Wilhelm Steinitz: "The best way to refute a gambit is to accept it"

W.S. wasn't playing someone rated 401 higher when he said that.

Bill Hartston had this to say about this move: "3...PxP? 4.P-K5 leaves Black in a terrible mess. Once again the correct reaction to White's aggression is the central break with 4...P-Q4."

Not sure I would say ...exf4 deserves a full ?, it is playable, but ...d5 equalizes much more easily.

5...Nh5 would transpose into the Schallop Defense to the Kings Gambit. Black defends her f4-pawn without weakening her kingside pawns. Of course, the loose knight on h5 can give rise to some tactics.

5... Nh5 is impossible, and if you meant 4... Nh5, it just loses the knight.

5. ... Nh6 was possible; but knights on the rim are grim. White would enjoy a healthier lead than what was played. **** 5. ...g5 would have been interesting!

Still, it takes time for the queen to take the b pawn and then save the queen, and material is even after that. This is a tactical opening, you want to counterattack, so worrying about isolated pawns in an endgame seems premature. The alternative you played, opening the center when behind in development, seems far worse to me.

The Qs had to come off anyhow. After 14...Kf8 15.Bc4 Black's Q has to intervene at e8 or e7 to stop the mate.

correct. if you are going to defend this position, you can't afford to have your king cut off from the pawn majority.

This sort of thing comes with experience, I guess. Ken's suggested move would have been the first thing I would have thought of, and, as there seems to be nothing against it, I would also have played it. In the late middle game/early endgame as we have here, the K is usually safe enough in the middle of the board, and can contribute greatly to attack or defence. But I have an inkling that against another opponent, Joanne might well have preferred the centralising move.

As the I.M at my chess club said, "The king needs safety by castling when the queens are still on the board; when the queens are off, Black needs to use his king more aggressively as the game is nearer to an endgame, if not there already"

I also would agree with Kd6. If it were earlier in the game, I'd probably would pull it back, now you had the opportunity to use the king as a major defender of pawns and a vital defender of the center.

A very interesting decision. Were I playing Black, I rather think I might have retained at least one pair of rooks myself, and played, say, 19...Rd8. I'm pretty sure that there was nothing objectively against White's removing the final pair of rooks...

When behind, all other things being equal, avoiding exchanges is a good idea if for nothing else than to complicate. With rooks off the board, this is a classic won endgame because the doubled pawns can still produce a passed pawn and the B is slightly more active and effective than the N when there are pawns on both wings. All doubled pawns are not equal--if only the White pawn was on b2 instead of d4 this strategy would be a good one. As it is, this advantage is enough to force a win.

I thought about removing the last pair of rooks; but the exchange would only force the Black King closer to a Queen-side defense. Not taking actually gives White more time to bring up the pawns and stabilize the structure to White's advantage. My intent with Rb1 was to tie down the Black rook to defense because I expected Re7. **** Objectively, White could get away with the exchange of the last pair of rooks. Yet, the win might be harder fought than most would expect. There is potential for error on both sides. The rooks provide complications that could prompt errors on Black's part. **** Also, count the pawn islands and pieces. The exchange makes it mathematically more possible for Black to potentially both defend his own and attack one of White's islands. I recalled a beautifully won endgame by Bobby Fischer where the Knight side won because of too many islands on the Bishop side of the board. I also saw the potential for an eventual ...c5. Although the white pawn could pass (or triple up the c pawns), the black Knight and King could potentially capture it or defend against any further push. So I decided to complicate instead.

Either b6 or b5 would have been an alternative; but, 20. ... Re7 was good. I don't see a problem with the move as played.

The thing about 20...b5, is that it would have left the c-pawn weak, on a square liable to attack by the White bishop. On the other hand, a b5-pawn might have enabled the knight to establish itself at d5, but one feels that c3-c4 would have broken up Black's Q-side. I have more sympathy with 20...b6, but my concern over the c-pawn would still remain. My own view is that the defender should avoid moving the pawns (except maybe to get them onto dark squares, where the bishop can't reach them) and get really active with her pieces - all three.

I think this is more flexible...move pawns only when it actually helps.

White 20...Rb8? is an awful place for Black's rook. The text retains optimum flexibility and control of the open e-file.

23...Rd8 might have been worth a look... The game might have reduced itself to a rook ending by: 23...Rd8 24.d5!? (Bg6+ is an alternative) 24...cxd5 25.cxd5 Rxd5 26.Rxb7 Ne4+ 27.Ke3! Rxd3+! 28.Kxe4! Ra3 29.c6 Kd8. White's advantage remains considerable, however.

I'm assuming you meant 23....Rd7 instead. Then what you said makes some sense. **** 23....Rd7 24.d5!? although playable is not nearly as good as the alternative that you suggest. 23....Rd7 24.Bg6+ Kd8 25.Kd3 and I judge the position actually getting slightly better for White. At the very least, it is just as strong as played. If instead 24.Bg6+ Ke7, then White plays the centralizing Kd3 and still has plenty of room for additional pawn movement before attempting a d5 push. White should be able to wait for the proper timing. If necessary, he could initiate a trade on e4 and support the pawn break even further when he retakes with the King. If Black does not take on e4, then the Bishop supports an eventual d5! after going to f3. **** The White rook can always be used to prevent any additional support from the Black King.

Personally, I would not have moved the K-side pawns beyond placing the h-pawn at h6. Black's K-side pawns are more vulnerable than White's are.

During the game, I was a bit surprised that she did play the pawn beyond h6.

It is rare to get an advantage from advancing pawns when your opponent has the more mobile majority, they are simply easier targets.