I wouldn't say that f3 is "utterly silly" there.
In the database games with 9...d5, White's play is almost uniformly abominable.

Willempie wrote on 10/22/09 at 11:20:16:


I have the exact feeling you do...that 9...Bxc3 followed by 10...d5 should be good for black. Well, I think now that 9...0-0 is best, but surely 9...Bxc3 is OK too!? But I just cannot find a concrete line to support this feeling.

I am not a French player, so maybe my judgement is a bit skewed, but I thought White had an easy-to-play kingside attack in the Qxc6 line. White goes e5, Bd4, Bd3, f3-f4 (maybe a later f5, springing the d4 bishop) Rf1-f3-h3 or g3 and throw the Queen out to g4 or h5. OK, Black will be allowed to play some moves too, but how and where does Black get his counterplay going?   And, not being a French player, that dud on c8 looks pathetic!

12...dxe4 13.c4 Nd7 doesn't really chase the bishop off the a3-f8 diagonal as White has the irritating 14.Bd6.

If 12...Nd7 immediately, then after 13.Ba3 Black cannot play 13...c5 or 13...e5 because he loses a pawn on d5. And if Black prepares ...c5 with 13...Bb7 then 14.Qd4 prevents 14...Nd7. I don't trust 14...e5.

If you can't play d5 afterwards 9 ... Bxc3 doesn't seems to be the right move. 9 ... 0-0 looks better. Rybka likes 9 ... d5 in one go. I'm not sure what to make of that.

But on principle f3 looks wrong in this position.

1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Qc7 6.Be2 a6 7.Be3 Nf6 8.0-0 Bb4 9.f3.

OK, all my sources say that this is last move is inaccurate. I rather like 9...0-0 now, but in a game I went with the principled 9...Bxc3 10.bxc3 and had intended to go 10...d5 when, to my dismay, I didn't know what to do against 11.Nxc6!? In my game, I changed course and played 10...Ne5 when White was better after 11.Qd2 d5 12.Bg5!.

In post-game analysis, I actually found 11.Nxc6 a tough nut to crack.

11...Qxc6 12.e5 Nd7 13.Bd4 with f3-f4 on the way looks nice for White. 

11...bxc6 12.Bc5 is very annoying. Despite White's ruined pawn structure and slumbering pieces, that damn bishop on c5 ruins my smooth development.

12...dxe4 is a pawn sacrifice on White's part, but is likely only temporary. 13.c4 looks like a good move then.

12...Bb7 intends a future ...Nd7 and ...c5 to follow, but 13.Qd4! prevents this.

12...e5, claiming some dark squares, was the best I could come up with. Now if White continues routinely like 13.Bd3 then Black gets the good play he's supposed to get after 9.f3?!, for example: 13...Be6 14.Qe2 dxe4! 15.fxe4 (15.Bxe4?! Nd7! 16.Ba3 f5! 17.Bd3 Qa5 with ...Kf7 and ...Rhe8 possibly to follow looks like a very good game for Black) 15...Nd7 16.Ba3 c5 was my sophisticated idea. Looking at it this morning, I am not sure again. 17.Bc4 Bxc4 18.Qxc4 0-0 19.Rad1 with a possible Rd5 in the future.

13.Qd3 is possibly more dangerous. Then my analysis went 13...Be6 14.exd5 cxd5 15.Ba3 Nd7 (intending Nc5 to enable castling kingside) 16.Qe3 f5! and I think Black is in good shape.

So what's up with this? Is 9.f3 Bxc3 10.bxc3 d5 not all that it's cracked up to be, or am I missing something painfully obvious?