I think I might have been more inclined to take with the c-pawn, as the move played leaves Black's c-pawn backward on a file open to White's rooks. A judgment call, I guess! Otherwise, White's IQP (isolated queen's pawn) guards the useful e5 square and leaves White with plenty of room for attack. Attack White must if her advantages aren't to be frittered away...

the IQP is no weaker than black's backward c pawn, thus taking wiht the c-pawn seems better. Black's pawn position is more open and subject to attack.

White's isolani hasn't proved too much of a liability so far!

Had she so chosen, White could have prevented Black's castling by 21.Bh5+ and if 21...Kf8 22.Qf3 with a strong-looking attack: 22...fxe5 23.Qxf5+ or 22...Be4 23.Qg3 fxe5 24.Qxe5 Rh7 25.f3 Bd3 26.Rfd1 Bc4 27.Re1. If such adventures are not to White's taste, then 22.Bh2 keeps options open. Black's alternate retreat, 21.Kd7, leaves the K open to pressure down the e- and c-files.

Quite a lot was happening in this last sequence of moves, it seems to me - with some significant strategic decisions being made. It seems as though White has been driven back considerably. Can we rely on appearances?

More interesting strategic decisions. I think that at the previous diagram (following 26...Bxc5), Black's greater command of space probably gave him an edge. But with queens coming off, and a pair of Q-side pawns as well, the territorial aspect seem to have lost in significance. White now has a passer of her own, and the "remoter" one at that, which could prove decisive in a pure pawn endgame. Further, White's pieces are by no means constricted: there's plenty of open lines for her bishops and rooks.

That centre pawn phalanx is looking formidable. Could White have prevented this? 31.Rfd1 and 32.Bd3 look apt, but I'm not sure about 33.Bxe4. Preferable seems 33.f3, forcing Black to decide whether of not to exchange. If 33.f3 Bxd3 34.Rxd3 White can then double rooks on the d-file, redeploy the bishop (Kf1 then Bg1) and advance the b-pawn to b3. Such a position would be very difficult to break down, so a draw seems likely.

Probably 34...d3 was preferable (35.Rc4 Rd4 36.Rxd4 Rxd4 37.Bg1 Rd5 38.f3 exf3 39.gxf3 Kf7 and black looks fine.

That remote passed pawn is coming into its own. But Black had to prevent White's attack on the advanced and vulnerable e-pawn by occupying the g1-a7 diagonal himself betimes: 39...Bb6! To attack the pawn, White would have had to relocate the bishop: 40.Bh2 Ke6 41.Bg1 Kd5 42.Ke2 Ke4 (=). Overlooking that resource, Black goes rapidly downhill.

lack's next is a good defensive try. Had White got in 43.Ba7...

White's advantage has been steadily increasing ever since Black omitted the vital 39...Bb6! Having the draw in hand is a good place from which to play for the win: you can't lose! But here Black has no sensible counteraction based on going after the White g-pawn as after ...Ke6 (say) Bxg5 the BK can't come at the Wg-pawn before White gets in Bg5-f6 followed by g4-g5. Note that in the earlier play Black was never given the opportunity for any similar kind of action.

Good finishing. A fairly even game until just one small oversight - missing the opportunity to occupy a diagonal before the enemy did - led to a lost endgame for Black.