Bonsai wrote on 05/02/14 at 08:51:46:


Actually, I'd rate both 18...Rd8 and 18...Ba4 higher than giving away a tempo with 18...Bb5, since White is likely to play c4 sooner or later anyway to make way for Nc3-d5/b5 but since Black may well be able to afford to give a tempo and still equalise it ended up as "my ML". Usually, once the c5-pawn is recaptured the equality is already beginning to lean towards Black imo, due to the pawn structure. So, I doubt this line will see much action otb or in corr.

I'll have a look at your suggestions, though I suspect that you overestimate Houdini's play in this kind of position.

Edit, had a look and:

1) I don't see how 22.Bf4 "indirectly defends c5" after simply 22...Qxc5 taking the pawn, and after a few more moves Black looks fully equal, for example: 23.Nb5 Rd8 24.h3 Qb6 25.e5 Rxd1+ 26.Rxd1 Nc5 =

2) 18...Rd8 ... 21.Bxb7 may indeed give White a tiny edge. Black could try 21...Ba4 22.c6 Bxc6 23.Bxc6 Qxc6 24.Qc2 Qc4 with compensating active pieces, though perhaps not quite full comp...

So maybe 18...Bb5 is best after all, despite giving the tempo - or 18...Ba4, which you didn't reply to, (though it might transposes to Bb5 in some lines perhaps)

I looked at that line, too, but it seems to me the Nc6-b8-a6 idea does not lead to completely clear-cut equality, white can at least try to set black some more problems. E.g. 20.Nc3 might be an interesting try.

TheDragon wrote on 04/29/14 at 17:26:19:

The contacts on this page would seem like the obvious way to contact him: http://www.chesspublishing.com/content/11/feedback.htm
I believe the idea is that the authors of the different section look into questions from subscribers, but obviously I don't know how fast he might respond or whether he will consider the question interesting.

This is definitely a quite critical line, but I wonder whether you/we are being overly influenced by the computer evaluations. It is a rather murky line and I wouldn't know exactly how to play this as black (some kind of Na5 + Bc8-d7-c6 plan looks logical, but I'm no GM), but while Houdini loves white (temporarily an extra pawn, some space advantage etc.), I am not truly sure whether his evaluation is totally right (white also has doubled isolated c-pawns, which are admittedly surprisingly strong). I played around a bit and the computer showed some really poor judgement for black (e.g. attempting to re-take on c5 with ideas like Bg7-f8 even if it means exchanging the bishop, which must be a bad idea).

That's why I was thinking of keeping the bishop on c6 in those positions - it obviously cuts out any c6 freeing moves, but also Ba4 etc. Ok, so white can go Nd4xc6 but maybe that's not too bad.

Finding a truly happy home for the knight seems a bit harder. Dark squared bishop/rooks fine I think.

I am not so sure... I mean, the Bishop on b3 really isn't a bad piece. It can go to a4, and following the freeing maneuver c5-c6 and then c4-c5, the bishop could become very active. Even if not, it performs a good defensive role. Half of the Black pieces are out of play/don't have a good square, like our light squared bishop, queen and knight. 

I really don't know what to do. It seems that the line is just teetering on the edge between 'playable line' and 'just plain dubious.' I was hoping that someone would analyse and see if he can resurrect the Black side with something incredible that I missed. I really do hope so, because otherwise I think I will have to switch lines which is a bit of a shame.

Thanks, 
TheDragon

MartinC wrote on 04/28/14 at 08:27:01:


I looked at the 14...Bd7/15...Ne5 line a little more closely following your post. I think the move I suggested for White, 18.Bxe5, is a little inaccurate and an improvement is 18.Ba4!. Following 18...Qf8 White plays 19.c6!, gets rid of the doubled pawns and has a clear advantage.

Instead of putting the Bishop on a6, you suggest putting the Bishop on c6. It is by no means losing immediately, but the Bishop doesn't do much there and Black is left without a plan. A good idea for White would be ...Bg5, ...f4 followed by ...e5.

By the way, I have computer-checked all these lines with Houdini 4 Pro. Not that I have been led by the engine, though... It just seems to find concrete problems in all Black's plans and Black in all lines seems to be grovelling for a draw without any counterplay.

I have read through the rest of the chapter, and the lines with 10...Qc7 followed by 11...e6 and 10...Na5 with 11... b6 are both very strong. The engine nor I could find any cracks in it.

It is not that Black has a lack of good options at his disposal, it is just that this line has put me off 'The Safest Grunfeld' mainline given by Delchev. It is not really a problem for me, but it is a little chink in an otherwise fantastic book.

Thanks,
TheDragon

gewgaw wrote on 04/27/14 at 21:01:17:

I also had a look at it, but it didn't really impress. 

By the way, the poll is only if you think the mainline is busted in Delchev's book, he offers a number of alternatives. I am talking about 10...Qc7 11. Rc1 Rd8 12. Bf4 Qd7 13.dxc5 Qe8. That is the line I am talking about, I know Black can deviate earlier. I just want to see if we can resurrect this specific line, because I analyse it for about an hour and it is just looking terrible for Black. Either he will have to defend a much worse endgame which I don't know if he can hold, or he will have to fight, down a pawn with no compensation.

Obviously this is all in theory. If someone could find these moves in practice, that is another story. But the question is speaking from a purely theoretical sense, not a practical point of view.

Thanks,
TheDragon

TheDragon wrote on 04/27/14 at 13:30:09:


That is the line you are speaking of. It is played and is playable, but it is far from ideal. I also looked at the rare continuation 11...e6 mentioned in 'The Safest Grunfeld' and it works quite well. But is it really that bad, that Black has to avoid on of the 7.Bc4 mainlines? Guys, also please post your reasons and some analysis following your choice on the poll.

Thanks, 
TheDragon

kylemeister wrote on 04/27/14 at 18:10:28:


Wow, has the move held up?

I started looking at your analysis with Houdini and it came up with a move earlier that looks ridiculous to my human eyes but who knows.  After, 1. d4 Nf6 2. c4 g6 3. Nc3 d5 4. cxd5 Nxd5 5. e4 Nxc3 6. bxc3 Bg7 7. Bc4 c5 8. Ne2 O-O 9. O-O Nc6 10. Be3 Qc7 11. Rc1 Rd8 12. Bf4, it suggests 12...Be5 which I'd guess has never been tested.

Hello there. 

I have been going through The Safest Grunfeld by Delchev and Agrest and I think it really is a fantastic book so far. I have been typing it in, engine-checking and coming up with improvements. Most of the improvements so far have been minor, but I think I have actually run into a problem which ruins the whole line for me. I personally have that you prove me wrong. 

The line that I speak of is: 1.d4 Nf6 2. c4 g6 3.Nc3 d5 4.cxd5 Nxd5 5. e4 Nxc3 6.bxc3 Bg7 7.Bc4 c5 8.Ne2 0-0 9.0-0 Nc6 10.Be3 Qc7 11. Rc1 Rd8 12. Bf4 Qd7 13.dxc5 Qe8 and now the move is 14.Qc2. 

This line is basically passed off in 'The Safest Grunfeld' and 'Chess Developments: The Grunfeld' as harmless, but I think it is pretty killer actually. Both books missed a simple idea and gave the line as 14...Bd7 15. Rfd1 Rac8 with good play based on ...Na5.



To be completely honest, the more and more I analyse the more hope I lose. I also analysed earlier deviations, such as 12...Be5, which is a little better but still not very good. Also please note that this is just a stub of my analysis - a condensed version. 

So, is it just me or will I have to switch systems? I hope you prove me wrong, but I am pretty sure this analysis is accurate. Any improvements for Black are greatly appreciated

Thanks.