The gameknot analysis version appears here: game

I keep thinking that the Bd3-Bxc4 move should be avoided as long as possible. It is a tempo loss, and castling is not a priority in the short term... cxd5, followed by the lsb's outcome, at a first sight looks better. As a well known issue, if black takes with the knight and then Nxc3, it is pretty playable for white.

Possibly White deliberately 'sacrifices' a tempo in order to induce this exchange - a feature of Queen's Gambit Declined openings as well. But here White is left with an 'isolani' - an IQP, isolated Queen's Pawn. This can be a long-term liability but a short-to-medium tern asset for the attacking player. How will it go here? Too early to say.

Blsck's conservative development of his knights might tend to recover for White his lost tempo...

Not having a knight in f6, Bd3 followed by Qc2 could be a cause of concern for black.

I think I would have looked seriously at 11.Bd3 eyeing h7. There would no real threat yet of a quick attack against h7, but the potential for one might worry Black.

The white's lsb was too good to exchange. Easy to say it now, though!

Black has arranged that very cleverly, and has established a firm blockade of the isolani with the ideal piece for the purpose: a knight. Observe that White's fine outpost at e5 is not really tenable for the long term - sooner or later he will be chased out of there. Meanwhile, white's attacking prospects on the K-side have been considerably diminished...

Had I been Black here, I rather think I would have kept that turbulent priest...

Black was very wise not to exchange the knight for the dsb.

Oh, yes! The Knight is a very powerful piece where he stands: rock solid, immovable from White's perspective; but from Black's point of view with far greater than White's bishop. At the moment forming part of a pawn chain - more like the country cure (how do you do acute accents on this thing?) among the peasantry, than the lofty head of a diocese. I rate Black with a small but distinct edge, here.

Hardly surprising!

I cannot understand this move, Joanne... Because of the black's answer. I would try b3, here, since no inmediate danger of having the dsb out of the defence of the diagonal. Or Qh4, aiming to e7. After Nd5, Bc5, and if b6, I'll try Qd6 or Bd6, or Bxb6, Rxa3 and accept the exchange.

I can't say I go a bundle over 25.b3 as a possibility: it seems to weaken the a-pawn too much after 25.b3 Nd5, where is the bishop to go? Yes, Black might suddenly be interested in exchanging Knight for Bishop, with the idea of obtaining an advantage in the heavy piece ending that follows - White's shattered pawn structure on the Queen side. This is one of the unsung advantages of having the superior minor piece - whatever the situation. The player with the better piece usually in the one in charge of when and if the minor pieces are to be exchanged. It's like having the weather gage in a sea battle...

So strong is this knight that I would have left it exchanged with my bishop, specially in the hindsight... but it is easy to say it now. I should have protected the a3 pawn somehow, and it is doesn't seem easy. Ra1 seems to be the only way.

The interesting thing here is that the game has come down to a Queen and minor piece ending. In general it is thought that this is one type of endgame in which the knight tends to outshine the bishop, mainly because the Q/N partnership works so well. In this position, the Knight dominates the bishop from his 'Krak des Chevaliers' at d5; the bishop has no worthwhile moves. But White has been defending fairly well, I think. The game is not lost yet! +++++++++++ Having made this comment, I became curious as to what the GK 'thought'. It gives the position as dead level, pretty much, which I don't see. However, Black does something now that I would not have expected; offers the Queen exchange.

Correct. I still rate Black - slightly better pawns, the better minor piece, and better placed to take and hold the initiative. The problem as I see it, is this. He has one obvious point of attack: d4. White can defend this easily enough. The trick for Black is develop another threat to stretch White's defences.

Somewhere here it should be a strategic mistake. It is difficult to say which move. I don't like g4, but I cannot give a reason. Maybe someone can.

I think White approach is the correct one here. Black hoped to break up White's king side - hence the tactic 34...f5. Jo wasn't having with that! Now she can close up the K-side.

One feels that White should be able to hold, from here. Her bishop is 'bad', sure, but 'bad' bishops tend to come into their own in defence, being able to protect the pawns.

I think I would have avoided playing h4 if I could, keeping it in reserve as a 'waiting' move. But Black is being clever here, because he has succeeded in developing a second point of aim: f4. Further, his Q-side pawns are just mobile enough to create threats on the Queen's flank. I am supposing Jo's 39th move was to retain the flexibility to close the Q-side if she gets the chance.

I'd probably play 44.Bc3 here, though I see nothing really against Jo's King move. The idea is to draw Black on: 44.Bc3 Ne4 45.Be1 c5 46.dxc5 bxc5 47.a4 b4 48.Kd3. White's protected passer ought I think to have kept Black from pressing too hard into White's position.

This pawn advance had to come: that was what Black has been angling for for quite a while. Except for poking and prodding at d4 and f4 this was really Black's sole chance. It seems to me that 45.Be1 must have been the mistake that gave Black the 'in'. I think I would have chosen 45.Bb4 with the idea 45...c5 46.dxc5 bxc5 (If Black tries 46...Nxc5ch, 47.Bxc5, whereat [a] 47...bxc5 48.a4; or [b] 47...Kxc5 48.Ke3! and Black has no breakthrough) 47.Be1 c4ch 48.bxc4ch bxc4ch 49.Kc2 or maybe 49.Ke3 look good ebough to hold. Also playable (possibly) was 45...Be3, again covering c5.

My first thought was that 47.Ke3 was a mistake, but on reflection, it becomes clear that the damage has already been done. The game can not be salvaged from here. After 47.Kc3 Ke4 48.Bf2 Kxf4 49.Bxc5 bxc5 50.a4 bxa4 ( the GK engine gives 50...b4ch, but I prefer the capture) 51.bxa4 Ke5 (possibly 51...Kg3 leads to a winning Q/P ending, but I prefer to round up the White passed pawn). Now: [a] 52.a5 Kd5 53.a6 Kc6 54.a7 Kb7 55.Kc4 f4 56.Kd3 Kxa7 57.Ke4 Kb6 etc; or [b] 52.Kc4 f4 53.Kd3 c4ch 54.Kxc4 Ke4 55.a5 f3 56.a6 f2 57.a7 f1=Qch 58.Kb4 Qa6 Black wins.