Pawn structure

Interesting question. I have two points to make.

First, if I were to play 1.d4 d5 2.e3 Bf5 3.Bd3 Bxd3 4.cxd3?!, I would be doing it not with the eye of playing e4. I would do it because after black eventually plays either ...c5 or ...e5 then I would play dxc5/dxe5 followed by d4, replacing the pawn which was just there. Throw in some nice control over the e4 square, the half open c file, and it's not really that bad. Of course I would nearly always play Qxd3 followed by c4 because that's more my style of play. My style is very classical.

Secondly, there is another one worth mentioning here. 1.d4 d5 2.e3 Bf5 3.Bd3 Bg6 4.Bxg6 hxg6. Very often the pawn structure for black is better after hxg6 than it is at the start of the game. This belief of mine was confirmed by an article in Chess Life a bit back by GM Larry Kaufman who said that the hxg3/hxg6/axb3/axb6 is the only example of doubled pawns which improve the position on a consistent basis.

This article:

http://home.comcast.net/~danheisman/Articles/evaluation_of_material_imbalance.htm

Yep, hadn't mentioned this as it wasn't in the question. I used to play 1.d4 d5 2.c4 Bf5 as Black, and often after Bd3 or Nh4 I'd drop my bishop back to g6; this was my usual decision.

Actually, I did but then my post got swallowed up by the ooops-you-timed-out- making-a-sandwich-in-between-finishing-and-posting-your-writing electronic void. So you'll have to wait (breathlessly, I'm sure ) for when I can find time to recreate it.  I'd do it now, but I have a (rare) social obligation to get to!

I'm also going to have to mention the "my-father-says-it-right" argument and point out that my chess engine, at 21 ply, slightly prefers cxd3 to Qxd3 (giving a slight plus to white, with both)! Take that Mr. NM Cobblepot and your snooty "gee looked at in the proper light cxd3 isn't that bad."

You do know that computers do not create opening books?

I do. And I also know that my appraisal of Qxd3 (as an error) in the "opening variation" of the original question is wrong. But there is a case to be made for cxd3 as a good move -- as a preferable move -- to Qxd3 (in the opening variation) and in light of that I think it is worth pointing out that an engine prefers cxd3 -- the preference is slight, and by no means do computers get the last word in the opening, but I've gotten the feeling that some of you think that cxd3 is fundamentally unsound in all cases. I think an engine preference for cxd3 shows that there is no easy refutation for the move, and it is not fundamentally unsound.

But I suggested that Qxd3 is fundamentally unsound (in the opening variation) and that is not supportable either. Qxd3 is ok in the opening variation (because e4 and c4 can still be played) My revised opinion is that cxd3 is more dynamic. A classically minded player, like ozzie, puts lots of value in "static" analysis, 'pawn structure  king safety, etc. the classic positional points. The dynamic player (Topalov, Kasparov) is willing to give up static pluses for initiative -- it's the contemporary version of gambit chess. In the old days attacking players threw away whole pawns for time. The contemporary player is working with the same idea, but he concedes things like a weak pawn for initiative, rather than an outright pawn sac/gambit.

So, to answer marvellosity -- yes Qxd3 leaves c4 a possiblity  and opening the c-file a possiblity... but cxd3 gets it done right away, and supports an e4 push right away, and without tying down the Queen to duties like supporting pawns. cxd3 followed by Qb3 puts immediate pressure on the e pawn. 

I'm not any longer arguing that Qxd3 is bad move... but I am arguing that in some positions, and particularly for some players (the attacking initiative type) cxd3 is as good or better.

cxd3 I don't think is bad, but what's so dynamic about it? Both ways e4 can be played. There isn't really a reason why e4 or the c file (especially the c file) need to be used right away. In fact, if you played e4 for dynamic reasons, you'd probably envision an e5 follow up, but I think this highlights the weakness of the doubled pawns. You even said you wouldn't play e4 right away, but then why not Qxd3? I don't see the point of commiting to a potentially weak and slower developing structure when you don't have to. I would say the queen is pretty happy on d3, not tied up, as it's a natural square anyway. The c4 lever opening up the c file is at most half of the value of c4. The real value is that it challenges the center.