MartinC wrote on 06/29/11 at 21:25:13:
Ah, not that its harmless, but I meant getting move ordered into some rather suboptimal lines of the English while trying to dodge the Keres:
cf 1 e4 c5 2 Nf3 e6 3 d4 cd 4 Nxd4 Nc6 5 Nc3 d6 6 g4 h6!? 7 Be3 Nf6?! 8 f3 which really isn't good. That sort of approach could easily be critical for the 6 .. h6 stuff.
Oh... I agree completely. I think it is more accurate to delay ...d6 a few moves, e.g. 5 Nc3 Qc7 6 Be3 (6 g4 h6 7 Be3 Nf6 8 f3 d5) a6 7 f3 Nf6 8 g4 h6 9 Qd2 b5! Since the modern way to combat g2-g4-g5 is ...d5, we need to preserve the option of playing it in one move against systems that depend on that motif (Keres, obviously, but also the English and the Be3/a3 system in the Taimanov).
...d6 is an important and thematic move, and the whole point is to transpose to a Scheveningen, but not at the cost of being worse. I know the 5...d6 move order is played often, but unless you really think you aren't going to play ...a6 and ...Qc7, it isn't essential to play it on move 5.
(...Qc7 isn't exciting against the English attack either, necessarily, but there are no perfect move orders.)