This is a sharp moment. Where the queen will go, and why?

I definitely agree Qg8 is inferior as e6xd7 is check. White avoids the trade, then black's queens is less active.

Mind you, she would have had a very active game for it. And that d-pawn, though isolated and weak, is a passer, which might be worth something. In broad terms, the suggested castling line wa worth a look.

Oh yes! That is why I mentioned it! It gives white far more chances to go wrong and therefore better practical chances for black, even if, with best play for both sides, black probably hasn't enough for the pawn.

That's a big mistake, I think. Black should opt for 13...Be7, he has compensation after your line.

What about a line similar to the one played later in the game? 14.Ne5 NxN 15.RxN. I can't say it's better than 14.a3, but it's worth looking at. Now black's bishop is undefended and needs to trade or retreat (or defend via a5). The pins are off, and white has a little more mobility. On the other hand, since the dark king is stuck in the center, white may want to keep his attacking chances by not trading as much, thus 14.a3. Just saying...

I think I would have preferred to retire the bishop to d6 or e7, with this sort of thing in mind: 14...Be7 15.Qxb7 Rb8 16.Qa6 Bh3 17.g3 Rf8, and it is hard to see that, with such a complete development and an active game, Black is at any real disadvantage. Observe that in this line, the Q-side rook will have found a way to activate itself. On the whole, the bishop probably ought to go all the way back to e7, as 14...Bd6 attracts the annoying 15.Ne4!.

I tend to agree with Ion. Black is actually a pawn up here, and maybe this was a good chance to give it back to activate her pieces.

And black pressure can increase with Qh5 - maybe the correct square to place the queen in before. bsb looking at white's king is better. Ne4 is good for white, but it leaves black a tempo to castle, for example.

How you feel about a position can make a lot of difference. Objectively, Black is probably not all that badly placed, provided she can bring the Q-side rook soon into effective action. That central passer has to be worth something - if fact the extra pawn and central majority might yet prove Black's salvation. But White's game does look the easier of the two...

In my view, she might as well play this at once, then decide whether to activate the rook by ...b6; and ..Kb7; or by ...a5.

Now the position is delicate for black, since he didn't contrast white's power on black diagonal. The BxN plan was probably the reason of the black's defeat.

er black was a "she" in this game!

in my opinion, ...b6 is absolutely necessary at this point. Now White plays 17.Ne5, and after the exchange Black again loses tempo to protect his pawn on d4, then White attacks strongly, for ex. 19.Rc1.

An alternative was 19.Qh3+ although what is White to play after 19...Bd7!?

Another plan was 19...Qf6. This protects the pawn for the moment, and also puts pressure upon the f2-square. Black would have had a bit of an attack going, with the threat of ...h5; ...h4 having to be answered. A possible line was, then, 19...Qf6 20.Re5! (threatening to double rooks) ...h5! (diverting the lead rook, which dare not ignore it: 21.Rae1? h4!) 21.Rxh5 b6 22.Qc4 Kb7 23.Re5! (of course, 23.Be5 meets a swift quietus after 23...Qxf2+) 23...Rae8 (White threatened 24.Re6) 24.Qxd4. So dies the d-pawn, but though a pawn shy, Black's drawing chances look reasonable after 24...Rxe5 25.Qxe5 Qxe5 26.Bxe5 (e.g. 26...Re8 hoping to induce 27.Bxg7? Rg8 28.Bc3 (say) 28...Rxg2+ 29.Kf1 Rxh2 (=)). At any rate, the bishops of opposite colour would be of some use in Black's defence.

Yes, in the middle game the opposite coloured bishops aid the attacking side, but in an endgame they can be a drawing influence.

Now the inactivity of the a8-rook is starting to bite.

Up to here, I reckon Joanne has defended very well, but this move should, I think, lead quickly to a White win. It has to be admitted, though, that for a 'best move', 20...Kd8 doesn't leap immediately to mind, but rather by a process of elimination: there's no good alternative. After 20...Kd8, these lines suggest themselves: [a] 21.Rxc7 Qe4 22.f3 Qd4+ 23.Bf2 Qe5 when it looks all up with the c7-rook; or 23.Kh1 Rc8 and Black's chances of survival don't look too bad. Though White can retain an edge by 23.Rxc6! bxc6 24.Qxd3+ Qd5 25.Qxh7, Black is still in the game. [b] 21.Rae1 d2 22.Qd1 ... This seems a deal better than 22.Bxc7+?! Kc8 23.Qh3+ Qf5 24.Qxf5 Rxf5 25.Rd1

Interesting analysis! Thanks Ion.