Chesstempo database gives white over 65% in all the variations after 6. dxc5 (130 games with both players 2200+). It has been used by Anand and Ivanchuk. So why the main line 6. g3 is way (2800+ games with both players 2200+) more popular where black scores much better?

It has to be because of black's preparation. But what should be the correct response? I don't see one using the database. Nor can I find one using chess engine.

[Event "?"]
[FEN ""]

1. d4 d5 2. c4 e6 3. Nc3 c5 4. cxd5 exd5 5. Nf3 Nc6 6. dxc5 d4 7. Na4 Bxc5 
( 7... b5 8. cxb6 axb6 9. e3  )
( 7... Bf5 8. e3 d3 9. Qb3 )
8. Nxc5 Qa5+ 9. Bd2 Qxc5 10. Rc1

2 Answers 2


6. dxc5 is just a good move, and you cannot really hope to refute it. It was given an exclamation mark in Grandmaster Reportoire 10 by Ntirlis and Aagard, and their main line starts as in your diagram.

Moves such as dxc5 were underestimated by older theory because they were thought to give the initiative to black. Nowadays with computers we can analyse more concretely and see if blacks initiative is something to be afraid of or not. In the case of 6. dxc5, the analysis seems to suggest that white ends up with a small advantage.

  • I'll add that 5...Nf6 avoids this variation and Black should play it. Then White has to go into the traditional main lines.
    – M.M
    Oct 27, 2016 at 5:52
  • how can 5... Nf6 stop 6. dxc5? Now you are down a pawn, d5 is under threat and can't play 6... d4
    – jf328
    Oct 27, 2016 at 9:41
  • @jf328 d5 would be protected by the Nf6, so Black can play 6.dxc5 Bxc5 quite happily. However, 5...Nf6? is considered very dubious because of 6.Bg5! with strong pressure.
    – Evargalo
    Sep 23, 2020 at 9:34
  • @M.M : White doesn't have to go for the main line since 6.Bg5! refutes 5...Nf6? That's why Black normally waits for g3 before committing his king's knight.
    – Evargalo
    Sep 23, 2020 at 9:35
  • 1
    @Evargalo I agree with your assessment , playing 3...c5 pretty much commits Black to the line that OP wants to avoid
    – M.M
    Sep 24, 2020 at 20:48

First, I think your statistics are misleading. 130 games is not a very large sample set, and you didn't give the minimum rating for the players. Are these master games, grandmaster games, or games by players of all ratings? The winning percentage of white might be meaningful if all the games were played by super-grandmasters, but I don't think that's the case.

I think that the 6. dxc5 line is more forcing than other lines. This means that a well-prepared black player might have an easier time accomplishing a draw. The 6. g3 variation is slower, and allows for a more playable game where world-class players like Magnus Carlsen have more room to outplay an opponent.

Let's look at it from a logical point of view, too. White's goal is to isolate black's d5 pawn and attack it. White doesn't have to do this right away. Allowing black to play d4 changes the character of the game. Why not keep things as they are, slowly develop your bishop to a great square on g2, and isolate the pawn when you feel ready? This way, hopefully, we don't allow d4.

In the position you give, 10... Qb6 is the best response according to the chess24 engine. If you play through the line, you see that the pieces get traded off and it is pretty forcing.

  • Hi, I edited the OP, it was 2200+ from both players
    – jf328
    Oct 24, 2016 at 8:21
  • Hm, even that might be misleading. Grandmaster games would probably offer more insight into an opening. The fact that the world's top avoid a line says a lot, too.
    – ktm5124
    Oct 24, 2016 at 18:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.