LeeRoth wrote on 07/15/08 at 03:02:19:

It is getting the a8 rook into the game that is the problem. Even in the line quoted above black's problem is the a8 rook. While i'm not saying the computer is right, I thought that line was simply too insane not to post. My point was there is clearly ideas to be tried here - it hasn't been tested in TWO decades. I think Bxc3 is a good try (though i'll never play this line cause I steer away from both 8... Nbd7 and 9... Qb6). In all the games with 14.. Re8 15. Qh3 15... Bxc3 is never tried either so its still a different ball park. 

The main idea with 14.... Bxc3 is to prevent this Qh3 idea where by the knights get traded on d5 opening the cfile. Here the knights have already been traded with the ever important c-file remaining closed. This keeps black's king a tiny bit safer. (14... Re8 15. Qh3 Nd5 ( I guess 15... Bxc3 could be tried here as well) 16. Nxd5 cxd5) At any rate 15... Nd5 is clearly bad because it gives white exactly the desired effect... an exposed king.

Also your suggestion 16. Qh3 is now white is down a piece... (16... Bxb2 or Bb4) the black king isn't truly exposed just has a bit of trouble getting the a8 rook into the game. Not saying Bxc3 wins or Qh3 loses... but its an idea. This thread seemed mostly focused around spawning new ideas so this is mine.

Personally I agree 8... 0-0 is the far more natural continuation than Nbd7. However, I don't even understand 9... Qb6  at this point I think 9... a5 critical! You don't want white to get in a5 and slow your queensdie. Your concession with the center dxc4 was so that white would concede back the b4 square with a4 creating a nice place to outpost a piece. Now is the time to secure the bishop in its place.

while I can only find a handful of games with 9... Qb6 10. e4 Bg6 11. Bxe6 fxe6 12. a5 Bxa5 13. Qxe6+ Kd8 14. e5 ... (all of them amazingly won by white!) I must say I haven't found one that wasn't from before 1990.... so this line hasn't been tested in nearly two decades. I must say upon first examination I do prefer black in it. This is the type of position I expect from the slav. I see only that 14... Ne4 and 14... Re8 have been tried.  I was wondering about maybe 14... Bxc3 so I put my idea into the computer and it spit out this crazy line 15 exf6 Re8 16 Qc4 Bxb2 17 fxg7 Bxc1 18 g8=Q Bh6?!  now if that isn't insanity I don't know what is! The computer actually thinks this is 100% equal! Two queens on the board but white can't really keep its second queen. example. Rfe1 Rxg8 Qxg8+ Bf8

CheckMate wrote on 07/14/08 at 18:49:49:


Black may want to try to retain the possibility of using the h-file in the event of Nh4xg6.

Thanks for that, Matemax.  I like the Bc5 mate idea.  I dug Play the Slav out of a box, and learned that Burgess attributes 11.Be3 to Ribli.  Burgess goes on to say that Ribli's evaluation of +/= "isn't conclusive, especially since the pawn grab 11..Bxc3 12.Qxc3 Ne4 just looks unclear."  But that's as far as Burgess goes; there's no further analysis.  So I guess we're on our own.  After 12..Nxe4 13.Qc2, I agree with you that 13.. Bxf3 14.d5 is logical, but maybe here Black can play 14..c5.  I looked at 15.dxe6 fxe6 16.gxf3 Nef6 17.Rfd1 and I'm not sure what's happening.  If 17..0-0 then 18.a5 Qc7 (18..Qc6 19.Qa4) 19.Bxe6 gets the pawn back and leaves White with the bishops.  But also the doubled f-pawns, so maybe this is OK for Black.  What do you think?

After 1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 dxc4 5.a4 Bf5 6.e3 e6 7.Bxc4 Bb4 8.0-0 Nbd7 9.Qb3 the move 9..Qb6 is considered risky for Black because of 10.e4



In this position, both 10..Nxe4 11.Nxe4 Bxe4 12.Bxe6 and 10..Bg6 11.Bxe6 fxe6 12.a5! Bxa5 13.Qxe6 are good for White.  But what about 10..Bg4.  ECO gives 11.Be3 and stops there with an evaluation of +/=.  

Is this right? I suppose 11..Bxc3 12.Qxc3 Nxe4 13.Qc2 comes next, when White has some comp, but is it really worth more than a pawn?