He should have accepted a draw

I played a 5|2 blitz with quite drawish positions throughout the game. Towards the end I offered a draw, not noticing that my opponent has under 20 secs left on the clock. He thought for a few seconds and declined.

2 moves later and it's mate in 2. I won on time. He hsould have accepted the draw.

Please analyze for any good/better moves I missed throughout the game. Thanks.

4...Bb4 (?) is suspicious (5.e3 followed by a3 and you have to trade the bishop for nothing).

6...c5 ? slaughters your pawn structure. After move 9 you have a position that occurs in the Rubinstein variation of the Nimzo-Indian defense, with the huge difference in favor of White that the bishop c1 has broken free. Otherwise, White could have played 8.dxc4 and (EDIT : this doesn't work, which makes Black's position even worse) your best is probably 8...Qa5 9.Bxf6 ! Bxc3+ 10.bxc3 (10.Bxc3 ! - Meilan1) Qxc3+ 11.Nd2 gxf6, and even if you will probably be up a pawn (c5 is undefendable) the terrible structure you get is a full compensation.

10...b6 is advantageously by ...Nc6, with maybe the idea of ...Nc6-a5-b3.

22.exd5 ? is almost a tactical mistake. Black gets too much counterplay due to his bishop. Had White played 22.e5 and then traded your bishop, you would have had huge problems on both sides.

24.a4 ? loses a pawn in view of what follows (the queen on c6 attacks it), but 24...Qc6 might be even better than trading rooks at once.

4...b4 clamps a pin on the knight. To play e3 followed by a3 costs white two moves and he ends up with a poor pawn structure.

6...c5 is the correct move as it challenges white's centre. 8...Qa5 is a mistake due to 9.Bxf6 exposing the king and doubling the pawns. Black's reply 9..Bxc3 is an outright blunder as the white bishop on f6 takes it, leaving white a knight and attack ahead.

22.exd5 is a mistake but only a minor one and white has at least drawing chances if not winning.

I agree with the fact 9...Bc3 ?? is a blunder in the variation I gave (this will teach me not to analyse in my head). But not with the other points.

4...Bb4 (?), as I already wrote, costs the bishop pair. The weakening of White's structure is delusional, he can always trade c4 vs. d5 and get rid of the doubled pawns. (this is the reason for which the ...d5 move is almost always a mistake in the Sämisch Nimzo-Indian d4 Nf6 c4 e6 Bb4 a3). The loss of tempi is a more serious thing, but is justified here - cf. the Rubinstein Nimzo-Indian.

6...c5 ? challenges White's center, indeed, but then why play 9.c4 that abandons the tension in the center and will get a weak pawn after the previsible e4 push that White will play sooner or later ? As the only other choice is to trade the c pawn against the white d pawn, you get an isolated d pawn (IQP) without the initiative to support it.

A better way to play this would be to trade first (...d5xc4) and only then play ...c5 (QGA-style) where it's White who has an IQP.

As for 22.exd5, it's not a blatant, tactical mistake, but look at this (this is no real analysis, and play could be improved in both lines) :

Here are some variations you suggested and why I agree or disagree. Thanks for taking the time.

c5 allows Qd6 later breaking the pin and protecting the knight. Without c5 white can play dxc4 either chasing taking away the queen from the knight's protection (or losing a pawn).

5.c5 ? : I never mentioned it. This move is almost never good in the QGD positions.

10...Nbd7 : better indeed than ...b6, but not as good as ...Nc6 that threatens to go to b3 via a5.

I agree your line after 22.e5 is better that what I thought, but White keeps a small edge after at the end of your line 26.Rb5 that almost forces 26...Nd7

And sorry, but I don't understand post #6.

The best way to break the knight-queen pin without running the risk of BxN and having to take back with the pawn (thus exposing the king) is to move the queen to d6. However unless black moves c4 white can play dxc5 (which also threatens the queen), forcing black to either take back with the queen (thus losing the gurad of the knight) or moving elswhere along the 6th rank but losing a pawn.

That is the thinking behind c4.

Why even ask for a draw? You should fight for the win, in whatever position you get. Even with a basic Philidor.

No, no, no. I mean : yes, if White just waits for you do to what you want, but it just won't happen.

in that kind of positions, you break the pin by ...Nbd7 and then move the queen (probably to a5 after ...c6). That's not a great prospect, but your opening was inferior.

If you think this is passive and the Bc8 can't go out easily, this is true. If this is too passive for you, consider stopping playing 1...d5 and have a look at 1...Nf6.

Again : (1) ...Bb4 (?) trades your best bishop, that's why the move shouldn't be played in any d4/d5 position, you need reasons to back it. and (2) ...c5 (?) as in the game gives you an IQP without any initiative.

If you don't understand these, or think this is unimportant, keep playing it that way. I predict you painful losses, with a slow positional agony, faced to stronger players, to whom you give for nothing the pair of bishop or an easy IQP target. Then you will learn (and if it doesn't happen this way, please come back and show me - then I will learn).

I think it is easier to find variations for white that might work like 5. e3 6. Nge7 keeping a N influence from c3 after the pin is captured through