Sunday 31 July 2022

Brașov Round Nine

TODAY in the final round I played another Romanian junior (born 2008).

Razvan-Alexandru Panait (1411) - Spanton (1889)
French Rubinstein
1.e4 e6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6!?
The main moves are 4...Nd7 and 4...Bd7 but there are 5,043 examples of the text in ChessBase's 2022 Mega database.
5.Nxf6+ Qxf6 6.Nf3 h6 7.Be3!?
The mainline runs 7.Bd3 Bd6 8.0-0 0-0 9.Qe2, when White has the upper hand, according to Stockfish15 and Komodo13.02.
7...Bd6 8.Qd2!?
Hinting at quick queenside castling.
8...0-0 9.Bd3 Nc6 10.c3 Re8!?
Possibly new. 10...e5 is known from three games in Mega22.
11.0-0-0 e5 12.Qc2?!
Capturing on e5 or playing 12.d5 are probably better.
12...exd4 13.Nxd4 Nxd4?!
Almost certainly better is 13...Ne5.
14.Bxd4 Bf4+?!
Grabbing a pawn by 14...Qg5+ 15.Kb1 Qxg2?? loses on the spot to 15...Rdg1 etc but the engines reckon 15...Be6 gives Black at least an even game. The text spends a tempo helping the white king to safety.
15.Kb1 Be5 16.Rhe1 Re7??
It was necessary to play 16...Re6 or 16...Bg4 17.f3 Bxd4, although in both lines Black is struggling for equality.
17.Bh7+?
This is good enough for a large advantage, but winning is 17.Qe2.
17...Kf8
Not 17...Kh8?? 18.Rxe5!? (18.Bxe5 also wins) Rxe5 19.Bxe5, and if 19...Qxe5? then 20.Rd8+ Qe8 21.Rxe8#.
18.Bc5 Bd6 19.Rxe7 Qxe7
Not 19...Bxe7?? 20.Rd8#.
20.Bxd6 cxd6
How large is White's advantage?
*****
*****
*****
*****
*****
The engines reckon White is the equivalent of about three-quarters of a pawn ahead.
21.Be4 Be6 22.Bf3 Kg8 23.Qe4
How should Black proceed?
*****
*****
*****
*****
*****
23...Rb8
The engines give 23...Re8!?, and if 24.Qxb7 then 24...Bf5+ 25.Ka1 (25.Kc1? Qg5+ 26.Rd2 Bg6) Qe1!, when the game is dead-equal, according to Komodo13.02 although Stockfish15 gives White a slight edge.
24.Qb4
Also strong is the simple 24.Qd4.
24...Bf5+ 25.Ka1 Bc2!? 26.Qxd6 Qxd6!?
Heading for a rook-and-pawn ending. The engines prefer 26...Qe8.
27.Rxd6 Re8 28.b3 Be4 29.Bxe4 Rxe4
How would you assess the ending?
*****
*****
*****
*****
*****
White is a pawn up and it is White to move, but the black king is slightly better placed than its opposite number. Stockfish15 reckons White is winning; the more-conservative Komodo13.02 gives White 'only' the upper hand.
30.Rd8+ Kh7 31.Rd7!?
Slower, but possibly better, is 31.Rd2.
31...Re1+ 32.Kb2 Re2+ 33.Ka3 Rxf2 34.Rxb7 f5?
Best is 34...a6, according to the engines, and they also prefer 34...Rxg2?!
35.g3?
White should get on with it, eg 35.Rxa7 Rxg2 36.b4 etc.
35...Rxh2 36.Rxa7 Rh3 37.Kb4
Stockfish15 reckons 37.b4!? is best but Komodo13.02 much prefers the text and 37.c4.
37...Rxg3 38.c4 f4?
Stockfish15 reckons 38...h5 and 38...Rg6 more-or-less equalise; Komodo13.02 disagrees, but both engines reckon the text is wrong.
39.c5!?
Stockfish15 reckons this and 39.Rf7 win, but Komodo13.02 much prefers the latter.
39...f3?!
39...Re3 was the best chance, according to the engines, although 40.Rf7 seems strong.
40.Rf7 Kg8 41.Rf4 g5
White to play and win
*****
*****
*****
*****
*****
42.c6!
Also probably good enough is 42.Rf5 Rg4+ 43.Kc3, but 43.Ka5? only draws and 43.Kb4?? loses.
42...gxf4 43.c7 f2 44.c8=Q+ Kg7 45.Qc7+ Kg6 1-0

Castle Dracula

SURELY no trip to Transylvania could be complete without visiting Count Dracula's mountain fortress?
And since yesterday was a free day in the chess tournament, I caught a bus from Brașov to Bran Castle, which is less than 20 miles away.
There is apparently no proof author Bram Stoker based his bloodthirsty character on Vlad the Impaler, also known as Vlad Dracula - dracul being Romanian for the Devil.
But it is certainly a reasonable supposition that the 15th century Viovode of Wallachia, infamous for impaling prisoners on stakes, was at least a partial source.
Then again Hungarian countess Elizabeth Báthory, a convicted serial killer rumoured - probably unfairly - to have bathed in the blood of her virginal victims, is also thought to have played a part in Stoker's imaginings.
Unfortunately Vlad never owned Bran Castle, which during his lifetime was in the hands of his political enemies, and there is no evidence the countess ever set foot there either.
But none of that bothers the owners of Bran Castle, and it was fun to play along for the sake of an interesting half-day out.
Bran Castle has an interesting history in its own right.
It may have been the site of a wooden fort built by the Teutonic Knights in 1212 and destroyed by Mongols in 1242.
Permission for a stone castle was granted by Louis the Great, King of Hungary, Croatia and Poland, in 1377 to Germans, known as Transylvanian Saxons, who lived in what was then Kronstadt but is now Brașov.
The castle was used as a defence against the Ottoman Turks and became a customs point, with the village of Bran - now town-size - growing up in its shadow.














Saturday 30 July 2022

City Life

THE city of Brașov - not to be confused with the nearby alpine resort of Poiana Brașov, where I am staying - is home to more than 250,000 people.
There are many cafes and the centre is easy to get around on foot.
A frequent bus connects the city to the resort at a cost of little more than £1, with the bus often hauling a trailer for passengers to attach mountain bikes.
Brașov has its own 'Hollywood' sign
A highlight of the city is St Nicholas's Church, founded in 1292 but rebuilt in stone from the late 1400s.
St Nicholas' Church

Exterior frescoes

Cemetery

Statue in the church grounds of Diaconul Coresi, who in the 1500s printed some of the earliest books in the Romanian language

My view while enjoying an espresso at the nearby Woof Street Cafe
 
Statue of the Unknown Soldier, unveiled in 1939 to commemorate those killed in WW1

Municipal flower display

Friday 29 July 2022

Brașov Round Eight

DOWNFLOATED today against another Romanian junior (born 2005).

Spanton (1889) - Stephanie Matasaru (1885) 
Sicilian Bb5(+)
1.e4 c5 2.Nf3 d6 3.Bb5+ Nd7 4.0-0 a6 5.Bd3!?
A relatively new move - the first example in ChessBase's 2022 Mega database is from 1984 - but it has taken over from 5.Bxd7+ as the mainline.
5...Ngf6 6.Re1 g6!?
The mainline in Mega22 continues 6...e6 7.c3 b5 8.Bc2 c4!? with an equal position, according to Stockfish15, although Komodo13.02 prefers White.
7.c3 Bg7 8.Bc2 0-0 9.d4 cxd4 10.cxd4 b5
How should White continue?
*****
*****
*****
*****
*****
11.e5!?
Trying to seize the initiative, but, although it might not be immediately apparent, the threat of being back-rank mated proves a major problem.
11...dxe5
11...Nd5 and 11...Ne8 have also been played, but the text is probably best.
12.dxe5 Ng4 13.e6 fxe6 14.Rxe6
Antony* Rode (2020) - Anton Demchenko (2628), Skopje Karposh ('Macedonia') 2017, saw 14.h3 Nge5 15.Ng5 Nc5, when Black had the upper hand, according to the engines (0-1, 27 moves).
What should Black play?
*****
*****
*****
*****
*****
14...Nde5
The engines reckon best is 14...Bb7!?, when Black's activity and lead in development more than compensate for the isolated king's pawn.
15.Qxd8 Rxd8 16.Rxe7 Bf6 17.Bb3+ Kh8 18.Nxe5!? Nxe5
Not 18...Bxe7? 19.Nf7+ Kg7 20.Nxd8 Bxd8 as White emerges a pawn up.
19.Rc7 Be6
How should White respond?
*****
*****
*****
*****
*****
20.f4??
Best seems to be 20.Nd2 Bxb3 21.axb3, when I feared 21...Nd3, missing that 22.Nf3 gives what the engines reckon is a dead-equal position. Komodo13.02 also likes 20.Bd2? but Stockfish15's 20...Bg8! causes White problems, for example Komodo13.02's suggestion of 21.h3 runs into 21...b4!, when White has major difficulties developing the queenside.
20...Bxb3 21.fxe5 Bxe5 22.Rc6 Bd4+ 23.Kh1
23.Kf1 is not much of an improvement - the white king is doomed either way.
23...Bd5 24.Rc2 Rf8 25.Nd2
The engines reckon best is 25.Bf4!?, which shows the position is well resignable.
25...Rae8
25...Rf2 also wins, but then so do many other moves.
26.h4 Re1+ 27.Kh2 Be5+ 28.Kh3 Rh1+ 0-1
*The player's name is more commonly spelt Anthony on the internet but Antony seems correct.

What Bears Do In The Woods

ROMANIA is home to Europe's largest brown-bear population, with more than 7,000 estimated in the country's forested Carpathian mountain region.
Attacks on humans are rare, but deaths do occur, and raids on villages and sheep pens are an accepted hazard of living in the mountains.
Not all bears are grizzly
The hotel I am staying at has a sign on the main entrance warning against going outside in the dark alone, and the tournament organiser advised me to make plenty of noise if walking out of the centre of Poiana Brașov even in daylight.
I was happy to accept their admonitions, until coming home from A&E late on Tuesday night when, having caught the 11pm bus instead of getting a taxi, I found myself walking in pitch darkness alongside a wood as I made it from the bus stop to my hotel.
I always carry a pocket torch on me, and this was one of the few times when I have been thankful I do as illuminating the blackness gave a feeling of security (probably false, but comforting all the same).
Reception in the hotel told me off, mildly, but what else was she supposed to say?
However, last night's "Extreme Alert" was about a bear on the woodside street I walked along on Tuesday night, so I am feeling a lot less sanguine.

Thursday 28 July 2022

Bear Necessities

I HAVE just been woken at 23:16 local time by my phone sounding a claxon alert I have never heard before.
The English part of the "Extreme Alert" reads:

The presence of a bear was reported in Predeal on Poinana Doamnei street! Avoid the area, stay indoor! Stay away from the animal and do not try to take pictures of it or feed it. Protect your pets/livestock without putting your life in danger. ISU Brasov.

As I was typing this, the alert was repeated. I guess they must have my number from using the hotel's wi-fi.
My phone is already on Do Not Disturb, but I guess the bear alerts override this.
I am turning my phone off ...

Brașov Round Seven

FACED another Romanian junior (born 2005) today.

Ilinca-Petra Batagan (1932) - Spanton (1889)
Catalan*
1.d4 Nf6 2.Nf3 e6 3.g3 c5
One point of Black's move-order is ...c5 can be played without White being able to push past with d5, although of course many blacks like White to push on to create a major imbalance.
4.Bg2 d5
The mainline in ChessBase's 2022 Mega database runs 4...Nc6 5.0-0 d5 6.c4, reaching a position Stockfish15 and Komodo13.02 reckon slightly favours White.
5.c4 Nc6 6.cxd5
How should Black recapture?
*****
*****
*****
*****
*****
6...Nxd5
The engines, and most strong players, it seems, prefer the most popular move in Mega22, 6...exd5, when 7.0-0 Be7 8.Nc3 0-0 reaches the main tabiya of the Tarrasch Defence - 15,057 games in the database.
7.0-0 cxd4
The commonest move in Mega22 is 7...Be7, with an almost even split between the replies 8.dxc5 and 8.Nc3 (the engines prefer the former).
8.Nxd4 Bd7 9.e4
White kicks the knight to gain space and let the queen's knight be developed without contracting two queenside isolanis.
9...Nf6 10.Nc3 Nxd4 11.Qxd4 Bc6
How would you assess this position, and how would you proceed?
*****
*****
*****
*****
*****
Black has no weaknesses and has a well-placed light-square bishop, but White has more space in the centre and is ahead on development. The highest-rated game to reach the position, Daniel Hausrath (2493) - Diogo Fernando (2482), ICC online blitz 2009, continued 12.Be3 Qxd4 13.Bxd4 Be7 14.Rfd1 0-0 15.e5 with advantage to White, according to the engines (1-0, 54 moves). The engines suggest an apparent novelty in 12.Qc4!?, claiming White has the upper hand, eg 12...Be7 13.Rd1 Nd7 14.b4!? a6 15.a4 with an initiative.
12.Qe3
This, like 12.Qc4!?, avoids an exchange of queens, but leaves the white queen's bishop short of squares. Nevertheless the engines reckon White's lead in development and extra space give a slight edge.
12...Bb4!?
Possibly a novelty. 12...Be7 is known from a nonsense game in Mega22 that continued 13.e5 0-1.
13.Rd1 Qa5
This is much better than the passive 13...Qe7?, after which 14.e5 creates major problems for Black as 14...Nd5 loses a pawn, and 14...Nd7 15.Bxc6 bxc6 16.Ne4 is even worse, according to the engines.
14.Bd2 0-0 15.a3 bc5 16.Qe2 Qb6 17.e5 Bxg2 18.Kxg2
After 18.exf6?! Qc6 White is holding on with 19.Qg4, according to the engines, but 19.fxg7?! Rfd8 looks good for Black.
18...Nd7?!
Better, according to the engines, is 18...Nd5, but 19.Nxd5 exd5 is clearly good for White.
19.Bf4
How should Black respond to White's threat to create an outpost at d6?
*****
*****
*****
*****
*****
19...Bd4!?
The engines prefer 19...Rfd8!?, but 20.Bg5 f6 21.exf6 Nxf6 22.Bxf6 gxf6 leaves Black with an extra pawn-island and an exposed king.
20.Kg1?
Too conservative. Strong is 20.Nb5 Bc5 21.Rac1 Rac8 22.b4 Be7, when 23.Be3 wins a pawn and leaves White with much the better position. But not 20.Qd2? Qc6+, eg 21.Kg1 Nxe5! After the text the game quickly peters out into a draw.
20...Bxc3 21.Rxd7
21...bxc3?! keeps the game interesting, but probably favours Black.
21...Qxb2 22.Qxb2 Bxb2 23.Ra2 Bc3 24.Rc2 Ba5 25.Rxb7 Rfb8 26.Rxb8+ Rxb8 27.Be3 Bb6 28.Bxb6
Or 28.Rb2 Bc7.
28...axb6 29.Rc7 ½–½
*ChessBase classifies the opening as a Catalan, although during the game I thought it was probably a Semi-Tarrasch.

First Time In An Ambulance

I GUESS I am lucky to have made it to 65 without needing an ambulance, but yesterday evening one was called by the hotel I am staying at in Poiana Brașov to take me to hospital.
I had been having sharp bursts of pain since lunchtime, and they became more frequent during my game yesterday and especially after it.
As I was waiting in the Romanian equivalent of Accident & Emergency - very smart looking, incidentally - I counted 35 stabs of pain in an hour.
The hospital conducted blood tests and put me on a painkilling drip, and I was discharged around 10pm with instructions to attend another, specialist, hospital this morning.
I will not go into where the pains were, but I ended up with three courses of tablets - antibiotic, anti-inflammatory and anti-pain, I believe - and a course of suppositories.
The good news is the doctors reckon there is nothing fundamentally wrong, although they could not tell me what I was suffering from, and the courses are more of a just-to-be-sure measure, I think I was told.
The interesting thing is, apart from using my passport to prove my identity, I was at no point asked to provide proof of health insurance or to produce my UK Global Health Insurance Card.
However, I did have to pay for the prescription - just over £15.
Unfortunately I had to default this morning's round-six game, and I am not out of the woods yet, pain-wise, but hope to play in round seven this afternoon.

Wednesday 27 July 2022

Brașov Round Five

UPFLOATED against another Romanian junior (born 2006) today.

Spanton (1889) - Alexia-Teodora Puia (1751)
Dutch Defence
1.d4 f5 2.Nc3!?
More popular are 2.Nf3 and especially 2.g3 and 2.c4, but the text has been tried by Garry Kasparov, Fabiano Caruana, Levon Aronian and other top grandmasters.
2...Nf6 3.Bg5 d5
A Stonewall set-up is easily the commonest response to White's system in ChessBase's 2022 Mega database.
4.e3
The mainline runs 4.Bxf6!? exf6 5.e3 Be6 6.Bd3 with an equal position, according to Stockfish15 and Komodo13.02.
4...e6 5.Bd3 Be7 6.Nh3!?
The engines do not like this, preferring 6.Nf3 or 6.Nce2!?
How should Black respond?
*****
*****
*****
*****
*****
6...0-0
This was played in all four games in Mega22 that reached the position after 6.Nh3!? The engines reckon equally good is 6...c5.
7.Nf4 Ne4
How should White proceed?
*****
*****
*****
*****
*****
8.h4!?
This positional pawn sacrifice may be a novelty. Weiqi Zhou (2638) - Yangyi Yu (2735), China Team Championship 2016, saw 8.Bxe7 Qxe7 9.0-0 c5 10.Nce2 with an equal game, according to the engines (½–½, 31 moves).
How should Black reply?
*****
*****
*****
*****
*****
8...Bxg5
The engines reckon White is much better after 8...Nxg5?! 9.Bg5 Qh5. They suggest declining the pawn, eg 8...Nc6!?, when both 9.Bxe7 Qxe7 and 9.Bxe7 Nxe7 are roughly level, but favouring White if anyone, according to the engines.
9.hxg5 Qxg5
White has more than enough for a pawn after 9...Nxg5 10.Qh5, according to the engines.
10.Rh5
The engines prefer 10.g4!?, giving White a slight edge.
10...Qg4??
Sensible is 10...Qe7 or 10...Qf6.
11.Qxg4?
This is good enough for a large advantage, but 11.Be2, as A-TP pointed out afterwards, is a complete killer.
11...fxg4 12.Nxe4 dxe4 13.Bxe4 g6 14.Rg5 Kg7 15.Bxb7?!
White is winning comfortably after 15.0-0-0 or 15.Ke2, according to the engines, while the text is unclear.
15...Bxb7 16.Nxe6+
What should Black play?
*****
*****
*****
*****
*****
16...Kh6?
Clearly White is much better - winning, according to the engines - after 16...Kf7 (or 16...Kg8) 17.Nxf8 Kxf8 18.Rxg4, but 16...Kf6 17.Nxf8 Kxg5 18.Ne6+ Kf5 19.Nxc7 Bxg2 is unclear. The engines reckon best play runs something like 20.c4!? h5 21.Nxa8 Bxa8 22.d5, when Stockfish15 gives White a slight edge but Komodo13.02 reckons the position is equal.
17.Rxg4
Threatening mate in one.
17...Kh5
Or 17...Rf5 18.Rh4+ Rh5 19.Rxh5+ gxh5 20.Nxc7 etc.
18.f3 Bxf3!?
The best chance, but losing.
19.gxf3 Rxf3 20.Rg5+ Kh4?
20...Kh6 avoids mate in the short run, but 21.0-0-0 Rh3 22.Rg3! and 23.Nxc7 wins.
21.Ke2 Rh3 22.Rag1 Rh2+
Or 22...h5 23.Nf4 Rh2+ 24.Kd3 and Nxg6#.
23.Kd3 1-0

Off ... But That's Fine

POIANA Brașov bills itself as an alpine resort, rather than a skiing or winter-sport centre.
In other words in the summer it hopes to attract hikers and others in search of sun and scenery.
But there is no doubt this is off-season - several of the hotels are shut and some days at least only a single gondola lift is working.
Going up
But that is fine as there is no queueing, and the main chalet at the top of the Postăvarul massif, on the northern slopes of which Poiana Brașov is situated, was open, the round trip costing about £10.
View from near the top

Tuesday 26 July 2022

Brașov Round Four

 FACED another Romanian junior (born 2006) in the second of today's double-round games.

Mihnea-Constantin Costescu (1800) - Spanton (1889)
QGD Semi-Tarrasch
1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 c5 5.cxd5 cxd4 6.Qa4+!?
This is a popular alternative to the mainline, which runs 6.Qxd4 exd5 7.e4!? (7.Bg5 is also common) Nc6 8.Bb5 dxe4 9.Qxd8+ Kxd8 10.Ng5 Be6, reaching a position that Stockfish15 and Komodo13.02 reckon favours White.
6...Bd7 7.Qxd4 exd5 8.Nxd5!?
This has been often played but commoner is the more-cautious 8.Bg5.
8...Qa5+ 9.Nc3 Nc6 10.Qe3+!?
The mainline in Mega22 runs 10.Qd1 0-0-0 11.Bd2, when White is at least slightly better, according to the engines.
10...Be7 11.Qg5 Qxg5?
Somehow I only saw the recapture 12.Nxg5?, when Black has the advantage. Better is 11...Qb6, not fearing 12.Qxg7?? Rg8 13.Qh6 Ng4 etc.
12.Bxg5
Black has a lead in development, but not enough to set off being a pawn down, especially as the black bishops are not active
*****
*****
*****
*****
*****
12...Nb4 13.Rc1 h6 14.Bd2 0-0 15.a3 Nc6 16.g3 Rad8 17.Bg2 Rfe8 18.0-0 Bg4 19.Rfd1 Bxf3!? 20.Bxf3 Nd4 21.Be3!
The engines agree this is best.
21...Nxf3+ 22.exf3 a6
How would you assess this late-middlegame position?
*****
*****
*****
*****
*****
White has an extra pawn, but it is doubled and cannot be converted into a passed pawn by ordinary means. On the other hand the white bishop is marginally more active than its black counterpart. I thought Black had decent drawing chances, but Stockfish15 reckons White is winning, although the more-conservative Komodo13.02 gives White 'only' the upper hand.
23.Rxd8 Rxd8 24.Ne4 Nd5
The engines marginally prefer 24...Nxe4!? 25.fxe4 Bf6.
25.Rd1 f5?!
Best is 25...b6, according to the engines.
How should White respond?
*****
*****
*****
*****
*****
26.Bc5!
Both 26.Nd2? Nxe3 27.fxe3 Bc5 and 26.Nc3?! Nxc3 27.Rxd8+ Bxd8 28.bxc3 b5! are good for Black, but 26.Nc5 also keeps an advantage, although not as big as the text.
26...fxe4 27.Bxe7 Re8 28.Rxd5 Rxe7 29.fxe4 Rxe4 30.Rd7 (1-0, 53 moves)

Brașov Round Three

FACED another Romanian junior (born 2008) in the first of today's double-round games.

Spanton (1889) - Ioan-Alexandru Iuga (1848)
French Alekhine-Chatard
1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Be7 5.e5 Nfd7 6.h4 h6!?
Stockfish15 and Komodo13.02 reckon this is the best way to decline the gambit.
7.Bxe7 Qxe7 8.Qg4
This is the mainline in ChessBase's 2022 Mega database, but the engines prefer 8.f4!?
8...0-0
One of the advantages of the ...h6 defence is Black can usually castle kingside in comparative safety.
How should White proceed?
*****
*****
*****
*****
*****
9.Nh3?
The white centre is too fragile to allow this. Normal is 9.0-0-0 or 9.f4, in both cases with equal chances, according to the engines.
9...c5!
This is not in Mega22, which has 9...Qb4?! and 9...f6?! The former can be answered with 10.0-0-0 while the latter runs into 10.Nf4.
10.Bb5
10.Nb5 is well-countered by 10...a6 or 10...Nc6. The engines suggest 10.dxc5!?, but agree Black is winning after 10...Nxe5.
10...cxd4 11.Qxd4 Nc6 12.Bxc6 bxc6 13.0-0-0 c5 14.Qf4 f6?!
This looks natural but the engines reckon it throws away much of Black's advantage. They recommend 14...Rb8 or 14...Bb7.
15.exf6 Nxf6 16.Qd2?!
Almost certainly too passive. I rejected 16.Qe5 because of 16...Ng4?! 17.Nxd5 Qb7, missing that after 18.Ne7+ Qxe7 White has the forking 19.Qe4. Then 19...Rb8 20.Qxg4 leaves White a pawn up, but Black's pieces are so much more active that the engines reckon Black is better. This is somewhat academic, however, as the engines reckon both 16...d4 and 16...Nd7 leave Black well on top.
White's best chance, according to the engines, is 16.Rhe1, when White has pressure against Black's impressive-looking centre, although they reckon 16...Qb7 gives Black an advantage.
16...d4 17.Ne2 Ne4 18.Qe1
How should Black proceed?
*****
*****
*****
*****
*****
18...Qxh4?
Black should continue calmly with, for example, 18...e5, maintaining a large advantage, according to the engines.
19.g3?
White at least equalises with 19.Nef4, according to the engines, eg 19...Rxf4 20.g3, or 19...Nxf2 20.Rf1, and if 19...Nd6? then 20.Ng6.
19...Qe7 20.Nxd4!?
It is a sign of how bad Black's position is that the engines are OK with this move. However, after ...
20...cxd4
... my planned 21.Qxe4 is answered by 21...Bb7.
21.Rxd4 Nf6
Black has bishop for pawn but is effectively much more than a rook ahead, according to Stockfish15, although Komodo13.02 puts the advantage at 'just' than a minor piece
*****
*****
*****
*****
*****
22.Nf4 Qc5!?
Giving up a pawn in return for simplification. The engines prefer 22...e5, when 23.Ng6?! Qf7 24.Nxe5 is met by 24...Qxa2.
23.Nxe6 Bxe6 24.Qxe6+ Kh8 25.Rhd1 Rac8 26.Qe2 Qa5 27.Kb1 Rfd8 28.Qd3 Rxd4 29.Qxd4 Qc5 30.c3 Qxd4 31.Rxd4 Re8 32.Kc2 Re2+ 33.Rd2 Rxd2+ 34.Kxd2
How would you assess this ending?
*****
*****
*****
*****
*****
Black has the only piece (apart from kings) and has possibilities of creating a passed pawn on the h file, but White's extra two pawns on the queenside look menacing. The engines agree Black is winning, but this is one of those unusual imbalances - unusual in the sense of not occurring all that often, except when the position is trivially winning for the player with the minor piece - that is tricky to play over the board.
34...Kg8 35.Kd3 Kf7 36.Kd4 Ke6 37.c4 h5
I-AI said afterwards he should have played 37...g5 to prevent my next move, and I thought he was probably right, but the engines reckon both moves win easily.
38.f4
Otherwise 38...g5 with ...h4 to come.
38...Kd6?!
Black is still much better after this, but 38...Kf5 is stronger, eg 39.c5 Kg4 36.c6 Ne8 with ...Kxg3 to follow.
39.b4
Black is still on top, but suddenly there are serious chances to go wrong
*****
*****
*****
*****
*****
39...Nd7?
Both 39...Ke6!? and 39...g6 are winning, according to Stockfish15, although Komodo13.02 'merely' gives Black the upper hand.
What is White's drawing plan?
*****
*****
*****
*****
*****
White needs to exchange off the black queenside pawn, even if this costs all three white queenside pawns, while at the same time penetrating on the kingside with the white king.
40.c5+ Kc6 41.a4 a6
The engines reckon Black can get away with, for example, 41...g6!?, but after 42.b5+ the white pawn-juggernaut means Black has to be careful not to lose.
42.Ke4!
The key switch of direction. Stockfish15 reckons Kc4 also draws, but recognises its error as soon as the move is inputted on the board.
42...g6 43.Kd4
Black is effectively in a sort of zugzwang.
43...Nf6 44.Ke5 Nd5 45.b5+ axb5 46.axb5+ Kxc5 47.b6
Diverting the knight.
47...Nxb6 48.Kf6 h4 49.gxh4
I-AI offered a draw.
49...Nd5+ 50.Kxg6 Nxf4+ 51.Kf5 Nh5 52.Kg5 Ng7 53.Kg6 Ne6
An only-move, but not difficult to find.
54.Kf6 ½–½

Wood You Believe It?

POIANA Brașov has a wooden church built in the regional Maramureș style.
Made entirely from wood, the churches were erected from the 1500s onwards in response to a ban by the Catholic Habsburgs on using stone to build new Eastern Orthodox churches.
Maramureș churches usually emphasise height
The example at Poiana Brașov is dedicated to St John the Baptist (Ioan Botezătorul in Romanian).
From the outside it looks interesting but somewhat plain - inside it is anything but plain.
Small but beautiful