French Advance

I get this kind of position with Black most of the time when I play the French. From my understanding of the opening, the White bishop shouldn't be on b5, especially as most White players go on to trade it, giving up their better bishop. But I still find it hard to get an advantage. After the pawn swap on d4, White is left with strong pawns on e5 and d4. The books say to undermine this centre with f6, but I often find that leaves me with a weakened king position and/or a backward pawn on e6. When I have traded the white-squared bishop, supposedly a good thing in the French, the white squares round the king are hard to defend. What am I doing wrong? Can White be punished for bringing the bishop to b5?

This kind of position is supposed to be common in class play. Basically there's rarely any reason to pin the knight and if White actually does capture, any kind of attack on Black's king will just be a crapshoot. Anyway if what they say is true you'll occasionally win a pawn after 5...Bd7.

I agree with Davydov and I'd add that Bb5 generally doesn't make any sense in most lines of the French (with a few exceptions; there's a line in the Tarrasch where it's normal, and it's also possible in some Nc6 lines of the exchange) because white's light bishop is a very valuable piece, doesn't belong there, and also shouldn't normally be traded for black's knights or black's light bishop, which it's poised to do on b5. Ideally it should be on d3/c2 (hitting h7), or, if that's not possible, e2. Bb5 is not logical and you should welcome such weak moves.

Also, yes you are often supposed to play ...f6 in the French, and the resulting backward pawn comes with the territory, especially in many lines of the Tarrasch. Often black's push e6-e5 can be very liberating and dangerous. That structure is important to understand if you play the French, especially if you play Nf6 against the Tarrasch.

You can play 4Qb6 and then 5Nc6 instead, which avoids this. However, 5Bb5 is not to be feared. RybkaShredder suggestion of 5Bd7 is fine. However, I prefer 5Qb6 and then if 6Bxc6 play bxc5. This gives Black another flank pawn to throw at the base of Whites pawn chain. Plus it allows Ba3, which develops the problem Bishop and prevents Kingside castling, which can be quite annoying for White.

Sure but why avoid it? White either gives up his critical attacking piece or loses a tempo moving it away. White gets nothing from Bb5 here.

Thanks all. I used to know about the Nxe5 trick, but had forgotten it, and must have missed quite a few chances to play it in recent games. And looking at the pawn recapture on c6, I agree it looks strong - again, I've been playing the wrong move.

Actually, 5 Bb5 is not uncommon - there are about 700 games in Chess King's GigaKing Database that reach that position; eyeballing the results (it doesn't total them) it looks like Black won more than White and scrolling through the games there are two typical Black responses with good results, 5...Bd7 and 5 Qb6 where the most common continuation is 6 Bxc6+ bxc6.

But i don't see it in the books either except one I have. It's not in Soltis's Beating the French Defense with the Advance Variation, It's not in Sveshnikov's 2-Volume French Defence Advance Variation, It's not in Eingorn and Bogdanov's Chess Explained: The French. IT IS in John Watson's Play the French where he comments:

"5 Bb5 is generally a mediocre move in advanced structures (barring White having an outpost on e5) because it is White"s good bishop that will be exchanged if Bxc6 follows, and in most cases that strengthens Black's center after ...bxc6.  This position is no exception, and Black has more than one way to try for a minor positional advantage, including 5...Qb6 and

a) 5...Bd7 (threatening 6...Nxe5) 6 Bxc6+ bxc6 7 O-O cxd4 8 Qxd4 c5 9 Qg4 Ne7 (or 9...h5) 10 Nc3 h5 11 Qh3 Qb8 12 Rb1 Nf5 =/+

b) 5...cxd4 6 Qxd4 (6 Nxd4 Nge7 7 O-O Bd7 8 Re1 Nxd4 9 Bxd7+ Qxd7 10 Qxd4 Nf5 11 Qd3 Bc5 =/+) 6...Nge7 7 O-O Bd7 8 Bxc6 Nxc6 9 Qf4 Nb4 10 Na3 Rc8 =/+"

The many database games reaching 5 Bb5 Bd7 (or 5...Qb6) varied considerably with many different move sequences, N's being prominent in most.

Note that both Stockfish 8, set 30 ply deep, and Houdini 2, going 22 ply, preferred 5...Qa5 as Black's best move.  Both engines expected Bxc6+ bxc6 and ...cxf4 cxd4 to occur over the next few moves.

Hope this helps!

5Bb5 is a weak player trade mark, doubt many games with white player FIDE 2000. Goodness knows what they find attractive about move, guess it pins a peice, but it involves white's best minor peice in these structures, so all wrong in principal.

It is one of the most popular questions in opening thread.