French Winawer

Which move is considered best here? cxd4 or Bxc3+ ? Both moves seem to have the same result to me,but do they? Please elaborate :)

About Qg4 i prefer the poison pawn var. just saying...anyway....

Ba5 i think is inaccuracy because of dxc5 and then threatening b5...what do you think?

@bugs_death ,thanks for stating your opinion

5...cxd4 is considered to be a bad move, but white shouldn't recapture

on c3 in the 7th move. Instead of recapturing he should sacrifice

a pawn by 7.Nf3 or even 7.Qg4. Many games have shown that white has

more than a good compensation for the pawn.

Love your idea. Thank you!

In 20+ years of playing the French, I have never seen cxd4.

Ba5 is bad because b4 is a strong positional sacrifice.

Ba5 might be slightly weaker than Bxc3 - however, there have been

several strong players who didn't fear b4, among them the former

world champion Michail Botwinnik. Of course, capturing on b4 is a

mistake after 6.b4, but after 6....cxd4 it's not so clear.

Thank you guys for your insights :)

@DrSpudnik It is played though,but with an approx. 53  % win for W (according to www.chess365.com that is why probably it is not opted and is considered bad (@DJ_Haubi)

5...cxd4 is probably fundamentally wrong (Negi gives it as ?! in his recent book), which is easily correct. Black gives up his key dark-squared bishop for practically nothing (sometimes a pawn if white goes for an ambitious gambit approach), while also opening the position to some extent. In contrast, 5...Bxc3 compromises white's structure and keeps the central tension, representing a MUCH sounder approach IMO

Neither have I. Asked Stockfish 5 about it and said: No fear! I'm going to give it a try as soon as I have the chance. It seems to be perfectly playable at amateur level and a nice surprise for your opponent.

5. ...Ba5 6. dxc5?! Bxc3+! is not anything to get excited about.

The French defence is a particularly concept rich opening. Will get furher studying well annotated games by strong practitioners like Korchnoi etc, rather than trying to play by the database stats numbers.

In speed chess, it may well be a time-killer that makes the difference between a win and a loss.

Concerning 5...cxd4, Yes, I've had to answer to that 5 or 6 times. It's not sound but implies a trap in if white plays 6.axb4 dxc3 7.dxc3?, ...Qc7! gives black a clear edge. White has to sacrifice a pawn and play 7.Qg4 cxb2 8.Bxb2. That's an 80-year-old recommendation btw, nothing new here. White will try to exploit his better development and his opponent's weaknesses on the dark squares. 

5. ... cxd4 is dubious, and rarely seen in games between strong players. The play above is probably best for both sides after 5. ...  cxd4.  8. ... Ne7 gets into trouble after the simple and obvious 9. Bd3, hence the try with f5. 7. Qg4 (first played by Bogoljubov) is strong but 7. Nf3 is simpler.  

Black plays the Winawer to mess up White's structure and get a complex and difficult game. He does not play it to suffer in a cramped position with a serious lag in development.   It isn't possible to demonstrate a forced win for White, but that is not the point.  Playing this way gives Whiten an easy game, and practical results in this line are poor for Black.

Since someone else dug up this thread:

5...Ba5 is not considered an inaccuracy because of 6.dxc5. In such case, I think GM Szusan Polgar suggested in a video of hers that Black can lose a tempo by playing 6...Bxc3 7.bxc3 Qc7 followed by 8...Nd7 after which White's tripled pawns and the e-5 pawn become targets.

However 5...Ba5, at least in several mainlines, leads to bloody sacrifices with crazy tactical passages of arms without clear theory (as far as I know).

It's called the Armenian variation since GM Lputian and GM Vaganian sacrificed some elo points in order to go through it during the nineties. While I modestly worked on it, I studied several games commented over the years by GM Psakis (I think, not sure) and it was clear each new improvement or ideas in this variation bamboozled him quite a bit regarding his previous assessments of the line. At my patzer level, it took me days (with all that previous work by others and the more recent major games in mind) to assess that on best play, White could get a better endgame which would be very hard for Black (let's say deadly if Black has never been seen before).

On the other hand, lines in the Qg4 subsequent variations are pleasant to play with a very stereotyped development and tactical blows/strategical ideas for Black (...Bd7 later possibly followed by ...Bb5 or ...Bc6 after ...d4 in some variations, ...Ndf5, ...Ncd4, rooks on semi-open -g & -h files, ...Qa5 or ...Qc5 if required, ...Rg4 if the d4 Knight cannot be supported by the Queen, weakness of White's -c pawn, ...etc) and a lack in development (and unseasy with White castle being a huge target) from White side for two sacrificed pawns.

Also, White can kill all that fun by simply playing an early Nb5 instead of going for the Qg4 lines (those winning g and h pawns).

I think Bxc3+ is the best and most solid and logical option

Because Black has many options to continue then (...Ne7+0-0, ...Ne7+Qc7, ...Qc7, ...Qa5-a4, Qd7)

5 cxd4 is a chapter in Watson's Dangerous Weapons in the French defence