Ah I see, thanks realpolitik and dom. Yeah, I suspected sometime before that Qxg7 is not the main way to go. But I have never yet played the Winawar other than several punts at it so my knowledge of it is rather shallow. It doesn't help that John Watson focused on Qxg7 in Dangerous Weapons...but no fault of him to do that. However, as you can see from my analysis, I did put some effort to convert myself to the Winawar...with his advocated "A Surprise in the Winawar" but again that ran into problems in my analysis, though from Black's point of view this time. So for now I give up trying to go to the Dark Side of the French Defence  . 

Though anyone who plays the Winawar and can put up a convincing argument of why it is more attractive instead of the Classical structure, feel free to enlighten me even more. I am quite easily persauded...unless you are telling me to suicide into the poison pawn variation proper without a higher, noble cause. 

Anyways, Zatara, I revise my view point. Seems that 5. Qg4 seems very viable, unless you have the pawn grabbing syndrome.

so you do or don't recommend 5.Qg4??  What about the 2 knights thanks,
Zatara.

I analysed this some years ago when I played the French 2N quite often. The win after 8... 0-0 is not trivial but I think it is there:

8... O-O 9. Bxh7+! Kxh7 10. Ng5+ Kg6 (Only move)  11. Qd3+ f5 and now:

a) 12. exf6+?! Kxf6 13. O-O-O Qa5 14. Nxd5+ exd5 15. Rhe1 Nd4 16. Nh7+ Kf7 17. Ng5+ Kf6 18. Nh7+


---------------------------------

@Paddy:

I believe 12 ef6 to be a !
If white plays a set-up with Qg3/Rhe1 and Bc7/Be3/Bd6 all will go to +-

Analysis:

12 ef6: Kf6:
13 0-0-0 
a) 
13........Qa5
14 Qg3! Nb4
15 Rhe1 +-
or
14.....Bb4
15 Rhe1 Nb6
16 Rd3 +-
or
14......Nb6
15 Rhe1 Bd7
16 Qh4! e5
17 Rd5! +-

16....Ba3
17 Ne6 Kf7
18 Ng5 Kg6
19 Qh7 Kf6
20 Nce4 +-

ofcourse you dont have to play 13..Qa5

Instead of this:

b)
13....Nb6
14 Rhe1 b1)e5
15 Be5 Kg5
16 Qh7 Ne5
17 Qg7 Ng6
18 Ne4 +-

b2)
14....Bb4
15 Qg3! Qe8 ( 15..Bc3 16bc3! Qe8 17Rd3 Nc4 18Nh7 Kf7 19Bh6 Ke7 20 Rd5)
16 Nb5 Qh5
17 Be5 +-

and
b3)
14...Qe8
15 Qg3 Qh5
16 h4! Rh8
17 Be3  Be3  (e5 18 Nce4+-)
18 Re3 Qh4
19 N G e4  +-

c)

13....Nb4
14 Qh3 Qe8
15 a3 Rh8  (..a5 16 Qg4)
16 Qg3+-

realpolitik, I'm not sure about Davies but Zatara and I were talking about Watson recommending 5. Qg4 as one of the dangerous weapons in his latest French book. 

But dangerous for whom? Well, move aside y'all for SniperOnG7's first TN on ChessPublishing!   
Whether my assessment found in my first post in this thread is harsh or not is up to you to decide...
1. e4 d6
2. d4 d5
3. Nc3 Bb4
4. e5 c5
5. Qg4 Ne7
6. dxc5 Nbc6
7. Qxg7 Rg8
8. Qxh7 d4
9. a3 Qa5
10. Rb1 dxc3
11. Be3 Bd7! (this move is found in the example game on p69 between Flipper and JEB_Stuart, Playchess 2003).

Now White has three pawns but Black is a piece up, while White's pawns can easily fall. Black has far better development (just look at White's Kingside!) and 11...Bd7 is meant to pronounce this contrast even more. 

After looking at the position for a while, I believe the most important element of the position is that White can capture back the piece but not without losing his pawns or letting Black accessing important squares. For example, if 12. axb4, 12...Qa2! is the winning shot. Basically it gains control of the queening square with tempo attack White's Rook. Thus play may continue: 
13. Rd1 Nxb4
14. Nf3 (trying to develop) cxb2
15. c3 Rg6! and Black wins

or 14. Bd3 cxb2
15. c3 0-0-0!!
16. cxb4?! (White is losing pretty bad anyway) Bb5! <deflection (or is the term decory??) from b1 square>
17. Bb1 Rxd1+
18. Kxd1 Rd8+
19. Bd2 Qc4! with mate threats on f1 and c1 simultaneously. 

or 
16. Bb1 Qc4 (or Ba4 if you want to be cool 17. Bd4 Qb3 18. Rd3 Rxd4!! 19. cxd4 (decoyed away from the c-file) Qc4!! again and Black wins)
17. Qe4 (forced in regards to ...Bb5) Qxe4
18. Bxe4 Bc6! (again deflection away from b1 square)
19. Bb1 Rxd1+
20. Kxd1 Rd8+ and Black wins since White lacks immediate reinforcements.


Okay, this is just the background, which is necessary to persuade us that Black's Bishop on b4 is quite immune. Obviously JEB_stuart knew this, allowing him to play 11...Bd7!

Now in Watson's book, White played 12. Nf3 and Black's response of 12...cxb2+ is poor. Firstly, since the Bishop is immune, why spend time on it? Is it not better to continue development in order to maintain the initiative? Secondly, after 12...cxb2+ 13 axb4, suddenly the position has equalized, if not slightly better for White. For, without the wedge on c3, White can play c2-c3 whenever which gives Rb1 support. Thus the situation has now totally reversed. Whereas beforehand Black's Pb2 was strong and White's Rb1 was pathetic, now Pb2 is just a weakling. 

Therefore my improvement is very logical: 12...0-0-0!TN As we concluded importantly before, Bb4 is actually very comfortable. Meanwhile, by maintaining a dark-square blockade on c3 and supported by Bb4 and Qa5, White's Rb1 is very isolated. For Black, he remains a solid piece up, has many pawn targets, has connected Rooks, has the initiative, has potential passed pawn on b2 etc etc. You can play around with the position but you will find that Black's superior development and assets give the second player a solid advantage. Consequently from this I conclude that Black has a really good time in the mainline 5. Qg4 variation. 

BTW, if you find something good for White or disagree with my analysis, please feel free to absolutely trounce it - it takes two to Tango. Though in this case, a French kiss might be more appropriate.  

MNb wrote on 03/07/08 at 02:18:49:


I analysed this some years ago when I played the French 2N quite often. The win after 8... 0-0 is not trivial but I think it is there:

8... O-O 9. Bxh7+! Kxh7 10. Ng5+ Kg6 (Only move)  11. Qd3+ f5 and now:

a) 12. exf6+?! Kxf6 13. O-O-O Qa5 14. Nxd5+ exd5 15. Rhe1 Nd4 16. Nh7+ Kf7 17. Ng5+ Kf6 18. Nh7+

b) 12 Nxe6!? Qb6 13 Nxd5 Bxf2+ 14 Kf1 +- Baker but Fritz's 12...Qh4 is less clear (to me anyway).

c) 12. Qg3 and now:

I. 12...Bxf2+ (! Baker) 13. Kxf2 Ndxe5 (13... Qb6+ 14. Kf1 Qxb2 15. Re1+-) 14. h3!  (not 14. Nxe6+? Ng4+ !15. Qxg4+ fxg4 16. Nxd8 Rxf4+ Baker) 14... Qb6+ (14... Kf6 15. Nh7+) 15. Kf1 Qxb2 16. Re1 Nc4 17. Nxe6+ Kf7 18. Qxg7+ Ke8 19. Qxf8+ Kd7 20. Nc5#

II. 12...Ndxe5!? 13. Nxe6+ Ng4 14. Nxd8 Bxf2+ 15. Qxf2 Nxf2 16. Nxc6 (16. Kxf2 Rxd8 17. Rad1+=) 16... Nxh1 when White can choose between Ne5+, Ne7+ and the simple 17. Nd4 - I think the Nh1 is doomed, but it is still not trivial; one possible line now is 17. Nd4 Re8+ 18. Kf1 Bd7 19. Kg1 Rac8 20. Rd1 Rc4! 21. g3 (not 21. Kh1 Rxd4!) 21...Nxg3 22. hxg3.

Conclusions:
1) No-one should play 8...0-0? as Black. There are several far better ways to face the French 2N/Steinitz.
2) Anyone playing this variation of the Steinitz for White needs to analyse the consequences of 8...0-0? 9 Bxh7+! Kxh7 10.Ng5+ Kg6 very carefully for themselves, since even with computer assistance it is not simple, so if you faced this over-the-board unprepared it would involve a lot of difficult calculation and assessment.

MNb wrote on 03/07/08 at 20:19:40:

Which is the Geller gambit? 
I personally would suggest this line against the Winawer: 1.e4 e6 2.d4 d5 3.Nc3 Bb4 4.e5 c5 5.a3 Bxc3 6.bxc3 Ne7 7.Qg4 
For some reason black lately always goes 7..0-0, which imo is completely unsuitable for club players playing black.

There is a choice:
1. bite the bullet, play 3.Nc3 and chose some aggressive main lines, like the Alekhine-Chatard and the Geller Gambit vs the Winawer.
2. play the Kortsjnoj Gambit also called Universal System: 3.Nd2, c3, Bd3, Ngf3, 0-0. Black can snatch pawn d4 but will suffer from bad piece coordination for a very long time.
3. chose 3.Be3. This does not promise an advantage, but is not that bad.
4. try 2.b3.
Some will promote the KIA (2.d3, 3,Nd2), but I don't have much faith in it.

MNb, you are right that there is no winning attack: after 8...0-0 9. Bxh7! Kxh7 10. Ng5+ Kg6, what White has is 11. Qd3+ f5 12. Nxe6 Ndxe5 13. Qg3+ Ng4 14. Bxd8 Bxf2+ 15. Qxf2 Nxf2 16. Kxf2 when White has a safer King, better Bishop and better pawn structure. Black can still fight...but it will be a long suffering.

However, you are also right in saying that 8...f6 is not the only move. In fact, 8...h6!? is rather interesting, preparing ...g5. I had a correspondence game (obviously as Black) in this variation once. It resulted in White queenside castling (while Black's stayed in centre). Typical of the modern French, the position provided me the opportunity to attack on both flanks with the goal of undermining White's centre and attacking his King simultaneously. It was a very exciting and chaotic game with me on the attack and gaining a sizeable advantage. Again, typically (but of me this time), I made a bad move after he surprised me with a move I didn't think about while analyzing. I reacted badly, which lost my advantage and the game at the same time  .