It's interesting to see differing GM opinions on those two lines after 6...h6 7. h4 Nc6 8. Rg1.  John Emms gave "8...h5!" (which by the way was played by Kasparov against Karpov in their first match), while Drazen Marovic gave "8...d5!" (though he also presented 8...h5 as being quite okay).  Sources that I know of on these lines include Marovic's two books "An Active Repertoire for Black" and "Dynamic Pawn Play in Chess," Andras Adorjan's "Black is OK!"* (mentioned earlier) and a letter by Junior Tay in NIC Yearbook 63. 

*I'm only familiar with the first book of that series; it had a chapter on the pawn sac (by Black) that can arise when White meets 8...h5 by 9. g5.  The book was published in the 1980s; the relevant chapter was called "A Blow to the Keres Attack."  It certainly seems to me that 9. gh has been more common in recent years, thus perhaps vindicating Adorjan.   

I was interested to learn of the upcoming book by Craig Pritchett.  He wrote a book on the Schev for Batsford back in the 1970s, which received good reviews (e.g. by the late US GM Edmar Mednis).

I think Black can also, as well as the line mentioned by Lee Roth, play Andersson's preferred 8...h5. This leads to quite a complex strategic position in which White may have an edge, but he also has the harder position to play. If you're interested in this or just the Schevy in general keep an eye out for Craig Pritchett's forthcoming Starting Out guide (Everyman).

J-dog wrote on 07/22/06 at 02:02:58:

Looking at the foreword of "The safest Sicilian" I guess another candidate is the system with Be2+a4+Kh1 followed by g4-g5, which Delchev and Semkov seem to value highly. As for move orders to reach the Siclian that same foreword mentions the 1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 move order (and then 5.Nc3 d6) as a possiblity to reach the Scheveningen that makes the Keres and English attack less of a threat. But admittedly this is just all from a short paragraph that explains why they would *not* recommend the Scheveningen.

I've tried just about every way there is to reach the Scheveningen and have never really been satisfied with any of them.  I finally decided to take another look at the Keres Attack.  Is it really that bad for Black?

I've convinced myself that the 6..h6/7..Be7 line is Black's best try.  1.e4 c5  2.Nf3 e6  3.d4 cxd4  4.Nxd4 Nf6  5.Nc3 d6  6.g4 h6  7.h4 Be7  when the main line is 8.Qf3 h5  9.gxh5 Nxh5.   

If 10.Bg5 Nc6 11.Nxc6 bxc6 12. 0-0-0, Sobura's 12.. Rb8 seems fine for Black.  I  don't know of any practical examples, but Gavrikov in Experts recommends 10.Be3 instead.   

After 10. Be3 Nc6  11.0-0-0 a6  12.Be2 Bd7  13.Qg2 g6  14.f4 Nxd4  15.Rxd4 Bc6  16.Bf3 Qc7 is critical.  I've looked at this some and think that Black's OK.  For example, White's most direct try 17.f4 gxf5  18.exf5 0-0-0  19.fxe6 fxe6  20.Bg5 Bxg5  21.hxg5 e5  22. Rdh4 split the point in Renet-Razuvaev, Belgrade 1988.

So, any thoughts or comments on the Keres?