engines agree Black has to find 14...h5!? to keep an advantage.

11...Ndxe5 12.Qg3 a6

This and 12...Qf6?! are known moves, but the engines do not like either, claiming Black has the upper hand after both 12...h5!? and 12...Nxd4.

13.h5!?

Seemingly a novelty, and perhaps a slight improvement on the known 13.f4.

13...h6 14.f4 Nxd4

Which knight should White capture?

*****

*****

*****

*****

*****

15.fxe5?

The engines reckon White has at least full compensation for a pawn after 15.Rxd4, eg 15...Nc6 16.Rd1 Qc7 17.Bd3 Nb4 18.Be2 Bd7 19.a3!? Nc6 20.Qf2, when Black remains a pawn up, but the king's rook is out of play.

15...Nf5 16.Qf2 Qg5+ 17.Kb1 Qe3

Not 17...Ng3? 18.Rh3, but 17...Ne3!? may be playable.

18.Qxe3 Nxe3 19.Rd3?

Necessary is 19.Re1, although the engines agree 19...Nxf1 20.Rhxf1 leaves White without enough for a pawn.

19...Nc4

What is White's best try?

*****

*****

*****

*****

*****

20.Nxd5!

Otherwise e5 drops, after which, with queens off the board, White is simply two pawns down.

How should Black respond?

*****

*****

*****

*****

*****

20...exd5

Almost certainly even stronger is 20...Nxe5 21.Nc7 Nxd3 22.Nxa8, when Black is a pawn up, and should soon get his rook into play.

21.Rxd5 Bxe6

White threatened 22.Rd8+ and 23.Rxh8.

22.Bxc4 Bxe6

This is obvious, at first glance, but the engines point out the superiority of 22...b5! 23.Bb3 Bxd5 24.Bxd5, and either 24...Rd8 or 24...Re8.

23.Bxd5

How would you assess this ending?

*****

*****

*****

*****

*****

White has a pawn for the exchange, the ability to stabilise the bishop in the centre with c4, and pressure against b7 and f7. Nevertheless, rooks are endgame pieces per excellence, and the engines reckon chances are equal.

23...Re8

Best, according to the engines, is 23...Ra7!?, intending a quick ...b5.

24.Bxb7 a5

The engines agree this is better than allowing White three connected passers with 24...Rxe5 25.Bxa6, but Dragon1, especially, fluctuates between the text and 24...Re6.

25.Re1 f6?

This is a serious mistake, according to the engines, as it allows White a passed pawn that can be easily protected by the bishop.

26.e6 Ke7 27.Bd5 Rhg8 28.Re3!? g5

White to play and win

*****

*****

*****

*****

*****

29.hxg6!

The only winning move, according to the engines, although they agree White is also better after 29.Ra3. However, they reckon 29.g4? f5! gives Black what is almost certainly a winning position, eg 30.gxf5 g4 31,Rg3 Rg5 32.Be4 Rd8 33.c3 Rd2 34.Kc1 Rf2 35.Bc2 Rxh5 36.Rxg4 Rh1+ 37.Bd1 Rxd5, when they reckon Black's advantage is roughly worth a minor piece.

29...Rxg6 30.Ra3 Rg5 31.c4

Also good is 31.Rxa5.

31...Rc8!?

Setting a 'trap'

*****

*****

*****

*****

*****

32.Rxa5!?

Except that the trap does not work. However, there was no way to save the a5 pawn.

32...Rxc4 33.Ra7+ 1-0