In terms of raw theory needed to survive, perhaps not so much (except vs 7 Qg4 ) because the positions are a tiny bit forgiving of oddities.

But the positions really do have a very individual character - really nothing like the Nimzo (or much else!) - so you really do need to play over some games etc. Much easier if you're already playing other French lines of course.
(if not there's a lot more than 3 Nc3 to prepare!)

This is the 1st post after 3 years of silence in the thread.

How much time does it take to learn the Winawer from scratch?
(At first only with sub-variations, such as 4..b6 and 4..Ne7. BTW, the Nimzo-Indian is also a favorite in my repertoire.)
The aim is to know it well enough for me to win against most of the u2200-opponents. What are the most frequent White tries at this level?
I have access to books by Uhlmann, Watson, Vitiugov, and Moskalenko.

As i remember the poisoned pawn variation is not good for black. There were some articles in NIC Yearbook about that. The variation with Ne7 and 0-0 also do not promise much chances for Black. By the way that are some of the reasons why I gave up the French with black a year ago. Now I am very happy with the Classical Sicilian and Richter-Rauzer 

The usual argument against 8 .. c4 is that it's just a slightly worse version of the main lines. It certainly looks that way to me. Nothing dreadful but I can't see any reason to prefer over the near equivalent main lines.

Normally to get this pawn structure white has to play Qg4 -> h5 to provoke Ng6, then Be3 to force c4 and then Qg4 again after BxN and fg. 

Compared to this black has lost a tempo here put kept his Q on d8 and white's knight on g1. If anything I suspect that this is slightly advantageous for white too - the normal manouver is Qc7 - > f7 then to e8 after Ng5 which then goes to h3 etc.

This is a wash in tempo terms. However it does perhaps give white a little extra flexibility - iirc there is even one line that Khalifman recommends where white plays Nf3 -> g1 of his own volition just to prevent Rxf3 ideas. 

He can also perhaps retreat his bishop on d2 where it's a little more secure and props up the Q side. Not sure if taking that square away from the king is a good move though.


The lines I want to work here are the ones with 8.. Nbc6 and 11.. Nce7. Then the 12.. Nf5 stuff is horribly tactical but might be objectively surviving reasonably. 

I've thought of trying a small move order finesse with 12 .. cd 13 cd Qc3+ 14 Ke2 Nf5 (to avoid certain lines) but 15 Rb1!?(!) might render this a bad idea. 

The line which is frustrating is 12 .. Bd7 because then after Karjakin's13 a4 black can go 13 .. c4 14 Bxg6 fg etc which looks like it should be an improved version of the main line. The bishop can even reach e8. The thing is that it doesn't seem to quite work like that for some reason. Hard game sometimes

sterpfi wrote on 05/15/08 at 08:17:08:


I agree, I don't think Black's results with c4 are good anyway...

Also look at 8.Bd3 c4. There is little theory on it, but imo it is heavily underestimated.

ArKheiN wrote on 05/11/08 at 20:03:52:


My understanding is that 8.Bd3 is considered best.  Why I do not know.