Can you clarify what is meant by the "modern treatment of the Smyslov variation"?

I take it this is 9..Qd7 or 9..h6?

Yes, the Petrosian is the system with 12...Bd7 which is dealt by Marin in his book. And yes, the modern treatment of the Smyslov looks quite interesting.

I recently drew against a higher-rated player with the 5.0-0 Bg4 6.h3 h5 line (like Arcticmonkey I doubt that Black is close to losing with best play, though it might be a rather strong += if White really knows what he/she is doing).  White played 7.d3 Qf6 8.Bxc6+ (in Svidler-Grischuk, 2008, Svidler played 8.c3 immediately here, but as Arcticmonkey discussed above 8.Nbd2 is probably more testing) 8...bxc6 9.c3 and I opted for the interesting endgame that results from the double exchange on f3.  The 7.d3 line is strongly reminiscent of White's optimal continuation against the 5...Bg4 6.h3 h5 line of the Exchange Ruy.

If I felt a need to give up on this line I would revert to using a different main defence to the Ruy- if I wanted to get a Closed Lopez type position I would just head straight for the Closed.

As for the 5.d4 line, I've caught a couple of players out with the Noah's Ark Trap line, and I agree with the 8...Bb7 being better than Taylor suggests.  White might be able to get enough compensation for a pawn, but that's all.

IMO 8...Bb7 is a serious alternative to play for win. Taylor dismisses to quickly 9.cxd4 Bxe4 because of 10.0-0 because of "a typical Open-Game e-Line attack". That maybe true but Black is a pawn up. It seems to me that this evaluation is superficially.

In the databases (i have Hugebase and BigDatabase 2011) only 5 games are know. The score is +3=1-0 in favour for White but an e-line- attack played no role in any game. In one game between lower rated players Black made an error in the 10th move and in the other games Black had at the end of the opening the better position or the position was at least unclear.

The "Chess Openings Encyclopdia 2010" (by Convekta) lists  the variation (after 8...Bb7)  as unclear (based on Velimirovic - Langeweg, 1988) or equal (Oll - Psakhis, 1987 - NCO sees here after 13.Qe2 an unclear position).

So i think (after a superficial look at the games) Black is in a very comfortable position. When he plays against a higher rated player he can choose the drawish line or force White to play a questionable gambit and when he is prepared he can kill the elo difference. Against lower rated player Black can choose 8...Bb7 and should be more comfortable because White has no real model games and if White is better is very, very questionable.

What has Taylor to offer in this line:

5.d4 b5 6.Bb3 Nxd4 7. Nxd4 exd4 8. c3

This can lead to dray by repetition and the varation is often seen as drawing line. Has Taylor analysed 8....Bb7 ?

I don't think 7.d3 Qf6 8.Nbd2 Nge7 9.Re1 Bd7 (probably best, Black can leave the bishop a bit longer with 9...Qh6 but White soon drives it away under more favourable circumstances, e.g. 10.Bb3 g5 11.Nf1 and Black gets driven back) 10.Nf1 Qg6 11.Kh1 is too bad for Black, at worst White may have a small edge.  Black can continue with 11...0-0-0, e.g. 12.Ne3 f5 13.Nh4 Qf6 14.Nhxf5 Nxf5 15.exf5 Qh4 16.c4 Qxf2 17.Rf1 Qh4.  11...f5 may also be playable, e.g. 12.Bg5 fxe4 (I'm not sure about the computer's suggestion 12...f4 as this closes lines on the kingside) 13.Rxe4 0-0-0 and Black unravels with ...Kb8 and ...Re8. 

I also find these lines rather more promising (at least in the sense of giving the sort of combative game that Black is typically after when embarking on this risky variation) than 7...Bd7, when after 8.Nc3 I think Black's kingside ideas are too slow.

Discussions earlier in the thread suggested that the "Bishop Defence" merely amounts to a modification of the Closed Lopez, with the bishop on g7 and rook on f8 rather than the rook on e8 and bishop on f8.  It's a legimitate way to play as Black, but not a good idea IMHO if Black wishes to avoid getting a Closed Lopez type of position.

Going back to 5...Bg4 6.h3 h5 7.d3, Grischuk played 7...Qf6 8.d3 Bxf3 (unfortunately 8...Nge7 9.hxg4! doesn't quite give Black enough for the piece here) 9.Qxf3 Qxf3 10.gxf3 with a roughly equal game (the bishop pair offsets the doubled pawns).  Haven't checked 8.Nbd2!? though which avoids the doubled pawns, though it isn't as critical as the main line.  I think it's an exaggeration to say that Black is lost in the main line of the Yandemirov variation, but White does have at least a couple of promising continuations that may yield a significant advantage.

NeverGiveUp wrote on 08/11/11 at 08:36:41:


Thats quite a confident assessment of the position. So you think that the inclusion of ...Bg4 - h3 favours black!? as i suppose there is some sort of a lever to attack, On the other hand though, the g5 square is weaker, although it shouldn't matter too much as your planning ...Nge7 and probably ...f6 though anyway.

As far as 7...Qf6, I think the best line is the simple 8.Nbd2 Nge7 9.Re1 Bd7!? (seems depressing to have to retreat) 10.Nf1 Qg6 11.Kh1 with an edge. White will follow up with Ne3, taking a bit of sting out of f5 and perhaps the computer suggestion of c4!? at some point, putting the pawns on light squares since it is semi-difficult to avoid exchanging the light squared bishops.

As for 7...Bd7, this just isn't the move i want to play against such a normal move as 7.d3. I also dont see how you will get a kingside attack after a normal sequence like 8.Nc3 Nge7 9.Be3 Ng6 10.Re1. White just seems to have an edge here, planning d4 at some point.

Regarding the mainlines, i agree. Black hasn't really demonstrated that he can equalise here, perhaps a slightly worse ending at best if White plays good moves.