I actually looked a lot more at 1.d4 d5 2.c4 e6 3.Nf3 dxc4 4.g3 and it's really not that bad for White (even in the c5 lines). A lot of lines have the potential to transpose back too.

It's an equal game but White has a bit of initiative, a lot of Catalan themes and way less forced play than in the 4. e4 lines. Were your opponent to know the e4 lines, he can force the game to very drawy wters (unless you're willing to be worse as White).

It can also have endgame lines but they aren't as forced. For example:

1. d4 d5 2. Nf3 e6 3. c4 dxc4 4. g3 c5 5. Bg2 Nc6 6. Qa4 cxd4 7. Nxd4 Qxd4 8.
Bxc6+ Bd7 9. Be3 Bxc6 10. Qxc6+ bxc6 11. Bxd4 Bb4+ 12. Kf1 Nf6 13. Kg2 O-O-O
14. e3 Rd5 15. Bxf6 gxf6 16. Rc1 Be7 17. Nc3 Ra5 18. Ne4 Rb5 19. Rab1 Rd8 20.
Rxc4 f5 21. Nc3 Rb4 22. Rxb4 Bxb4 23. Rc1 Rd2 24. Na4 Kc7

It's kind of funny though that every time I study an opening in depth it turns out to offer nearly nothing for White if Black plays the main defenses. I might as well play the Colle or something and just know from the start I won't have an advantage! To be fair, more theoretical openings tend to have way more pitfalls in the positon, more tension, and can punish big mistakes harder. This was my problem with 1. b3 during a stint with it. The positions were fine but I'd get random moves like 1. b3 a5 and STILL couldn't pounce on them or do much. Literally almost any move Black continued to be equal. If you go 1. e4 and your opponent goes a5, he will have a hard time withstanding your initiative.

Interesting MNb, I hadn't really thought of that.  Of course there's 5.a4 which should transpose to 7.a4, but White may not have wanted to play that way.

So I guess 3.Nf3 dxc4!!! does deserve all three exclams after all  .

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


As Eric said, I think it was c5. I do know that it lead to some kind of worse version for White since I considered it too.

Anyway, I just like to have an internally fully consistent repertoire even if there's no practical advantage to it. I guess I can always just play e3 vs that one and consider the  immediate 1. d4 d5 2. c4 dxc4 a slight inaccuracy by Black. Although, I'm sure a Black player would tell me that 3. Nf3 is the real inaccuracy since 3. Nc3 suffers from none of these problems.

BTW: Yeah, I guess that position is OK OTB and it's obviously going to be fine in my games if Radjabov can pull it off. However, it's definitely favoring Black and he was fortunate that his opponent played 16. ...Ba6 instead of Nc6. Ba6 gives White an equal game.

Stigma wrote on 07/14/17 at 23:44:12:

Makes me wonder what's wrong with 1.d4 d5 2.c4 e6 3.Nf3 dxc4 4.g3. Is it this?

Stigma wrote on 07/15/17 at 07:05:03:

If yes, do you have a concrete line to demonstrate it?

My guess is that Avrukh wanted to recommend 3.e4 against the QGA and then noticed this later, and to patch up the line recommended 4.e4 despite it not being the strongest move.  But there's still scope in there for outplaying your opponent; it's gambit play, and these days it's actually beneficial that engines are high on Black's extra pawn, because in over-the-board play it's extremely unlikely that Black is going to defend with that kind of precision.  Here's one nice game where White gets a menacing position straight out of the opening in the line you gave:



I would be pretty happy with White's game at move 16 I think, regardless of what the engine says.  At least I would in OTB play. 

Objectively, I'm sure that 4.e3 is best, and if you can't find a single line in there that you like as White, perhaps the Queen's Gambit isn't for you?  You'll never be able to force an advantage by analysis in a well-respected opening in this day of engines, you just have to pick a few lines that suit you and do some homework to set your opponent problems. 

Also, I think that you could probably go your entire life and never face 1.d4 d5 2.c4 e6 3.Nf3 dxc4 as Stigma hinted at, so I don't think that it's much of a problem in reality.  But you can still play 3.e4 against the QGA in the vast majority of your games, and then just familiarize yourself with a particular other line via the 3...dxc4 move order for that one game in a decade that will go that way.  If you knew that an opponent had prepared this line against you, you could try 1.d4 d5 2.Nf3 to see what his second move was; after 2...e6 you could play 3.g3!?.  Of course after 2...Nf6 3.c4 dxc4 you'd still need another line to play against the QGA than 3.e4, but my point is that if you suspect Black of some trickery, you can engage in some of your own.  That's true of every opening repertoire, isn't it?

exigentsky wrote on 07/14/17 at 20:52:47:

This isn't the regular QGA of course, just a move order Avrukh felt the need to use because he tried to combine the Catalan and the 3.e4 QGA in the same repertoire. Maybe that's not really possible, at least not the way he tried to do it. Or Eric is right that White needs to be ready to play QGA e3 lines as well if he wants to reach the Catalan with a 1.d4 move order. Very annoying for White if a rare move order like this causes such problems!

You also need a broad repertoire to exploit this as Black without bluffing; first offering a QGD including both 3.Nc3 lines and Nc3+Nf3 main lines, and then switching to the QGA.

Still, it's a nice catch, assuming you're right that this is a big hole in Avrukh's book.

I guess you're probably right but I don't like to make assumptions.

Although, it seems that opting for a Catalan automatically makes the QGA really good vs you so I'm not sure what to do.

The e4 line from Avrukh's work is nearly with forced moves for White or he stands significantly worse. I looked at deviations even early on and couldn't find improvements even with several engines, engine matches and my own exploration. It seems this is a great way for Black to exploit the move order with Nf3:

1.d4 d5 2.c4 e6 3.Nf3 dxc4!!! 4.e4 b5 5.a4 c6 6.axb5 cxb5 7.b3 Nf6 8.bxc4 Nxe4 9.c5 Bxc5 10.Bxb5+ Bd7 11.Bxd7+ Qxd7 12.dxc5 Qxd1+ 13.Kxd1 Nxf2+ 14.Ke2 Nxh1 15.Bf4 O-O 16.Bd6 Rc8 17.Nc3 Nd7 18.Na4 Nf6 19.Nd2 Ne8

The alternative is 4. e3 and that's fine but I don't think White gets almost anything in those lines. The dxc5 line gives a tiny bit more comfortable end-game and the Bb3 line seems equal but with more play.

BTW: I realize these concerns are more theoretical than practical but I like the process of building a repertoire and thinking about various move orders.

Thank you. Can you just check the references though and tell me if it includes Avrukh's work? Ideally, he'd have looked at it. Although, of course, it doesn't mean that it can't be a great book if he hasn't and it may not be too relevant if it's more of an overview style.

If it's as you say that it's like Sadler's book, that's one of the highest compliments. I already have Avrukh's books so I wonder if you'd recommend this in order to improve understanding of the positions. Are there really in depth verbal explanations of the ideas?