I'm liking the Online Chess format more and more, because I find having lots of time to think is the best way to improve my game. I play much more impulsively in Live Chess, and therefore play much worse.
This was an interesting game I played against "greasemonkey2660". I played Black. After only a few games, my rating of 1,496 was very provisional. Since I am working on the assumption that giving a lot of thought to each game, and comprehensively analysing it afterwards, is the best way to improve, a thorough analysis follows:
The game opened:
1. e4 e5 2. Nc6 Nf6 3. d3 Nc6 4. g3 Bb4
At this point, I thought Black was ahead in development. White had made three pawn moves in four moves, and Black had made one, giving Black three pieces in play compared to White's single developed piece.
With 4...Bb4, Black is threatening 5...d5 6. exd5 Nxd5 and then 7...Nxc3 8. bxc3 Bxc3 winning the rook. This can be avoided by 5. Bd2 so is not a serious concern for White, but developing with threats is good on general principles, and if nothing else it might tie up White's dark-squared bishop for a while, which may be useful given that he's already behind in development.
5. a3 Ba5
Black doesn't actually want to trade with the knight, as he wants to keep the bishop pair and doesn't want to helpfully open the b-file for White's rook. If now: 6. b4 Bb6 and the knight is unpinned, but the bishop is safe and White has created a permanent weakness on his c3 square.
6. Nf3 d5
After a normal developing move like 7. Bg2, Black is threatening 7...dxe4 8. dxe4 Qxd1+, forcing an exchange of queens and robbing White of his right to castle. Then, after 9. Kxd1 Bxc3 10. bxc3 Nxe4 then Black wins a pawn, and threatens to win another pawn or the rook after 11...Nxf2+, or to win another pawn with tempo after Nxc3+. After 7. Bg2, Black also has the threat 7...d4 winning the pinned knight. If White takes the pawn, then Black will reply with 7...Nxd5, adding more pressure to the c3 square.
7. b4 Bb6
White first decides to kick out the bishop, unpinning the knight, but creating a weak c3 square.
8. Qe2? d4
White fails to respond to Black's threat. After 9. Nd5 Nxd5 10. exd5 Qxd5 Black has won a pawn, has two central pawns and has claimed plenty of space and is ahead in development, and White's lack of space will make his development even more difficult. Best was probably 8. Nxd5 Nxd5 9. exd5 Qxd5 immediately, saving the pawn, and then 10. c4, kicking out the queen.
9. Nd5 Nxd5 10. exd5 Qxd5
as expected, winning a pawn and gaining a good position.
11. c4? dxc3 e.p.
White's move was a blunder, and he probably didn't see that the e-pawn could take en passant. 10. c4 works in the preceding variation, and White may have thought it through, but after 8...d5 it fails, so White loses a second pawn and Black has a dangerous passed pawn on the sixth rank.
Black took stock of his position, here, as follows. In terms of material, Black is two pawns ahead, has a passed pawn on the sixth rank, and a better pawn structure (White has an isolated pawn on the d-file, and a 2 vs. 3 pawn minority on the queenside).
Despite not having the first-move advantage, Black is a piece ahead in his development. Both sides have held onto their bishop pairs, and Black controls the a7-g1 diagonal with his dark-squared bishop. White will soon play Bg2 and take control of the long diagonal, which will be a positional advantage, but there is no feasible way for Black to prevent it, and White's knight on f3 is temporarily blocking. Black's compensation for White's control of the long diagonal is better developed pieces, which is a temporary advantage Black needs to exploit quickly or it will fade.
Black has better control of the center, with one central pawn compared to White's zero central pawns, occupying the e5 and d5 squares, guarding the d4 square with four different men while White guards it with one, and guarding the e5 pawn with two men while White also attacks it with two.
Neither king has castled, and the e-file is semi-open. The White queen opposes Black's king on the e-file, which is not an immediate threat as e4 is defended twice, but which may very well soon become one.
Black's strategic priorities are now to (a) castle; and (b) develop his light-squared bishop which will also connect his rooks. Then, Black will look to capitalize on his better position by finding an attack, before White can better his development.
11. Bg2 O-O
A natural developing move for White, which also attacks the Black queen if the knight now moves. Black continues with his strategic objectives, and now needs only to develop his light-squared bishop.
12. Ng5?? Qxg2
A game-losing blunder for White. White was probably looking to play 13. Qh5 after White takes the bishop, looking to checkmate on the h7 square. However, this attack is easily repelled by 13...h6. Black expects 13. Rf2, or 13...Qxh1 both loses the rook and leaves White with no time to carry out his attack anyway. Black sees that after 13. Rf2 Nd4 attacks the queen and threatens mate or ruinous loss of material if Black can kick the knight off g4 by playing h6, which would then naturally be followed by hxg5 if White does something to avert the threat of Bh3, winning the knight for a pawn and going two pieces up at a minimum. With the knight gone, White's kingside attack would also be over - Black doesn't see any way for the dark-squared bishop to help after Bxg5. So Black happily took the bishop, deeming both that White's threat is not credible, and that White's pieces will be in serious trouble if he doesn't play precisely as a response. Black took a significant amount of time to analyze this position - a benefit of correspondence chess! - calculating some variations up to 8 moves ahead, and was very comfortable with the played move.
Better for White was to just castle, instead of moving the knight.
14. Rf1 Nd4 15. Qd1 h6
Move 14 was as Black expected. White's move 15 was a surprise, a possibility that Black had not calculated, but it seems rather sub-optimal. Black now has the threat of c2, attacking the White queen, up his sleeve if the position demands it. Note that White's queen can now move only to h5 or a4 without being taken by a minor piece or pawn. Due to the knight on d4 and the c3 pawn (a surprising negative side-effect for White of 11. c4? dxc3 e.p., showing the restrictive power an advanced pawn can have), White's king cannot get off the first rank without going all the way to b1, which would take five moves including moving the queen and bishop out of the way, which would be too slow to fend off mate. Further, if White's king moves even one square, then unless the queen is on e2 (in which case it will be taken by the knight) then Black will take the rook and it'll be mate.
White's position starts to look hopeless. At this point, Black thinks that at a minimum White must either lose his queen and more material, or face a quick mate. The first step is to kick the knight out of the way with 15...h6, so that 16...Bh3 can be played. Black calculated several possible moves for the White knight, all of them losing.
16. Qh5? hxg5 17. Bxg5 Bh3 18. Qxh3 Qxh3
White has now lost his knight for a pawn, and his queen for a bishop. He had little choice. 18. Qe2 was the only other move which could have fended off immediate mate, and the queen would have been lost to 18...Nxe2 anyway. If the king then recaptured the knight, then 19...Bg4+ 20. Ke1 (forced) Qf3 is checkmate on the next move anyway. If the king doesn't move, then 19...Qxf1#. If 19. Kd1 then 19...Qxf1+ 20. Kc2 (forced) Qxa1 and White has lost all his pieces except his bishop, while Black has only lost one bishop, and White's king is in big trouble.
Black now has a rook fork on c2, and a bishop fork on f3. If 19...Nc2+ 20. Ke2 (20. Kd1? then 20...Qxf1+ and White loses both rooks at once.) Nxa1 21. Rxa1 Qg4+ and Black wins the bishop, too).
19. Rg1 Nf3+
losing the exchange, but all White's options are losing at this point.
20. Ke2 Nxg1+ 21. Rxd1 Qg4+ 22. Kf1?
winning the bishop. White's move 22 is a blunder not because it loses the bishop, which would have been inevitable, but because it gives Black a forced mate. 22. Ke1 was better, although it would only have prolonged White's suffering.
Now White is almost finished. If 23. Rg2, then 23...Qd1#. If 23. Ke1, then 23...Bxf2+ 24. Kf1 Bxg3#. If 23. Be3, then 23...Bxe3, and the same result as before if 24. Rg2 or Ke1. On any other move, then 23...Qxf2#.
23. a4 Qxf2#
I was pleased with this game, not only because I won, and won with Black at that. But it seemed like a good strategically played game with post-game computer analysis revealing no blunders on Black's part, and eliminating blunders - especially in games where I have 3 days to think through each move - is a short-term objective of mine. I think I developed well, took control of space, took care of king safety before embarking on dubious attacks, correctly assessed my opponent's threats, and spotted and took decisive opportunity of his mistakes. I'm particularly pleased I spotted the checkmate after 22. Kf1, and avoided the easy alternative of taking the hanging bishop and then trying to win the rook.