Following are a few comments about your game. Great game, bye-the-way, and well played from both sides, more-or-less! Please take my notes in the spirit they are sent (i.e. I'm not pretending to be a know-it-all! I'm sure you could make similar comments to my games. So please don't feel patronised!) One theme keeps coming up in this game: Development. If there was any basic error, this was it: lack of concern for development, especially as it concerns the centre.
1.e4 e5 2.f4 Glad to see it! A fellow King's gambiter!
2...exf4 3.Nf3 g5 4.h4 g4 5.Ne5 Nf6 6.Bc4 d5 7.exd5 Bg7 8.0-0 Nxd5 [A game played in 1994 went 8...Nh5 ]
9.Nxg4 [9.d4 Might be more flexible. The g- pawn will probably fall anyway, at least if black plays ... 9...h5? which doesn't work cos the f4 pawn drops, and the knight cannot really take it due to the f7 weakness. (Better would be 9...Nc6 10.Nc3 Nxc3 11.Bxf7+ Kf8 12.bxc3 And, although the position is tricky, I'd prefer to play as white. Why? Among other reasons, the f-pawn will fall and the black king on f8 shouldn't be too happy about that! ) ]
9...Bd4+?! Not so sure about this. It seems to leave the Bishop on a more vulnerable square, a fact White indeed makes use of ...
10.Nf2 OK, so at least the Bishop check pretty much forced the Knight to f2, but g4 wasn't exactly the best square either. From f2 he perhaps also has a promising future.
10...Nb6?! Why not leave your Knight in the centre? This reminds me of the Ng1–f3-d4-b3 manoeuvre that one sometimes meets in a Sicilian. But usually, I think, there is a specific reason for such a blatant decentralisation. Here, I cannot see one. Maybe you did have a good reason? Perhaps it was to get your d5-Knight defended (being attacked the c4-Bishop) so that you can take the h4 pawn? Instead, better would have been a development move like 0–0 or Nc6 because the big problem in the following moves is Black is way behind in development, especially as it relates to the centre. Development is all the more important now, as the position is open.
11.Qe2+ Kf8 [Another option is 11...Qe7!? 12.Qxe7+ Kxe7 13.Bb3 I still fancy white here: Kingside pawn structure not so messed up as black's; King relatively safer; more central space (after c3 - hitting that vulnerable bishop! - and d4). But, Black is not facing an immediate tactical defeat. Black, with the Queen block on e7 fights against White's superior development. But strategically this is, I think, far worse for Black. Guess it is a matter of taste. Either fall behind in development, with queens on the board, and risk getting brutally mated, or counter White's lead in development and end up in a strategically difficult position. I dunno! What do the experts suggest is the best defensive strategy?]
12.c3 A critical position.
12...f3?! [A) After 12...Bxf2+ Black's king looks decidedly alone on the Kingside. It makes the position of the b6-Knight look even more dubious. And so to remove a defender of the dark squares around the king ... naah; B) 12...Nc6 is an interesting idea. This addresses the issue of poor development, something that the imaginative 12. ...f3?! doesn't do. But I'll leave you to work out the details. ]
13.Qxf3 Nxc4 14.cxd4 Qxd4 Perhaps this was the position you had in mind when you went in for the f3 pawn-push? I think the problem is that your Queen and c4-Knight are just too far into enemy territory, and can be chased away with pawn moves that will aid White's development. I don't think that White does that perfectly in the following moves, but well enough to keep a development advantage.
15.Qc3 Nc6 Good to develop the queenside at last. But it might be too late.
16.b3 Such an innocent little prod, but it soon becomes clear how White gets the initiative
16...N4e5 17.Re1 f6 Another time-wasting pawn move [17...Qxc3 Is, perhaps, the better chance to hang on and dull the sharp blade of White's better -yep, you've guessed it- development! 18.Nxc3 Otherwise White looses his development advantage, and black is left with just the, not very important, weakness of the kingside pawn structure 18...f6 19.Re4 Bf5 20.Rf4 Bc2 And, apart from the bleak prospects for the black f-pawn, the d-pawn will be pushed scattering the black knights. (20...Bg6 21.d4 Nd3 22.Nxd3 Bxd3 23.d5 Ne5 24.Rxf6++- ) ]
18.Ba3+?! But the game is not +-. Black needs only a little time to get developed and the game is perhaps only +/= or even = if white's pawn structure remains split too. Ba3+ only helps black to co-ordinate his pieces, by getting the king out of the way of the development of the h8-rook. Why this is significant will be seen later.
18...Kf7 19.Re4 Qd5 Good move. Now things look managable for black. [19...Qxc3 Notice, compared to before, this now works as after 20.Nxc3 Rd8! stops white pushing his pawn to d4]
20.Qc5?! One bad move begets another! White seemed to really want those queens off.
20...Be6 [But after the simple 20...Qxc5! 21.Bxc5 b6 ... and it is White's pieces that get driven back from insecure points in the black camp, giving black tempo and a comfortable game. For example: 22.Be3 Bb7 and the black rooks are joined, and the white e4-rook looks decidedly odd.]
21.Nc3 One can see the difference between the position after blacks 20 ...Be6 as compared with the previous note about 20 ... Qxc5! In the game black has had the chance to develop his b2-Knight, basically winning a tempo. (It would have cost White a tempo to regain material equality after the Queen was taken.)
21...Qxd2? But, from a practical point of view, this just looks too risky! OK, a pawn is a pawn, and I think a few modern GM's would take the pawn unless they saw a clear refutation. But this leaves White with a wonderful position, and black has to find the "only" move for a while, something you don't manage.
22.Rd1 White, with gain of time, develops his last piece, and the centre, for now, is HIS! I'd rather have White here any day, (even if a pawn down) especially as the black rooks still sit in the corners, the black queen cowers on h6. All White needs to do is find a point to co-ordinate an attack. But, black has the extra pawn, so perhaps with perfect play it should be a draw.
22...Qh6 23.Nd5 attacking an undefended pawn. If white can win a pawn, and keep his superior development and control of the centre, he should win, if he can keep the tactics under control. [or: 23.Nd3 Nxd3 24.Rxd3 Rad8 25.Rxe6 Rxd3 26.Qc4 Rd4 27.Re7+ Kg6 28.Qf7+ Kf5 29.Qe6+ Kg6 30.Qf7+ Kf5 31.Qe6+ Kg6 32.Qf7+= This forcing line is the computer answer to the "down a pawn but better developed" imbalance. Draw! But I wouldn't like to have deliberately gone in for this position as black unless I was very sure of my calculations!]
23...Rag8? But this leaves White's centre control unchallenged. It look like black has given up the fight for the centre and is hoping for some lucky shot against the White king. [23...Rad8! Notice that with this move, Black challenges two things: 1) White's control of the centre files, and 2) White's lead in development. 24.Nxc7 Rxd1+ 25.Nxd1 Bf5 26.Qd5+ Kg6 27.Ne3 Bxe4 28.Qxe4+ Kf7 29.Qd5+ Kg6 30.Qe4+ The computer is still happy with a draw here. After some nifty tactics, we get a draw by repetition. This was the last hope for a draw.; 23...b6?! Somebody recommended 23 ...b6 (or at least Chessmaster did). But I find this unconvincing. After the queen moves black can then either develop a rook into the centre (or even use his Bishop to pester the White rook on e4?) but basically the same position will arise as after 23 ...Rad8 with the only difference that the c6-Knight has one less defender. That may not be important now, but it may be a further weakness to worry about later. Like all the chess teachers say: you can't take a pawn move back! Better to avoid unnecessary pawn moves and use the pieces, especially when the initiative is with the opponent.]
24.Nxc7 After the Rad8 manoeuvre, Black could have, after this, 1) swapped off a pair of rooks to dull White's initiative 2) and by so doing, forced a Knight to the poor square d1 3)and as a result, have won an important tempo. Now, white simply wins a pawn for nothing!
24...Bf5 25.Qd5+ Kg6 So much for the aggressive intentions behind Rag8! The rook is performing the duty of protecting the back of the King's head! Not what he had hoped for! And, on top of this, white has won back his pawn and kept all of his advantages. The end could be nigh.
26.Nd3!! Hey! Clever! With the nasty threat of Nf4+ winning material. Black, therefore, is obliged to take off the rook, the f4 square defender, winning the exchange .... but loosing an important defender of the white squares
26...Bxe4 27.Qxe4+ Kf7 28.Bc1!! What is this guys rating?!! All the tactics work for this.
28...Qg6 This allows a forced mate, but unless you wanted to give up your queen ... (or after ... Qh5 allow certain death!) [Trying to save the queen by 28...Qh5 29.Qd5+ Kf8 (after 29...Kg6? and a Knight joins the slaughter! 30.Nf4+ ; 29...Kg7? and another piece joins the attack. This doesn't even need to be calculated! 30.Ne6+ ) 30.Ba3+ Natural forcing moves like this just finsh black off. 30...Ne7 31.Bxe7+ Kxe7 32.Nxe5 Qxe5 33.Qd7+ Kf8 34.Ne6+ and goodnight!]
29.Nxe5+ But this is a nice finish. Your uncle played this part of the game flawlessly! Well done to him!
29...Nxe5 30.Rd7+! Nxd7 31.Qe6+ Kg7 32.Qxd7+ Qf7 33.Ne6+ Kg6 34.Nf4+ Kg7 35.Nh5+ Kf8 36.Ba3+ Qe7 37.Qxe7# Actually, a brilliant finish. You both played a good and instructive game. The errors you made were not of the "Oh no, that looses my rook to a Knight fork" variety, but more strategic. I feel I learnt something going through this game (which is why I spent so much time annotating it if you want to know my secret!!) After you handed him the centre and a lead in development for free (not even at the cost of a pawn), after Rag8, it was a slaughter. If your uncle comes to Germany, warn me cos I don't wanna play this guy! 1-0