World Team 65+ Round One

Bo.	19	France - 2	Rtg	-	46	England - 5	Rtg	0 : 0
19.1		Boulard, Eric	2228	-		Freeman, Richard C P	1897
19.2	FM	Fayard, Alain	2181	-	CM	Stokes, Michael	1861
19.3		Theon, Pierre	2046	-		Spanton, Tim R	1908
19.4		Lebret, Jean-Michel	2016	-		Marshall, Michael	1859

Pierre Theon (2046) - Spanton (1908)

QGD Tarrasch

1.d4 d5 2.c4 e6 3.Nc3 c5 4.cxd5 exd5 5.dxc5!?

This is very much a sideline, compared with 5.Nf3, but there are 668 examples of the move in ChessBase's 2026 Mega database, and it has been played by Carlsen and Kramnik.

5...d4 6.Ne4

How should Black proceed?

*****

*****

*****

*****

*****

6...Nf6!?

The stem game, Amos Burn - Siegbert Tarrasch, German Chess Congress (Breslau*) 1889, went 6...Bxc5?! 7.Nxc5 Qa5+ 8.Bd2 Qxc5 9.Rc1 Qf5 10.Qa4+ Nc6 11.Nf3, with the upper hand for White, according to Stockfish17.1 and Dragon1, although the game was eventually drawn. The engines prefer the text.

7.Nd6+

Winning the bishop-pair, but falling behind in development. The engines reckon 7.Nxf6+ Qxf6 8.b4 Nc6 9.Bb2!? gives equal chances. They also like 7.Bg5, when they agree Black's best reply is 7...Bf5!? (7...Be7 8.Bxf6!), eg 8.Nxf6+ gxf6 9.e4!? Bxe4 10.Qe2!? Qd5!? 11.Bxf6 d3 12.Qg4! Nc6 13.Bxh8 d2+ 14.Kd1 h5! 15.Qe2 0-0-0, when they give Black complete equality, despite Black being rook and pawn down.

7...Bxd6 8.cxd6 Qxd6 9.Nf3 Nc6 10.e3 Bg4

Also very interesting is 10...Qb4+!? 11.Qd2 Qxb2 12.Qc1!? Qxc1+ 13.Rxc1 dxe3 14.Bxde3 0-0, when the engines reckon Black's extra pawn is worth more than White's bishops.

11.exd4!?

Possibly 11.Be2 is better.

11...0-0-0

The engines agree 11...Bxf3 gives Black the upper hand, eg 12.gxf3 0-0!? 13.Be3 Nd5!? 14.Qb3 Rfe8 15.0-0-0!? a5, when Black's king looks decidedly safer.

12.Be3 Rhe8 13.Be2 Bxf3?!

Black has at least a slight edge after 13...Nd5, according to the engines.

14.Bxf3 Qb4+!

White will have no problems if allowed to castle.

15.Qd2

Also seemingly playable is 15.Kf1!? Nxd4 16.g3.

15...Qxd2+ 16.Kxd2 Nxd4 17.Rac1+ Kb8

This is almost certainly better than 17...Nc6+ 18.Kc2.

18.Bxd4 Rxd4+ 19.Kc3 Red8

How would you assess this late-middlegame?

*****

*****

*****

*****

*****

White has the better minor piece for working with rooks, especially on such an open board, but Black's coordinated rooks and already-tucked-away king give full compensation, according to the engines.

20.Rhe1!?

Not 20.Rhd1? as Black has 20...Ne4+, eg 21.Kc2 (21.Bxe4?! Rxd1) Rc8+ 22.Kb1 Nd2+ 23.Ka1 Rxc1+ 24.Rxc1 Nxf3 25.gxf3 Rd3, with much the better of a rook-and-pawn ending.

20...Rd3+ 21.Kc2 Rd2+ 22.Kb1 Rxf2

Black's initiative has netted a pawn, but now White gets to apply pressure

*****

*****

*****

*****

*****
23.Re7 Rd7 24.Bg4!?

How should Black respond?

*****

*****

*****

*****

*****

24...Rd8

The only move, as both 24...Nxg4?? and 24...Rdd2?? allow a quick mate.

25.Bf3 Rd7

Again forced.

26.Rxd7!? Nxd7

How would you assess this rook-and-minor-piece ending?

*****

*****

*****

*****

*****

White's rook-and-bishop combination give full compensation for being a pawn down, according to the engines.

27.Re1 Nc5 28.Re7 b6 29.Rxf7 Nd3 30.Rb7+!? Kc8 31.Rxg7 Rxb2+ 32.Ka1 Rc2 33.a3 a5!?

This may be too slow. Perhaps best is 33...Ne1!?, when 34.Kb1 Rf2 35.Be4 Re2 36.Bd5 Kd8 is completely equal, according to Stockfish17.1, although Dragon1 gives White a slight edge.

34.Be4 Rc1+?!

Almost certainly better is 34...Rc3.

35.Ka2 Nc5 36.Bxh7

White is a pawn up in a position that favours White anyway, thanks to a bishop being almost always much better than a knight when there are rival pawn-majorities

*****

*****

*****

*****

*****

36...Rf1 37.Kb2 Rf2+ 38.Kc3 Ra2 39.Kc4?!

Probably stronger is starting to push the passed h pawn, either immediately, or after first playing Bf5+.

39...Rxa3

The extra pawn has gone, White's king is more active than Black's, and bishop-v-knight still greatly favours White

*****

*****

*****

*****

*****

The only thing in Black's favour is that the black pawns are more advanced.

40.Kd5 Rc3?

40...a4 almost certainly provides better drawing chances, eg 41.h4 Ra1 42.h5 a3 43.Bf5+ Kb8!, when the engines give White only the tiniest of edges. However, there are many alternatives, and the lines are very tricky.

41.Bf5+ Kb8 42.h4 Nb7

Black queens first in the pawn-race, ie 42...a4 43.h5 a3 44.h6 a2 47.h7 a1=Q, but then White queens with mate.

43.h5 Rc5+ 44.Ke6 Nd8+?

Better, but still losing, is 44...a4, eg 45.Rg3 b5 46.Rh3 Rc6+ 47.Ke7 Rc7+ 48.Kf8 Nd6 49.Bg6 Nc4 50.h6 Rc8+ 51.Kg7 Rc7+ 52.Kf6 Rc6+ 53.Kg5 Rc5+ 54.Bf5 Nd6 55.g4, when the only way to stop the h pawn queening is to give up the knight.

45.Kf6 a4 46.h6 a3 47.Bb1

White wins the pawn-race after 47.h7? a2 48.h8=Q, but this time Black queens with check, and so draws.

47...Rc6+ 48.Ke7 Rxh6

Forced.

49.Kxd8 Rh1 50.Ba2

Not 50.Be4?? Rh8+ 51.Kd7 a2 etc.

50...Rd1+ 51.Rd7 Rh1 52.Bd5 Rh8+ 53.Ke7 Kc8 54.Rd6 Rh7+ 55.Ke6 Rg7 56.Rc6+ Kd8 57.Kf6 Ra7 58.Ba2 Rd7!?

This speeds the end, but Black is lost anyway.

59.Be6 a2!? 60.Bxa2 Rd2 61.Be6 Rf2+ 62.Bf5 1-0

Team Result

Boulard (2228) = Freeman (1897)

Fayard (2181) 1-0 Stokes (1861)

Theon (2046) 1-0 Spanton (1908)

Lebret (2016) = Marshall (1859)

France 2 3-1 England 5

*Now Wrocław in the Lower Silesian region of Poland.

Getting There (Durrës)

CAUGHT British Airways' 13.50 from Heathrow to the Albanian capital Tirana, where three coaches were waiting to ferry players to the seaside resort of Durrës, which is where the world senior team championships are being held.

I was checked in and down to dinner before 20:00 central European time, later enjoying a poolside drink with some of the other England players.

View from my balcony

Play starts tomorrow at 15:00, with 54 teams in the 65+, where I am on board three (of four) for England 5.

Our team's average rating of 1881 means, barring late changes, we should face France 2 (average rating 2118) in round one.

There is also a 36-team 50+.

Tricky Brain Exercise

SAW this serial number on a Central Line carriage: 92136.

As usual, each number should be used once, and once only, and must be added, subtracted, multiplied or divided to make a balanced equation.
There is a perfect solution, ie one that uses the numbers in the order they appear, but it is far from obvious.
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****
*****

*****
*****
My solution: 9 ÷ (2 + 1) + 3 = 6

Balkans Bound

AM making final preparations for flying to Albania tomorrow for the world 65+ senior team championships, where I am on board three (of four) for England 5.

There are nine afternoon rounds over 10 days, including a free day after round five, with a time limit of 40 moves in 90 minutes, a further 30 minutes to finish, and a 30-second increment throughout.

The 54 teams are headed by German club side Lasker Schachstiftung GK, England, Iceland, Montenegro, France and Russia.

There is also a 50+ tournament with 36 teams, where the top three seeds are USA, England and Kazakhstan.

Venue hotel - Grand Blue FAFA, Durrës

Minatures 34

IN this occasional series I am going through my decisive games of 20 moves or fewer.

Bertrand Barlow (173 BCF) - Spanton (147 BCF)

Hastings 3rd Week A 1991

King's Gambit

1.e4 e5 2.f4 Bc5 3.Nf3 d5!?

How should White respond to Black's unusual third move?

*****

*****

*****

*****

*****

4.exd5

This is the commonest continuation in ChessBase's 2026 Mega database, but Dragon1 prefers 4.Nxe5, while Stockfish17.1 suggests 4.Nc3!?

4...e4 5.d4!?

More popular are 5.Qe2 and especially 5.Ne5, but the engines like the text, which goes back to at least 1839.

5...exd3 6.Qxd3!? Nf6 7.Nc3 0-0 8.Bd2 Ng4?

The obvious 8...Re8+ is better.

9.0-0-0?

White has a large advantage after 9.Ng5, eg 9...Re8+ 10.Be2 g6 11.Nce4 Bf5 12.Bxg4 Bxe4 13.Nxe4 Qh4+ 14.Kf1! Rxe4 15.g3 Qe7 16.c3!?

9...Nf2 10.Qc4

What should Black play?

*****

*****

*****

*****

*****

10...Nxd1?

Best is 10...b6, protecting the loose bishop while maintaining a fork of the white rooks.

11.Qxc5 Nd7 12.Qd4 Nxc3 13.Bxc3 Nf6

White's bishop-pair and extra pawn give full compensation for the exchange, according to the engines

*****

*****

*****

*****

*****

14.Bc4 Qd6 15.Re1 Bd7 16.Ne5 a5?!

Black probably needs to disrupt White's build-up, perhaps by 16...b5, and if 17.Bb3, then 17...a5 18.a3 Rfb8, with equal chances, according to the engines, which reckon 17.Be2!? is a promising alternative.

17.Re3 b5?

Too late. 17...R(either)d8 keeps the game going by allowing the black queen to keep defending f6.

18.Nxd7 Qxd7 19.Qxf6! 1-0

If 19...gxf6?, then 20.Rg3+ leads to mate, while 19...Qg4 20.Qd4 f6 21.Bxb5 leaves White with an advantage worth more than a rook, according to the engines.

LESSON: when enemy pieces are lining up against your king's position, precise calculation is usually required.

Miniatures 33

IN this occasional series I am going through my decisive games of 20 moves or fewer.

Spanton (147 BCF) - N Foster (200 BCF)

Barbican (London) Rapid1990

Nimzowitsch Defence

1.e4 Nc6 2.d4 d5 3.e5 Bf5 4.Nf3 e6 5.g3!?

The main continuations in ChessBase's 2026 Mega database are 5.c3 and especially 5.Bd3.

5...f6!?

This immediate attack on the white centre is a rare continuation but is liked by Stockfish17.1 and Dragon1.

6.exf6 Qxf6!?

How should White proceed?

*****

*****

*****

*****

*****

7.Bg5?

The engines suggest 7.c3 or 7.Bg2, with what they reckon is equality.

7...Bxc2! 8.Qe2

Or 8.Bxf6 Bxd1 9.Kxd1 (forced) gxf6, when Black is a sound pawn up.

8...Qf5 9.Nc3?

Better is 9,Nbd2, developing and protecting the king's knight.

9...Bb4?

Missing the strength of 9...Be4.

10.h4?

The bishop should be moved, probably to e3.

10...Be4 11.Bg2 Bxf3 12.Qxf3 Qxf3 13.Bxf3 Nxd4

Winning a second healthy pawn.

14.Bd1 Nf6 15.0-0 0-0 16.Rc1 c6 17.a3 Bd6 18.Be3 Nf5 19.Re1?! 0-1

LESSON: success against a much higher-rated opponent is made infinitely harder if basic tactical mistakes are committed while still in the opening.

Thoughts On Daventry V

IN round seven, after taking a halfpoint bye in the morning, I was Black against David Flynn (1867 ECF/1816 Fide).

The game featured a sharp line from the Hyper-Accelerated Dragon in the Sicilian, and I think the line is worth a closer look.

It began 1.e4 c5 2.Nc3.

This usually means a Closed Sicilian, but as has been pointed out by many chess instructors, White is not committed to keeping the position closed (although White has ruled out a Maróczy Bind).

I played 2...g6, partly because I had seen in ChessBase's 2026 Mega database that after 2...Nc6 DF 'always' (seven games out of seven) plays 3.Bb5.

Now 3.Bb5 is not something Black should fear, although my record against it is a disappointing +1=0-3, including two losses to much lower-rated opposition.

But it seemed to make sense to deprive my opponent of the ability to play his 'favourite' variation.

He could have stayed in another type of Closed Sicilian with 3.g3, but instead opened the game with 3.d4 cxd4 4.Qxd4.

After the further moves 4...Nf6 5.Nf3 we reached a position occurring 2,722 times in Mega26.

The usual Hyper-Accelerated move-order is 1.e4 c5 2.Nf3 g6 3.d4 cxd4 4.Qxd4 Nf6 5.Nc3

*****

*****

*****

*****

*****

Following 5...Nc6 6.Qa4 the threat of 7.e5, forcing the black king's knight to move, more-or-less obliges Black to play 6...d6.

But 7.e5 comes anyway, and after 7...dxe5 8.Nxe5 the game remains sharp.

The position after 8.Nxe5 occurs 1,784 times in Mega26

*****

*****

*****

*****

*****

On three occasions reaching this position I have continued 8...Bd7, which is easily the main move in Mega26.

The drawback is White can immediately win the bishop-pair with 9.Nxd7, and, although I have drawn this position in correspondence play, I lost with it both times over-the-board, albeit against much stronger opposition.

This time I played the gambit move 8...Bg7!?, expecting 9.Nxc6 bxc6 10.Qxc6+ Bd7, which occurred in Hanna Kyrjebo (1831) - Spanton (2002), Hastings 2014-5.

That game continued 11.Qf3 0-0 12.Bd3 Qb6 13.0-0 Bc6 14.Qe3 Qb7 15.f3 Rad8, with Black enjoying full compensation for a pawn, according to the engines (0-1, 44 moves).

However DF surprised me with 9.Bb5.

Of course this should not have been a surprise, especially as it is the most popular continuation in Mega26.

One of its points is that unless Black reverts to giving up the bishop-pair with 9...Bd7, Black will either have to give up a second pawn or allow some simplification.

The game continued 9...0-0 10.Nxc6 bxc6 11.Bxc6

How should Black proceed?

*****

*****

*****

*****

*****

I chose 11...Rb8!?, offering a second pawn, and the continuation can be seen here.

It is the main move in Mega26, but the engines marginally prefer the alternative, 11...Bd7.

After 12.Bxd7 they reckon Black should avoid further simplification, which would occur with 12...Qxd7?! 13.Qxd7, and instead play 12...Nxd7, which looks a little retrograde, but the knight will gain a tempo on the white queen with ...Nb6 or ...Nc5, and Black's dark-square bishop has been unmasked.

The engines continue 13.0-0 Bxc3!? 14.bxc3 Nb6, claiming Black has full compensation (Stockfish17.1) or is no more than slightly worse (Dragon1).

Indeed the engines reckon White should meet 11...Bd7 with 12.0-0, when, to avoid serious disadvantage, the engines agree Black has to play 12...Rc8.

Now 13.Bxd7 should be met, they reckon, with 13...Qxd7!?, not fearing the simplification that occurs with 14.Qxd7 Nxd7.

How should White proceed?

*****

*****

*****

*****

*****

15.Bd2 might seem natural, but is perhaps too passive. After 15...Nb6 (threatening ...Nc4) 16.Rfd1 Rfe8 17.Be1!? Nc4, the engines give Black full equality.

Instead they suggest 15.Nd5!?, when 15...Rfe8?! 16.c3 is good for White.

However 15...Rxc2 16.Nxe7+ Kh8 leaves Black with great activity. A plausible continuation runs 17.Rd1 Ne5!? 18.Nd5!? Rd8 19.Rb1 h6!? 20.h3!? Nc4 21.Ne3 Rxd1+ 22.Nxd1, after which Black remains a pawn down, but has enough compensation (Stockfish17.1) or is only slightly worse (Dragon1).

The conclusion seems to be that the pawn-sacrifice 8...Bg7!? is a dynamic and fully viable alternative to giving up the bishop-pair with 8...Bd7, even though White retains a slight edge in many lines.

There is certainly much left to explore in the resulting positions.

Thoughts On Daventry IV

THE following position was reached after 27 moves in round five, in which I was White against Iranian Hooman Honarvarmahalati (1865 ECF/1837 Fide).

Black has just played 27...Bf7-g6

*****

*****

*****

*****

*****

The game continued 28.Bd6+ Kc8 29.Bxe5 Bxg3 and was eventually drawn.

Stockfish17.1 and Dragon1 suggest 28.Re1!?, which, naturally, I considered, but rejected for being too passive.

Here I want to try to to discover what is behind the engines' surprising - to me - recommendation, by seeing how the game could develop after their suggestion.

Their main line runs something like 28...h4 29.Kb2 Kd7 30.Re2 Ke6 31.Be1 Bh5 32.Re3 g5 33.g3!? Bg6 34.g4 Rd8.

Position after 34...Rd8

*****

*****

*****

*****

*****

As far as I can see, White has not made progress, and the engines seem to agree. Their previous evaluation of White either winning or having the upper hand has deteriorated to White having the upper hand or just a slight edge.

Their main line continues more-or-less (they do not always completely agree on the best move): 35.Ba5 Rf8 36.Bc7 Rc8 37.Bb6 Rf8 38.a4 Kd7 39.Rd3+ Ke8 40.Rd6+!? Bxe4 41.Re6+ Kd7 42.Rxe5 Bg2 43.Rxg5 Bxh3 44.Rg7+ Ke6 45.Rh7 Bxg4 46.Rxh4 Bf5.

Position after 46...Bf5

*****

*****

*****

*****

*****

Stockfish17.1 gives White a slight edge, although Dragon1 reckons White has the upper hand.

With pawns on only one side of the board, and opposite-colour bishops, I suspect most games between strong players would have been agreed drawn by now, and even club players as White would only play on more in hope than expectation.
The pleasing thing from my point of view is that while 28.Re1!? may, objectively, be a little stronger than my 28.Bd6+, I was probably right in regarding it as too passive to give much hope of victory.

Thoughts On Daventry III

MY round-four game, in which I was Black against China's Ran Li, featured a new-to-me move in the Vienna.

After 1.e4 e5 2.Nc3 Nf6 I had previously faced and/or played what I thought was every reasonable move: f4, g3, Bc4, a3!? and Nf3.

However my opponent surprised me when she played 3.d4!?, explaining afterwards that, although the move might not be found in books, there has been quite a lot posted about it online.

ChessBase's 2026 Mega database shows the move goes back to at least a simul given by Isidor Gunsberg in Leeds in 1887.

Not only that, but the move's score in Mega26 of 59% is very good, although in games between rated players White only performs seven elo above White's average rating for those games.

Part of the move's success must be due to its surprise value.

The day after our game I was having breakfast in the same cafe as another U2000 competitor, Matthew Wilson.

Naturally we got to talking about our upcoming pairings and, when he learnt I had already faced his opponent, and that I had written the game up at my blog, he spent the next 30 minutes feverishly preparing.

Here is the result.

Ran Li (1701 ECF/1699 Fide) - Matthew Wilson (1858 ECF/1810 Fide)

Vienna Game

1.e4 e5 2.Nc3 Nf6 3.d4!? exd4 4.Qxd4 Nc6 5.Qc4!?

This keeps play in independent Vienna lines, whereas 5.Qe3 transposes to the Centre Game.

5...Bb4

I played 5...Be7, which was also Alexei Shirov's choice in the position in a blitz game, but the text is most popular, and is recommended by Stockfish17.1 and Dragon1.

6.Bd2 d6 7.0-0-0 0-0

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

*****

*****

*****

*****

*****

White has more space in the centre, but the queen will likely be harassed by the black light-square bishop. Stockfish17.1 gives Black a slight edge, although Dragon1 reckons the game is equal.

The same position was reached in my game, except that the black dark-square bishop was on e7 rather than b4. In that case both engines call chances equal.

8.f3 Be6 9.Qe2 a5 10.g4?!

The engines strongly dislike this, suggesting the paradoxical 10.a3!?

Can you see how Black takes advantage of White's last move?

*****

*****

*****

*****

*****
10...Ne5

The engines give 10...Bxc3! 11.Bxc3 Nb4, reckoning White has nothing better than giving up the a2 pawn as 12.a3 Na2+ 13.Kd2 Nxc3 is deeply unpleasant whichever way White captures on c3, eg 14.Kxc3 d5!? 15.g5 Nh5!? 16.Qe3 c5 17.exd5 Bxd5 18.Bc4 Re8! 19.Qxc5 Nf4 20.Rxd5!? Nxd5+ 21.Qxd5 Qc7 22.Nh3 Rac8 23..Rd1 Re5 24.Qd4 b5, or 14.bxc3 Nd7 15.Qe3 d5 16.e5 f6!? 17.exf6 Qxf6!? with a very strong attack.

11.g5 Nfd7 12.f4?!

This looks like an oversight, but the engines are not too unhappy with it.

12...Bg4 13.Qg2 Bxd1 14.Nxd1 Bxd2+ 15.Qxd2 Nc6

White has lost/sacrificed the exchange, but White's kingside attack is probably ahead of Black's on the queenside

*****

*****

*****

*****
*****

16.h4 Nc5 17.Nc3 Qd7

The engines prefer 17...Re8.

18.h5 f5?

Opening a diagonal to the black king can hardly be right. However, the engines agree Black needs to do something about White's attack, suggesting 18...Rae8!?

19.Bh3?

White is winning, according to the engines, after 19.Bc4+ Kh8 20.Nf3.

19...Ne7 20.Nge2 Qc6?

The queen is a target on c6. Perhaps ...Rae8 is still best, although the engines prefer White.

21.Nd4 Nxe4?

The queen should be moved, albeit White is winning, according to the engines.

22.Nxc6 Nxd2 23.Nxe7+ Kf7 24.Nxf5 Nc4

Instead of having bishop for rook, White has emerged with bishop and knight for rook

*****

*****

*****

*****

*****

25.Nd5 c6 26.g6+! hxg6 27.hxg6+ Kxg6 28.Kg5?

Black has to play 28.Kh7, but would be the equivalent of more than a rook down, according to the engines.

28...Rg5#

Thoughts On Daventry II

AFTER taking a halfpoint bye in round two, I was positionally outplayed in round three by a girl aged nine or 10 (probably the former - she is listed by the English Chess Federation's as England's second-best under-10-year-old).
Tara Tamilselvan (1879 ECF/1820 Fide) went on to finish tied for second in the U2000, scoring +4=3-0 with a rating performance over 2050.

That may not make her a prodigy, but she has plenty of juvenile years ahead of her in which to get even better, and of course there are many adult players who can only dream of being so 'weak'.

She beat me positionally, but in the following round won against a Ukrainian in the tactical style associated with juniors.

Tara Tamilselvan (1879 ECF/1820 Fide) - Dmytro Chernov (1920 ECF/no Fide)

Pribyl Defence

1.Nf3 d6 2.e4 Nf6 3.Nc3 c6 4.d4 Bg4

This is easily the most popular move in ChessBase's 2026 Mega database, although Stockfish17.1 and Dragon1 prefer transposing into a Pirc with 4...g6.

5.h3 Bh5 6.Bd3 Qc7?!

The engines strongly dislike this, claiming White is now winning.

How should White proceed?

*****

*****

*****

*****

*****

7.Bf4

This seems to be a novelty. The engines' main line runs 7.g4!? Bg6 8.e5 Bxd3 9.Qxd3 Nd5 10.e6!? f6 11.0-0, assessing White as having a positionally won game.

7...Nbd7 8.g4 Bg6 9.e5

The engines prefer 9.Be3!? or 9.g5.

9...dxe5 10.Nxe5 Qa5 11.Qe2

White has the upper hand after 11.Nc4, according to the engines.

11...Nd5?!

11...Nxe5 equalises (Stockfish17.1) or cuts White's advantage to a slight edge (Dragon1).

12.Nc4! Nxf4

More-or-less forced as 12...Qd8?? allows 13.Nd6#.

13.Nd6+ Kd8 14.Nxb7+ Kc7 15.Nxa5 Nxe2 16.Kxe2 Rb8 17.Nb3 e5

White has emerged from the tactics a pawn up, so Black tries to open lines in the centre to get at the white king

*****

*****

*****

*****

*****

18.d5 Bb4 19.dxc6 Bxd3+ 20.cxd3 Kxc6

The engines prefer closing the c file with 20...Bxc3!? before capturing on c6.

21.Rac1 h5?

The black king is too vulnerable to allow time for this.

22.Ne4+ Kb6 23.Rc4 Be7 24.Rhc1 Rbd8?!

Almost certainly better is 24...Rh6, preventing White's next.

25.Rc6+ Kb7 26.Na5+ Kb8 27.Rc7 Rc8!?

White was threatening to deliver mate, starting with 28.Rb7+ Ka8 29.Rcc7.

28.Rb7+ Ka8 29.Rxc8+ Rxc8 30.Rxd7 Bb4 31.Nc4 1-0