Thoroughly confused

Hello

Here is a game I recently played. I lost, but I am not sure where I went wrong.

I don't know for sure, but I don't like 18. 0-0 -- you're moving your king away from the center into a less active position in the endgame, while your opponent's king is on a more active square. That means you're automagically making yourself slightly worse. 

Why not 18. Nd2, you're attacking the knight, forcing it to move, and allowing you to play f3 and e4 to use your pawns to control the center?

Once you've castled black has better development and more active pieces. His position is just easier to play. 

Instead of 7. Bd3, you could have tried the string:

7. cxd5  Nxd5 8. Bxe7  Qxe7 9. Nxd5  exd5 10. Bd3 0-0 11. 0-0

Maybe this would give you better positioning and wouldn't have catapulted  the game straight from opening to endgame and given him the initiative.  

On move 19 what about Ne5 preparing Rd7 check? It allows you to also now move the f pawn to kick the knight and make an escape square for the king.

Both of you play better than me so I can't be much help here, but I am interested to see other feedback from better players.

im not strong in opening theory so for that part you would probably want a stronger player to look at that but on move 17 you should have played ke2 so you can challenge black for the open files and not allow black to invade with his rooks after which you have nothing to do but play passively while he builds up pressure so dont underestimate open files and invading rooks

@ chrisr2212: what would you suggest for white instead? is Qc2 good?

@ DaneHitchins: I do normally play the exchange on move 4, but I also wanted to try the mainline QGD, since my favorite line vs the qid/nimzo is 3.Nf3 b6 4.a3, and 3... d5 leaves me unable to play the exchange properly.

Thanks to everyone for their comments!

I guess I never thought about delaying the exchange until after ...Nbd7 prevents...Bg4

is it similar to the normal exchange?

Yikes, this was longer than I planned, I went back and bolded the main points if it was too long to want to read :)

5...a6

You're right, it's not book, but it's what I'd call book-ish.  First of all, it's a non-developing move, so it's useful to see if you can take advantage of it with active moves like Qb3 or Ne5.  I don't see a way to punish it, so I would just find a useful way to transpose back into a book line.

So now I give a long winded explination of some pawn structures... but really there are only 2-3 possible ones here (and only about a dozen or so in all of chess), each one with its own ideas and time tested plans, so it's useful to help with your overall understanding of what to try for.

The first is after white plays cxd and black re-takes exd for what Soltis calls the orthodox exchange.  In this structure white will go for a minority attack on the queenside, advancing his a and b pawns.

Black can't really play c5 in this formation because after dxc his isolated d pawn is a weakness, so black plays for a kingside attack / piece play.

Therefore in this orthodox formation, the moves a6 and c6 are often useful in slowing down white's minority attack.  White will get in the b5 break anyway, but it will take longer.

So one option here was for you to simply play 7.cxd and it would transpose into book and your idea (a minority attack) would be clear.

The 2nd idea you could have done was to play 7.c5  I don't know if this has a name, but it's just basic chess... you usually punish an early c5 with b6 and or e5, undermining the pawn chain but here black can't play e5 and black's move a6 which weakened b6 for black also points to the idea of c5.

In this case your plan would be a queenside expansion of your own.  0-0 with b4, and set up b5 break by moving your heavies to the queenside (rooks and queen).

The other formation is after black plays dxc and you re-capture with the bishop where you get a slav structure.  In this case a6 has a totally different purpose, which is part of a black queenside expansion, b5 and c5.  In this case you really can't stop c5 unless black screws up somehow.

But here was black's trick, he didn't play a slav on move 2, so he wants to get his c5 break in with just 1 move, the greedy jerk.  So your instincts were correct, 7.Bd3 gave black equality just about right away for this reason... you actually reach a book like position but black's up a tempo.

Pawn structure aside, usually there is a tiny battle for that c4 tempo, and white delays Bd3 as long as he can, usually by playing 7.Rc1 first.  Notice this also hits indirectly at black's c5 idea.  If black simply plays 7...0-0 I would go for dxc5 because as we discovered Bd3 loses a tempo when black enters a slav, and now c5 isn't quite as good because Rc1 doesn't make sense in terms of the b5 break you want in that structure.

-----------------------------------------------------------------

Ok enough about that, lets talk basic chess.

Indeed, after all those trades it's a dead even position... but black is ahead in development.  One quick and dirty tip is the open files furthest from your king are the most valuable for penetration (idea stolen from Michael Stean's book Simple Chess).  So here 17.Rc1 guards the most valuable file.  17.b4 was actually very weakening, it moved your pawn weakness from the 2nd rank (at b2) to the 3rd rank (at a3) and gave up the c3 square and c4 square.

21.Rdd1 is a passive move.  Would you rather defend mate by moving Kf1 or g3, or tie up one of your most valuable pieces to defense?  Activity is paramount in the endgame.  Shedding a pawn here or there for rook activity is a very common idea.

After this you could activate your a1 rook with your a4 break (one reason I really liked your Rd4 move, which guards b4 and prepares this break).

You try this giving up a pawn for activity eventually on your 26th move, but at that point it's a bit too late.

Wow, this took a lot longer than I anticipated, I think I'll leave it at that!  There may be other moves worth pointing out, but these are what popped out at me.  Hope this helped.

Seeing as everyone else has spoken about moves preceding the 20th move, I'll point out that 33.Rxh7 might have been greedy. Rc6+ would pick up the a-pawn, thus defending your own a-pawn, and you could later have gone after the h-pawn. You should double check that though; that's just my own opinion.

I'm not sure about 17.b4 (?) Were you concerned about a rook hitting the b2 pawn? It seems you have more urgent issues to deal with, like finishing your development, meeting black's rooks coming to the c and d files, and chasing away that pesky knight. b4 doesn't deal with any of those, and in fact hands the c3 square to black. Maybe 17. Nd4 or 17. 0-0, or 17. Ne5 (17...Rac8 18. Rd7+ Kf6 19. f4!). I do think Black is at least equal after his 16th move, but I think he has an advantage after 17. b4. 

Thank you all for your very thoughtful replies; seeing my own positional mistakes in a game is much more helpful than reading a book on middlegame theory .

I don't think anyone has mentioned 21.Rdd1 yet, but it looks suspect to me. This move worsens the rook's position, and doesn't do a very good job of defending the threat of back rank mate that caused you to move it there-you play 22.h3 anyway. Also, 21.Rdd1 allows 21...Ne4, which is obviously dangerous. Perhaps 21.Kf1 or 21.h3 or 21.g3 would have been better. Not sure which move is best, but surely not Rdd1.