MNb wrote on 07/05/11 at 03:33:32:
Sure White has many options after 6...b5 and sure that famous Kasparov-Topalov is not the final word and sure White has many other options. None of them promises White even an edge. I wish it would or I would return to 4.f3 immediately (I gave it up when I realised that 4.Be3 is more precise).
The main lines with 7. g4 are extremely complicated, calling it just "=" is a substantial overstatement. White has very real chances to try for an edge in the critical lines with 7. g4 due to the extremely complex resulting positions.
Quote:After 6...Qa5 7.Nge2 b5 8.Nc1 White is better prepared for Black's counterplay on the queenside. With the Black King in the centre it isn't a problem that the manoeuvring is slow. The important thing is that White finds good squares for both Knights.
That entire plan is ineffective - black plays b5, Nbd7, Qc7, and white's entire maneuver will ultimately serve as the main source of black's counterplay (the Nb3 isn't just out of play - it's a target). Hallier-Tashkov, E-Mail 2000 is a good example of how little white's entire plan does.
Quote:6...b5 asks the question what White is going to do with those pieces on the kingside.
White's not compelled to do anything immediate with those pieces, preferring to push pawns. 7. g4 h5 8. g5 Nfd7 9. f4 being extremely common, which gives white more typical placement options for the Ng1/Bf1.
Quote:White's play can be improved too.
The "improvement" on move 13 (tested in Geller-Zakharevich, St. Peterburg 2002) is also =+. The entire idea white employed beginning on move 9 is plainly inferior, leaving 8. Bd3 to preference given there's nothing special about 8. 0-0-0 if e5 doesn't work.
Quote:Well, that is my general evaluation too. In Dangerous Weapons Vigus seems to agree that this is critical (and not 6.Nf3 or 6.f3). He might have changed his mind since then; after all this is difficult stuff.
It's an area for investigation, but definitely isn't one to put off black from the entire concept. I've experimented at length against Houdini trying to prove an edge of some sort (engine or human recognized) and it's not happening yet.