MNb wrote on 07/16/17 at 18:40:00:

One thought occured to me: Why even play ...a6 here? Lines in the QGA where Black plays ...b5 and meets a2-a4 with ...b4 have been giving White headaches for a few years now. If Black can get there without spending a tempo on ...a6 (early on; any further a4-a5!? should still usually met with ...a6), that can only be to Black's advantage.

So 1.d4 d5 2.c4 e6 3.Nf3 dxc4 4.e3 b5!? 
This seems to be a specialty of Austrian IM Norbert Sommerbauer and Israeli GM Tamir Nabaty, among others.* For instance 
5.a4 (almost everybody plays this) 5...b4 6.Bxc4 Bb7!? (or 6...Nf6 first) 7.0-0 Nf6 8.Nbd2 Be7. Here White has tried several moves, but Black generally scores well. Perhaps the 9.a5 a6 10.Qa4+!? of Postny-Nabaty, Tiberias ch-IRS 2016 is critical. Nabaty felt compelled to play the passive 10...Nfd7 and lost, while something like 10...Qd7 11.Ne5 Qxa4 12.Rxa4 Nbd7 13.Nxd7 Nxd7 14.Nb3 0-0 15.Bd2 Rfc8 may be slightly better for White even though he can't take the b4 pawn.

Black can also fiddle with the move order and for instance castle before playing ...Bb7, but that allows White to get in e3-e4 (probably Black should then try to answer with a quick ...c5).

The alternative 5.b3 doesn't look too dangerous: 5...cxb3 6.Bb5+ c6 7.Bc4 and here Nabaty has played 7...Bb4+ 8.Nbd2 Bc3!?, but I don't see much wrong with simply 7...bxa2, i.e. 8.Rxa2 Nd7 9.Bd2 a5 10.0-0 Ngf6. Maybe White has enough compensation for the pawn, but if he wins it back, all the queenside pawns are likely to be elimated, which tends to make the position drawish - a typical scenario in these ...b5-b4 lines.

P.S.: Another reason it would be nice to get this to work for Black is nothing stops him from using it via the regular QGA move order: 3.Nf3 e6. You just have to choose which you're more afraid of: 3.Nc3 in the QGD or 3.e4 and 3.e3 in the QGA.

* Edit: Some of the games I've looked at started something like 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 b5. I'm not sure how much difference ...Nf6 (versus ...e6) makes here. It certainly takes away Nabaty's 5.b3 cxb3 6.Bxb5+ c6 7.Bc4 Bb4+!? idea.

If statistics are gruesome we should first ask ourselves if we think the variation worth bothering. When I want to play an opening that makes things difficult for White it seems to me that a Closed Catalan that might be somewhat inferior due to an early ...c6 is not the right choice for Black.
Scherbakov writes: "As White, strong players often try to get a position where they can safely aim for two results, a win or a draw, without any risk of losing the game. But in the Noteboom White never feels comfortable and any result is certainly possible."
This version of the Closed Catalan gives White exactly that kind of two results play, as Scherbakov implicitly admits in the chapter on 4.Nbd2 and the gruesome statistics confirm. Avoiding bad moves later on won't be sufficient to change that, I'm afraid.

an ordinary chessplayer wrote on 07/17/17 at 20:05:44:

Unfortunately Ruslan Scherbakov disagrees and I have the weird impression that he understands a little more about chess than any chessplayer up to 1946 ever could, including worldchampions. The reason is 6.b3 and White does very well. But even worse - what would a Catalan player do at move five!? Quick, you have only one guess. And then I suggest you to search your database for statistics. They are gruesome.
The Triangle simply isn't a problem for the Catalan player like 1.d4 d5 2.c4 e6 3.Nf3 dxc4 might be.

Indeed I would not mind playing 4.Qc2 dxc4 5.Qxc4 Nf6 and 6...b5 as Black. That kind of risky, active play is what Black is looking for. It's also not what Black gets after 4.Nbd2.

Stigma wrote on 07/17/17 at 17:19:13:

That's not really a consolation as soon as you realize how well Scherbakov's suggestions do for White in practice: 1.d4 d5 2.c4 e6 3.Nf3 c6 4.Nbd2 f5 5.g3 Nf6 6.Bg2 Bd6 (Be7 fares even worse) 7.O-O Nbd7 8.Ne1 or 7...O-O 8.Ne5. It seems to me that such positions are what Catalan players dream of.
But of course, to quote JEH: that's something for another thread, even if relevant if you play the Catalan.
It will bec lear why 3...dxc4 interests me. I plan to take a closer look at the Rubinstein Variation of the QGA in the near future.

an ordinary chessplayer wrote on 07/17/17 at 20:05:44:

That's quite a bold statement. Maybe you're right in the Torre case (I know even less about that), but surely in the London White just goes 3.Bf4 Bd6 4.Bg3 followed by e3, Nbd2 etc., which is liable to transpose to the main line? I don't see how Black gets in ...e6-e5.

There is a lot going on in this thread.
1.d4 d5 2.Nf3 e6 3.Bg5 Be7 and 3.Bf4 Bd6 are both instant equalizers. Seriously. Black gets in ...e6-e5.
1.d4 d5 2.c4 e6 3.Nf3 dxc4 4.g3 c5 5.Bg2 Nc6: I once lost to 6.Ne5!? (maybe it's ??).
1.d4 d5 2.c4 e6 3.Nf3 dxc4 4.Qa4+ (possibly !) 4...c6!? looks like the 3...c6 4.Qc2 line. Not my preference, just noticing.
1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nbd2 e6 5.e3 c5! = was given by Alekhine, and he is correct on this one. At worst black gets a decent IQP position.
1.d4 d5 2.c4 e6 3.Nc3: I have even played Steinitz's 3...dxc4?! here. 4.e4 should be +/- but somehow my opponents haven't found the way.

There is one further option for White after 1.d4 d5 2.c4 e6 3.Nf3 dxc4, namely 4.Qa4+. This is similar to the Mannheim variation of the QGA (2.c4 dxc4 3.Nf3 Nf6 4.Qa4+) and can directly transpose there.

Both lines were recommended by Burgess in A Cunning Chess Opening Repertoire for White, and I've been meaning to look into them for ages. These lines are not regarded as giving White much, but just leafing through Burgess' book they clearly lead to interesting, complex middlegames with everything to play for. Could be something to look into for the desperate Catalan player worried about 3...dxc4!!!. 

MNb wrote on 07/17/17 at 16:38:12:

At least after 4.Nbd2 f5 it's a Stonewall where White has given up most of his options: No Nh3 line, no Bf4 line, no early b3/Ba3, and no fashionable Nc3/Qc2. It should be possible for Black to prepare this limited part of the Stonewall if he has any affinity for the defence - admittedly I don't!

But I wonder why you think 4.Qc2 is less annoying? With that White keeps the option of playing the critical Nc3/Qc2 setup that both Avrukh and Wojo's Weapons recommended against the Stonewall.

Maybe Black should simply go 4...Nf6 against 4.Qc2 and allow the Anti-Meran and the Closed Catalan (where arguably White would prefer not to play Qc2 this early).

MNb wrote on 07/17/17 at 00:52:19:


I'm actually working on doing that 

MNb wrote on 07/15/17 at 15:38:58:

I will try to post some relevant lines on this, but I've just had too much work over the weekend. Stay tuned...

MNb wrote on 07/17/17 at 00:52:19:

I'm not sure if this was a joke or not. But actually, building a repertoire around the Von Hennig-Schara is exactly where this line 1.d4 d5 2.c4 e6 3.Nf3 dxc4 would come in very handy! You already have your line against 3.Nc3, but need to do something different against 3.Nf3, which prevents the VHS...

Though maybe the problem is the Von Hennig-Schara itself is dubious. I wouldn't really know why; I'm just lucky that nobody's played it against me while I've allowed it in serious games for years without knowing much more than Qa4+ and Qxd4.

Another repertoire quandary where 3...dxc4 could be the right puzzle piece is the combination French Defence/English Defence. 1.e4 e6, 1.d4 e6 2.c4 b6 and 1.c4 b6 are all covered, but 1.d4 e6 2.Nf3 is a problem. So why not 2...d5 intending 3.c4 dxc4!!! (retaining the three exclams for consistency!) 

But If I were to try this, I would be worried about the London and the Torre: 1.d4 e6 2.Nf3 d5 3.Bf4 and 3.Bg5 (or 3.c3 planning Bg5) respectively. I know White players of these openings are usually happy to see an early ...d5 and ...e6 from Black, when they can look forward to either using the e5 square for the knight and going f2-f4 soon, or getting in e3-e4 at a favorable moment. Though maybe this is "merely" a stylistic problem, not an objectively threatening one?