Here: http://www.chesspub.com/cgi-bin/yabb2/YaBB.pl?num=1170253141

is a discussion that belongs where it is, but which touches closely upon the Saemisch Variation, which I think may be a better try for White than its current popularity would indicate.

Another related question. Is there any concrete reason that after 4 Qc2 d5 5 a3 BxN+ no one plays 6 PxB?

Ok black can win c4 with 6 ..dc but 7 e4 b5 8 a4 c6 seems to offer some compensation. So maybe 6.. c5 7 e3 etc is safest - white wouldn't I think normally play Qc2 this early if at all. Still it's maybe not enough to ruin his position?!

I'd guess this is perhaps only worth playing as a surprise to get black into a pawn structure he doesn't normally play (since the queen is slightly misplaced on c2).

Of course it really would also be a painfully unthematic move to play :)

@ Markovich: Isn't Black at least OK after 4.a3 Bxc3+ 5.bxc3 c5 6.e3 (6.f3 d5 is the 4.f3 system) Nc6 or b6 ? There have been some Jussupow-Karpov games in the past.

@Martin C: Euwe-Colle, Amsterdam 1928, went 4.a3 Bxc3+ 5.bxc3 0-0 6.Qc2 d5. I have also found games by Yanofsky, Lilienthal and Pachman - not exactly nobody. But they all only played it once. I suppose one idea is to play Bg5.

MNb wrote on 02/01/07 at 20:38:25:

@ Markovich: Isn't Black at least OK after 4.a3 Bxc3+ 5.bxc3 c5 6.e3 (6.f3 d5 is the 4.f3 system) Nc6 or b6 ? There have been some Jussupow-Karpov games in the past. @Martin C: Euwe-Colle, Amsterdam 1928, went 4.a3 Bxc3+ 5.bxc3 0-0 6.Qc2 d5. I have also found games by Yanofsky, Lilienthal and Pachman - not exactly nobody. But they all only played it once. I suppose one idea is to play Bg5.

Well, a lot of people would say that Black is more than O.K. in several lines against 4. a3.  But in practice life is not always so easy.  If memory serves, the latest idea for White is to deviate from Yusupov-Kasparov (after  5...c5  6. e3 Nc6  7. Bd3 0-0  8. Ne2 b6  9. e4 Ne8  10. 0-0 Ba6  11. f4 f5 -- I hope I have all that right) with  12. d5 Na5  13. e5, followed soon by d6.  White lets go of his c-pawn but makes it difficult for Black to coordinate his pieces.  This seems like a promising idea.  In some lines I looked at, White invades Black's queenside.  That's a switch, for the Saemisch.

In many lines of the Saemisch, the c-pawn simply has to go.  Once the doubled c-pawns are on the board, the c4 pawn is pretty worthless anyway, its main purpose being to exchange itself if Black tries ...d5.

As for Martin's question, the queen on c2 is not very useful in these Saemisch positions.  White, already behind in development, needs every tempo he can get to mobilize his big center before Black blockades him all to blazes (as Karpov did to Yusupov).

Jussupow,A (2645) - Karpov,A (2725) [E29]
Linares 11th Linares (11), 02.1993

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 c5 5.Bd3 Nc6 6.a3 Bxc3+ 7.bxc3 0-0 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.Ng3 g6 13.Be3 cxd4 14.cxd4 d5 15.cxd5 Bxd3 16.Qxd3 fxe4 17.Qxe4 Qxd5 18.Qxd5 exd5 19.Rac1 Rc8 20.f5 Nd6 21.fxg6 hxg6 22.Rxf8+ Kxf8 23.h4 Nc4 24.Bg5 Nxd4 25.h5 gxh5 26.Rf1+ Ke8 27.Nxh5 Nxa3 28.Ng7+ Kd7 29.Rf7+ Kc6 30.Rxa7 Nac2 31.Bf6 b5 32.g4 b4 33.Ra2 b3 34.Rb2 Kc5 35.Nf5 Rg8 36.Nxd4 Rxg4+ 37.Kf2 Nxd4 38.Bxd4+ Kxd4 39.Rxb3 Re4 40.Ra3 Re8 0-1

12. d5?! is refuted easily exd5! 13. cxd5 Bxd3 14. Qxd3 fxe4 15. Qxe4 Nd6 when Black is better. So White would have to start with 12. e5 but after d6 13. d5 leads nowhere because of  exd5 14. cxd5 Bxd3 15. Qxd3 dxe5

unknown_master wrote on 03/09/07 at 12:01:06:

Jussupow,A (2645) - Karpov,A (2725) [E29] Linares 11th Linares (11), 02.1993 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 c5 5.Bd3 Nc6 6.a3 Bxc3+ 7.bxc3 0-0 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.Ng3 g6 13.Be3 cxd4 14.cxd4 d5 15.cxd5 Bxd3 16.Qxd3 fxe4 17.Qxe4 Qxd5 18.Qxd5 exd5 19.Rac1 Rc8 20.f5 Nd6 21.fxg6 hxg6 22.Rxf8+ Kxf8 23.h4 Nc4 24.Bg5 Nxd4 25.h5 gxh5 26.Rf1+ Ke8 27.Nxh5 Nxa3 28.Ng7+ Kd7 29.Rf7+ Kc6 30.Rxa7 Nac2 31.Bf6 b5 32.g4 b4 33.Ra2 b3 34.Rb2 Kc5 35.Nf5 Rg8 36.Nxd4 Rxg4+ 37.Kf2 Nxd4 38.Bxd4+ Kxd4 39.Rxb3 Re4 40.Ra3 Re8 0-1 12. d5?! is refuted easily exd5! 13. cxd5 Bxd3 14. Qxd3 fxe4 15. Qxe4 Nd6 when Black is better. So White would have to start with 12. e5 but after d6 13. d5 leads nowhere because of  exd5 14. cxd5 Bxd3 15. Qxd3 dxe5

That's far from a refutation, it seems to me.  Though I agree that Black's game seems easy enough after 15...Nd6, I don't think he is better.  Still, it's curious that J. Polgar didn't play this way against Milov at the FIDE W.Ch. in Moscow, 2001; or Gyimesi against Cebalo, Croatia 2002.  

It's always difficult to find the correct way if you're surprised at the board with a new move. But if the variation was that strong Milov surely would have tried it again.

Birarov,V (2518) - Joao,N (2515) [E29]
Veinger mem-A corr ICCF corr, 2002

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 0-0 5.a3 Bxc3+ 6.bxc3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qf3 Ne5 17.Qh3 Qf6 18.g4 Nec4 19.Qd3 Rae8 20.Ng3 Na5 21.g5 Qf7 22.Rb1 c4 23.Qf3 Nf5 24.Nxf5 Qxf5 25.Be3 Re4 26.Rbd1 Rfe8 27.Bd4 d6 28.Qg3 Nb3 29.Bf2 Re2 30.Rfe1 Qc2 31.Rxe2 Qxd1+ 32.Re1 Rxe1+ 33.Bxe1 Nc5 34.Qe3 Kf7 35.h4 Qxd5 36.Bf2 Qd3 0-1



Pitterson,J (2221) - Oms Pallise,J (2496) [E29]
Turin ol (Men) 37th Turin (10), 21.05.2006

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 c5 6.e3 Nc6 7.Bd3 0-0 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qd3 Na5 17.f5 Qf6 18.Ng3 Rae8 19.Ra2 Nb3 20.Bf4 c4 21.Qf3 Nxf5 22.Raf2 Nxg3 23.Qxg3 Qg6 24.Bd6 Rxf2 25.Qxf2 h6 26.Be7 Kh7 27.d6 Nc5 28.Qd4 Qe4 29.Qf2 Qe6 30.Re1 Qd5 31.Qe2 Nd3 32.Qe4+ Qxe4 33.Rxe4 b5 34.Kf1 Kg6 35.g4 h5 36.gxh5+ Kxh5 37.h4 Kg6 38.Rg4+ Kf7 39.Re4 a6 40.Ke2 Nc5 41.Rf4+ Kg8 42.Ke3 Rc8 43.Rf5 g6 44.Rg5 Kf7 45.h5 gxh5 46.Rxh5 Ke6 47.Kf4 Nd3+ 48.Ke4 Rc5 49.Rh6+ Kf7 50.Rf6+ Kg7 51.Rf8 Rh5 52.Ra8 Nc5+ 53.Kd4 Ne6+ 54.Ke3 Rh3+ 55.Kd2 Nc5 56.Ra7 Kf7 57.Rc7 Ne4+ 58.Ke1 Ke6 59.Ra7 Nxc3 60.Rxa6 Nb1 61.a4 b4 62.Rb6 Rh1+ 63.Kf2 b3 64.a5 Nc3 65.Bf8 Nd5 66.Rb5 b2 0-1

I'll leave it to you to find an improvement over these games. I think this is exactly the kind of position Black is aiming for in the Sämisch and I find it hard to imagine that White should not be worse here.

unknown_master wrote on 03/10/07 at 23:17:48:

It's always difficult to find the correct way if you're surprised at the board with a new move. But if the variation was that strong Milov surely would have tried it again. Birarov,V (2518) - Joao,N (2515) [E29] Veinger mem-A corr ICCF corr, 2002 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 0-0 5.a3 Bxc3+ 6.bxc3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qf3 Ne5 17.Qh3 Qf6 18.g4 Nec4 19.Qd3 Rae8 20.Ng3 Na5 21.g5 Qf7 22.Rb1 c4 23.Qf3 Nf5 24.Nxf5 Qxf5 25.Be3 Re4 26.Rbd1 Rfe8 27.Bd4 d6 28.Qg3 Nb3 29.Bf2 Re2 30.Rfe1 Qc2 31.Rxe2 Qxd1+ 32.Re1 Rxe1+ 33.Bxe1 Nc5 34.Qe3 Kf7 35.h4 Qxd5 36.Bf2 Qd3 0-1 Pitterson,J (2221) - Oms Pallise,J (2496) [E29] Turin ol (Men) 37th Turin (10), 21.05.2006 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 c5 6.e3 Nc6 7.Bd3 0-0 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qd3 Na5 17.f5 Qf6 18.Ng3 Rae8 19.Ra2 Nb3 20.Bf4 c4 21.Qf3 Nxf5 22.Raf2 Nxg3 23.Qxg3 Qg6 24.Bd6 Rxf2 25.Qxf2 h6 26.Be7 Kh7 27.d6 Nc5 28.Qd4 Qe4 29.Qf2 Qe6 30.Re1 Qd5 31.Qe2 Nd3 32.Qe4+ Qxe4 33.Rxe4 b5 34.Kf1 Kg6 35.g4 h5 36.gxh5+ Kxh5 37.h4 Kg6 38.Rg4+ Kf7 39.Re4 a6 40.Ke2 Nc5 41.Rf4+ Kg8 42.Ke3 Rc8 43.Rf5 g6 44.Rg5 Kf7 45.h5 gxh5 46.Rxh5 Ke6 47.Kf4 Nd3+ 48.Ke4 Rc5 49.Rh6+ Kf7 50.Rf6+ Kg7 51.Rf8 Rh5 52.Ra8 Nc5+ 53.Kd4 Ne6+ 54.Ke3 Rh3+ 55.Kd2 Nc5 56.Ra7 Kf7 57.Rc7 Ne4+ 58.Ke1 Ke6 59.Ra7 Nxc3 60.Rxa6 Nb1 61.a4 b4 62.Rb6 Rh1+ 63.Kf2 b3 64.a5 Nc3 65.Bf8 Nd5 66.Rb5 b2 0-1 I'll leave it to you to find an improvement over these games. I think this is exactly the kind of position Black is aiming for in the Sämisch and I find it hard to imagine that White should not be worse here.

Thanks for that info.

People interested in this line might want to look at Radjabov-Anand from Melody Amber on Sunday. I'm no expert, but I think Anand simply ignored the received wisdom that Capablanca's 9...Ne8 is important so as to stop the Bg5 pin, and won anyway, although whether he was always doing well I hesitate to say.

The line in question is 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 O-O 6.e3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.O-O Ba6 11.f4 f5  In CBM 122, Moskalenko examines the idea discussed by Markovich and Unknown Master -- playing for d5-d6 and e5, cramping the Black position at the cost of the c4 pawn.  

Moskalenko agrees that 12.d5 is imprecise and argues that White should start with 12.e5, the main idea being 12...Na5!? 13.d5.  If instead 12..d6, as proposed by Unkown Master, then 13.d5?! doesn't work -- see Carlsen-Aronian, Elista 2007 -- but White has the new tries 13.g4!? and 13.Qa4.



Thoughts?
 

Markovich wrote on 03/11/07 at 14:19:48:

unknown_master wrote on 03/10/07 at 23:17:48: It's always difficult to find the correct way if you're surprised at the board with a new move. But if the variation was that strong Milov surely would have tried it again. Birarov,V (2518) - Joao,N (2515) [E29] Veinger mem-A corr ICCF corr, 2002 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 0-0 5.a3 Bxc3+ 6.bxc3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qf3 Ne5 17.Qh3 Qf6 18.g4 Nec4 19.Qd3 Rae8 20.Ng3 Na5 21.g5 Qf7 22.Rb1 c4 23.Qf3 Nf5 24.Nxf5 Qxf5 25.Be3 Re4 26.Rbd1 Rfe8 27.Bd4 d6 28.Qg3 Nb3 29.Bf2 Re2 30.Rfe1 Qc2 31.Rxe2 Qxd1+ 32.Re1 Rxe1+ 33.Bxe1 Nc5 34.Qe3 Kf7 35.h4 Qxd5 36.Bf2 Qd3 0-1 Pitterson,J (2221) - Oms Pallise,J (2496) [E29] Turin ol (Men) 37th Turin (10), 21.05.2006 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 c5 6.e3 Nc6 7.Bd3 0-0 8.Ne2 b6 9.e4 Ne8 10.0-0 Ba6 11.f4 f5 12.d5 exd5 13.cxd5 Bxd3 14.Qxd3 fxe4 15.Qxe4 Nd6 16.Qd3 Na5 17.f5 Qf6 18.Ng3 Rae8 19.Ra2 Nb3 20.Bf4 c4 21.Qf3 Nxf5 22.Raf2 Nxg3 23.Qxg3 Qg6 24.Bd6 Rxf2 25.Qxf2 h6 26.Be7 Kh7 27.d6 Nc5 28.Qd4 Qe4 29.Qf2 Qe6 30.Re1 Qd5 31.Qe2 Nd3 32.Qe4+ Qxe4 33.Rxe4 b5 34.Kf1 Kg6 35.g4 h5 36.gxh5+ Kxh5 37.h4 Kg6 38.Rg4+ Kf7 39.Re4 a6 40.Ke2 Nc5 41.Rf4+ Kg8 42.Ke3 Rc8 43.Rf5 g6 44.Rg5 Kf7 45.h5 gxh5 46.Rxh5 Ke6 47.Kf4 Nd3+ 48.Ke4 Rc5 49.Rh6+ Kf7 50.Rf6+ Kg7 51.Rf8 Rh5 52.Ra8 Nc5+ 53.Kd4 Ne6+ 54.Ke3 Rh3+ 55.Kd2 Nc5 56.Ra7 Kf7 57.Rc7 Ne4+ 58.Ke1 Ke6 59.Ra7 Nxc3 60.Rxa6 Nb1 61.a4 b4 62.Rb6 Rh1+ 63.Kf2 b3 64.a5 Nc3 65.Bf8 Nd5 66.Rb5 b2 0-1 I'll leave it to you to find an improvement over these games. I think this is exactly the kind of position Black is aiming for in the Sämisch and I find it hard to imagine that White should not be worse here. Thanks for that info.

Now that this has come up again, I must agree that "unknown master" was right, and I was wrong, about this particular line.

But I still think that in general, White's prospects in the Saemisch are better than commonly believed.  The ideas pointed to by Lee Roth are interesting, and White can also try 13.Ng3 g6 and then swapping, in some order, on f5 and c5.

unknown_master wrote on 03/09/07 at 12:01:06:

12. d5?! is refuted easily exd5! 13. cxd5 Bxd3 14. Qxd3 fxe4 15. Qxe4 Nd6 when Black is better. So White would have to start with 12. e5 but after d6 13. d5 leads nowhere because of  exd5 14. cxd5 Bxd3 15. Qxd3 dxe5

Dearing recommends 12 ... d6 13. dxe6 Qe7 in "Play The Nomzo-Indian" ISO 12 ... exd5 etc. Wander if Dearing didn't know about 12 ... exd5? Is this development younger than the book?

Oops! I meant 12 ... Na5 13.e5 d6 14. dxe6 Qe7 in the last post!
And it's Nimzo-Indian, not Nomzo-Indian.

CheckMate wrote on 07/03/08 at 13:15:30:

unknown_master wrote on 03/09/07 at 12:01:06: 12. d5?! is refuted easily exd5! 13. cxd5 Bxd3 14. Qxd3 fxe4 15. Qxe4 Nd6 when Black is better. So White would have to start with 12. e5 but after d6 13. d5 leads nowhere because of  exd5 14. cxd5 Bxd3 15. Qxd3 dxe5 Dearing recommends 12 ... d6 13. dxe6 Qe7 in "Play The Nomzo-Indian" ISO 12 ... exd5 etc. Wander if Dearing didn't know about 12 ... exd5? Is this development younger than the book?

I've looked, and I think that White has a little less than nothing after 12...exd5!  So that is what I would play.

Markovich wrote on 07/03/08 at 13:22:02:

CheckMate wrote on 07/03/08 at 13:15:30: unknown_master wrote on 03/09/07 at 12:01:06: 12. d5?! is refuted easily exd5! 13. cxd5 Bxd3 14. Qxd3 fxe4 15. Qxe4 Nd6 when Black is better. So White would have to start with 12. e5 but after d6 13. d5 leads nowhere because of  exd5 14. cxd5 Bxd3 15. Qxd3 dxe5 Dearing recommends 12 ... d6 13. dxe6 Qe7 in "Play The Nomzo-Indian" ISO 12 ... exd5 etc. Wander if Dearing didn't know about 12 ... exd5? Is this development younger than the book? I've looked, and I think that White has a little less than nothing after 12...exd5!  So that is what I would play.

I agree! But Dearing's suggestion also looks very good. White has less than nothing!

Yes, after 12.d5.

What about this line: 1. d4 nf6 2. c4 e6 3. Nc3 Bb4 4. a3 Bxc3+ 5. bc c5
6. e3 Nc6 7. bd3 0-0 8. g4!?

White "Schaffer,Hendrik"
Black "Seifert,Volker"
1.d4 e6 2.c4 Nf6 3.Nc3 Bb4 4.e3 c5 5.Bd3 0-0 6.a3 Bxc3+ 7.bxc3 Nc6
8.g4 d5 9.g5 Ne8 10.Nf3 dxc4 11.Bxc4 e5 12.dxe5 Qe7 13.Qc2 Bg4 14.Be2
Rd8 15.h4 b6 16.c4 Nxe5 17.Nxe5 Qxe5 18.Bb2 Qe6 19.Bd3 Rxd3 20.Qxd3
Bh5 21.Rg1 Qc6 22.Be5 Nc7 23.Bxc7 Qxc7 24.Rg3 Rd8 25.Qe4 Bg6 26.Qf4
Qc6 27.e4 Bxe4 28.Rd1 Re8 29.Re3 Bc2 30.Rd5 g6 31.Kd2 Bf5 32.h5 Rf8
33.h6 Qa4 34.Qe5 Qc2+ 35.Ke1 f6 36.Qe7 Qc1+ 37.Rd1 1-0

CheckMate wrote on 08/28/08 at 20:17:51:

What about this line: 1. d4 nf6 2. c4 e6 3. Nc3 Bb4 4. a3 Bxc3+ 5. bc c5 6. e3 Nc6 7. bd3 0-0 8. g4!? White "Schaffer,Hendrik" Black "Seifert,Volker" 1.d4 e6 2.c4 Nf6 3.Nc3 Bb4 4.e3 c5 5.Bd3 0-0 6.a3 Bxc3+ 7.bxc3 Nc6 8.g4 d5 9.g5 Ne8 10.Nf3 dxc4 11.Bxc4 e5 12.dxe5 Qe7 13.Qc2 Bg4 14.Be2 Rd8 15.h4 b6 16.c4 Nxe5 17.Nxe5 Qxe5 18.Bb2 Qe6 19.Bd3 Rxd3 20.Qxd3 Bh5 21.Rg1 Qc6 22.Be5 Nc7 23.Bxc7 Qxc7 24.Rg3 Rd8 25.Qe4 Bg6 26.Qf4 Qc6 27.e4 Bxe4 28.Rd1 Re8 29.Re3 Bc2 30.Rd5 g6 31.Kd2 Bf5 32.h5 Rf8 33.h6 Qa4 34.Qe5 Qc2+ 35.Ke1 f6 36.Qe7 Qc1+ 37.Rd1 1-0

After 9.g5, Black should play 9..Ne4.

The only other game I know in this line is Hernando Rodrigo-Levin, 2005, where White tried 9.cxd5 exd5 10.f3 Ne8 11.Ne2 instead.  GM Emms now says 11..f5!  If you are a subscriber, you can check out the game, with his annotations, in the archive.  ;)  

CheckMate wrote on 08/28/08 at 20:17:51:

What about this line: 1. d4 nf6 2. c4 e6 3. Nc3 Bb4 4. a3 Bxc3+ 5. bc c5 6. e3 Nc6 7. bd3 0-0 8. g4!? White "Schaffer,Hendrik" Black "Seifert,Volker" 1.d4 e6 2.c4 Nf6 3.Nc3 Bb4 4.e3 c5 5.Bd3 0-0 6.a3 Bxc3+ 7.bxc3 Nc6 8.g4 d5 9.g5 Ne8 10.Nf3 dxc4 11.Bxc4 e5 12.dxe5 Qe7 13.Qc2 Bg4 14.Be2 Rd8 15.h4 b6 16.c4 Nxe5 17.Nxe5 Qxe5 18.Bb2 Qe6 19.Bd3 Rxd3 20.Qxd3 Bh5 21.Rg1 Qc6 22.Be5 Nc7 23.Bxc7 Qxc7 24.Rg3 Rd8 25.Qe4 Bg6 26.Qf4 Qc6 27.e4 Bxe4 28.Rd1 Re8 29.Re3 Bc2 30.Rd5 g6 31.Kd2 Bf5 32.h5 Rf8 33.h6 Qa4 34.Qe5 Qc2+ 35.Ke1 f6 36.Qe7 Qc1+ 37.Rd1 1-0

8...d5 looks good for Black.  In the position after Black's 7...Nc6, invariably I have played 8.Ne2.  8.g4 smells bad to me.

A game with some theoretical relevance:

MNb - Kostanjšek, em WS/H/187 ICCF

1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 O-O 5.Bd3 c5 6.a3 Bxc3+ 7.bxc3 Nc6 8.Ne2 b6 9.e4 Ne8 10.O-O Ba6 11.f4 f5 12.Ng3 g6 13.Be3 cxd4 14.Bxd4 Rc8 15.exf5 gxf5 16.Bf2 Na5 17.c5 Bxd3 18.Qxd3 bxc5 19.Nh5 d5 20.Rfe1 Rc6 21.Rab1 Rf7 22.Re3 Kf8 23.Rbe1 Qc8 24.Qe2 Nf6 25.Nxf6 Rxf6 26.Qh5 Qc7 27.Rh3 Nc4 28.Qh4 Rg6 29.Qxh7 Qxh7 30.Rxh7 Rg7 31.Rh6 Kf7 32.a4 Nd2 33.Rd1 Ne4 34.Rxd5 Rb6 ½-½

Are there any lines in the Saemisch that can be avoided via the 4. f3 move order or does playing 4. f3 simply require the white player to learn more theory?

vasilicus wrote on 03/06/10 at 04:45:39:

Are there any lines in the Saemisch that can be avoided via the 4. f3 move order or does playing 4. f3 simply require the white player to learn more theory?

Either way Black has options not available the other way.  I have a hunch that Black's systems with an early ...b6 are the main reason for the very strong modern preference for 4.f3 (as opposed to the other, anyway).  The main reason for preferring 4.a3 is if you plan to play e2-e3 against Black's main line, which is ...c5, ...Nc6, ...O-O in some order.  Black's stock in that line has risen, however.  If White intends instead and eventual f2-f3, e2-e4 against that, then avoiding the ...b6 ideas is perhaps a good reason for playing f3 on already on the fourth move.

You wind up having to know a more theory with 4.f3, however, since Black has his 5...Be7 and 5...Bd6 ideas after 4...d5 5.a3, and he further has 4...c5, which creates situations quite unlike those that arise form 4.a3.

I remember Yusupov once wrote that the most accurate move order for white is actually 4. e3 0-0 5. Bd3 c5 6. a3

LeeRoth wrote on 03/08/08 at 03:33:48:

The line in question is 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 O-O 6.e3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.O-O Ba6 11.f4 f5  In CBM 122, Moskalenko examines the idea discussed by Markovich and Unknown Master -- playing for d5-d6 and e5, cramping the Black position at the cost of the c4 pawn.   Moskalenko agrees that 12.d5 is imprecise and argues that White should start with 12.e5, the main idea being 12...Na5!? 13.d5.  If instead 12..d6, as proposed by Unkown Master, then 13.d5?! doesn't work -- see Carlsen-Aronian, Elista 2007 -- but White has the new tries 13.g4!? and 13.Qa4. position after 12..d6 [fen]r2qnrk1/p5pp/bpnpp3/2p1Pp2/2PP1P2/P1PB4/4N1PP/R1BQ1RK1[/fen] Thoughts?

I'm coming back to this relatively old thread since I've been looking at the merits of the Saemisch from white's point of view and this thread is very relevant for this.

The position above (Saemisch main line) seems critical since after 12.e5!? Na5 13.d5 Bc4: 14.Bc4: Nc4: 15.d6! white is clearly better since black's knight on e8 is buried alive - a very interesting positional idea stemming from Bronstein. So black should go either 13. ... d6!? after which 14.g4!? is interesting, or 12. ... d6!? straight away as in the diagram. Then white should avoid 13.d5?! ed5: 14.cd5: Bd3: 15.Qd3: as in Carlsen-Aronian, but both suggestions 13.Qa4!? or 13.g4!? are interesting, and also 13.Be3!? seems a good choice to me.

In addition not much seems wrong with 12.ef5:!? ef5: 13.dc5: bc5: 14.Ng3 g6 15.Be3 d6 16.Bf5: ef5: 17.Qd5+ Rf7 18.Qc6: with a very resonable position for white, I'd say.

And if white is still not happy then there is the interesting gambit 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3: 5.bc3: c5 6.e3 Nc6 7.Bd3 0-0 8.e4!? with the point 8. ... cd4: 9.cd4: Nd4: 10.e5 Qa5+ 11.Kf1 Ne8 (11. ... Qe5:?? 12.Bb2) and now white can go 12.Bd2 Qc7 13.Bb4 d6 14.Bh7:+ Kh7: 15.Qd4: a5 16.ed6: Qd7 17.Bc5 with a complex position where white is a pawn up rather than down, or Botvinnik's 12.Bb2 Nc6 13.Nf3 f5 14.Re1 followed by Rg1 and g4 with fair compensation for the pawn. 

All these lines may be viable ways for white to play the main line Saemisch. So what's your verdict on this line, then?      

Well you can find 2600's on the database easily enough, so its not going to be rubbish :) Its also one of those lines 'well known' not to be terribly good, so I'm sure it'd do well up to quite a decent level if you put a bit of work in.

8 e4 I'm not sure about - for one thing not 100% clear what it gains over 8 Ne2/e4 if black doesn't take, and there's (from Dearing's Nimzo repitoire book which is quite thorough here) 8.. cd 9 cd d5 which seems efficient enough.

12 e5 is quite commital of course, but if you like it :)

MartinC wrote on 07/27/11 at 14:48:56:

8 e4 I'm not sure about - for one thing not 100% clear what it gains over 8 Ne2/e4 if black doesn't take, and there's (from Dearing's Nimzo repitoire book which is quite thorough here) 8.. cd 9 cd d5 which seems efficient enough.

I recently smashed a strong opponent in 24 moves who didn't take, where the game went 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3: 5.bc3: c5 6.e3 Nc6 7.Bd3 0-0 8.e4 Ne8?! 9.Nf3 b6 10.e5 (threatening Bh7:+) 10. ... f5 11.d5 Na5 12.Bg5 Qc7 13.d6 Qc6 14.Be7 Rf7 15.Rg1 h6 16.g4 fg4: 17.Nh4 and I won quickly. So the knight on f3 makes a big, big difference.

And my computer came up with the amazing line 8.. cd 9 cd d5 10.e5 Ne4 11.Ne2 Qa5+ 12.Bd2 Nd2: 13.Qd2: Qd2: 14.Kd2: Na5 15.cd5:!? Nb3+ 16.Kc3 Na1: 17.d6 where white will win the Na1 and then has has excellent compensation for the exchange, and is probably better. A very remarkable idea indeed.   

15 cd seems an interesting move (having played it on move 9 :)). Actually Bg4 and not Qa5+ is the main idea for black and the position does look quite stable/reasonable for black.

Obviously not awful for white either though, so if happy with it then go ahead :) You'll probably see black allowing Nf3 or snatching d4 at least as much.

I tend to agree with MartinC.  8.e4 runs into 8..cxd4 9.cxd4 d5! 10.cxd5 exd5 11.e5 Ne4 12.Ne2.  Milov evaluated this as better for White on his old ChessBase CD, but Emms (cited by Dearing) pointed out 12..Bg4 13.Be3 (12.f3 is met by 12..Bxf3 followed by Qh4+) 13.Be3 Qa5 14.Kf1 f6 with what Rybka gives as an edge to Black.  Never Give Up, your 10.e5 Ne4 line is nice but Black can play 10..dxc4. 

Although its a Black repertoire book, Dearing's book from 2005 has very good coverage of the Samisch that is also worth looking at as White.  I haven't really kept up with theory that carefully since then, but I'm not aware of any major changes or improvements for White.  That includes the 12.e5 d6 line discussed above.  At least according to ChessBase, no one's tried 13.Qa4 or 13.g4 yet.  Maybe the biggest disappointment for Samisch fans is that Aleksandrov, who was a big specialist in this line in the early 2000s, seems to have stopped playing it.

After 12.exf5 gxf5 13.dxc5 bxc5 14.Ng3 g6 15.Be3 d6 16.Bxf5 gxf5 17.Qd5 Rf7 18.Qxc6 White hopes for 18..Bxc4, but 18..Bb7! 19.Qa4 h5! is considered to give Black good comp for the pawn.  The Bb7 is better than the Be3, and Black has ideas of h5-h4 and Rf7-g7 hitting g2.  Now, 20.Rab1 Ng7 21.Qd1 of N.Pedersen-Schandorff improves on the old Hollis-Hovde postal game.  In these positions, the engines tend to favor the extra pawn, but, I still think I would prefer Black in these positions, at least in practical play. 

I would look into 14.Be3 d6 playing for the Spassky plan of Bf2-h4, Ng3-f1-e3-d5, etc.  There's a famous game from the 60s, Spassky-Bykhov, where this worked well and some early 2000s examples from Aleksandrov with more mixed results (for these see ChessPub, Dearing's book or Peter Well's annotations in ChessBase.)  Dearing views this line as ok for Black, but it's a game and there's chances for both sides.      

LeeRoth wrote on 07/27/11 at 18:28:22:

After 12.exf5 gxf5 13.dxc5 bxc5 14.Ng3 g6 15.Be3 d6 16.Bxf5 gxf5 17.Qd5 Rf7 18.Qxc6 White hopes for 18..Bxc4, but 18..Bb7! 19.Qa4 h5! is considered to give Black good comp for the pawn.  The Bb7 is better than the Be3, and Black has ideas of h5-h4 and Rf7-g7 hitting g2.  Now, 20.Rab1 Ng7 21.Qd1 of N.Pedersen-Schandorff improves on the old Hollis-Hovde postal game.  In these positions, the engines tend to favor the extra pawn, but, I still think I would prefer Black in these positions, at least in practical play.

I've looked at this again and the line above may well be critical (maye for the Saemisch as a whole?). After 18. ... Bb7 black has some compensation for the pawn, but I don't think it's sufficient. So I think the engines are probably right that white is better here. As Steinitz said, being a pawn up is worth a little discomfort.

Very un-Saemisch this line with black trying to show he has enough for the pawn rather than white!

Not trivial how play should continue after 18. ... Bc4: but I guess 19.Rfd1 Qc8 20.Qc8: Rc8. 21.Rab1 leaves white better placed. The b-file is a major asset.

Speaking of engines, Houdini doesn't see an advantage for white there (at depth 22 it begins sliding into the negative numbers with 21. Rfe1 being the best move at - 0.14, meaning the position is actually trending in black's direction if you were to go by the engine). More importantly, looking over the games that've occurred in correspondence play serves to highlight white lacking any edge, in fact one of the correspondence games demonstrates that black is the one playing for the full point (Souto-Van Tienhoven, E-Mail 2007).

Black doesn't need to try to show compensation, he already has it - hence why this hasn't seen any sort of uptick in popularity for the white cause.

Thanks BPaulsen - you may be right as you've been before!

I confess I haven't taken the position through the computer. I will do that and come back to you.

Would you happen to have the moves of this correpondence game you're referring to? Thanks.

The game is in the www.chesslive.de database, you can find it online.

BPaulsen wrote on 09/14/11 at 08:44:02:

Speaking of engines, Houdini doesn't see an advantage for white there (at depth 22 it begins sliding into the negative numbers with 21. Rfe1 being the best move at - 0.14, meaning the position is actually trending in black's direction if you were to go by the engine). More importantly, looking over the games that've occurred in correspondence play serves to highlight white lacking any edge, in fact one of the correspondence games demonstrates that black is the one playing for the full point (Souto-Van Tienhoven, E-Mail 2007). Black doesn't need to try to show compensation, he already has it - hence why this hasn't seen any sort of uptick in popularity for the white cause.

I looked, and again you are absolutely right. Black has excellent compensation for the pawn.

I'm now going to look into 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 O-O 6.e3 c5 7.Bd3 Nc6 8.Ne2 b6 9.e4 Ne8 10.O-O Ba6 11.f4 f5 12.e5 where I suppose white is doing fine after 12. ... Na5 13.d5 Bc4:(?!) 14.Bc4: Nc4: 15.d6, but critical are 12. ... d6 or 13. ... d6. How do you guys evaluate this variation? Thanks. 

As usual in the critical continuations black is perfectly content.

The likelihood of finding an edge of some sort in any line via 4. a3 is very low. If you like the positions, just be content with "unclear" stuff you might be able to outplay your opponent from.

BPaulsen wrote on 09/15/11 at 10:59:26:

As usual in the critical continuations black is perfectly content. The likelihood of finding an edge of some sort in any line via 4. a3 is very low. If you like the positions, just be content with "unclear" stuff you might be able to outplay your opponent from.

Agree. In the end...you'll become an expert in this variation and you'll outplay all of your opponents. It seems you Never Give Up.  ;)

Thanks a lot guys for your wonderful help with these variations. This has really helped me get a good knowledge and understanding of this lot. There is so much to it.
The Nimzo is a great opening.
I got Dearing's book now (could get it cheap) and that's quite good.
I will do more work understanding the Saemisch by reading Sokolov's Understanding Chess Middlegames - a truly brilliant book.

Objectively black is doing fine in the Saemisch, but I'm pretty sure if I throw some of the variations we have been looking at at strong players I may give them a hard time OTB, because it's not easy and straightforward at all what black should do.

One more line for you guys which I've been looking at - strange looking but very interesting and challenging:
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3:+ 5.bc3: c5 6.e3 Nc6 7.Bd3 0-0 8.Rb1!?
This looks like a beginner's move but is mentioned by Emms in his easy guide. He says "This stange move has some hidden points and is not to be taken lightly."
The point of Rb1 becomes clear in the main line 8. ... b6 9.e4! Now 9. ... cd4: 10.cd4: Nd4:?? loses a piece to 11.e5 Ne8 12.Be4. So black will probably go 9. ... Ne8 after which white can go 10.Nf3 Ba6 11.e5 (threatening Bh7:+) and may be slightly better because the knight on f3 (rather than e2) gives him other possibilities not present in the main line. The knight may go to g5, and to e5 or d2 to support the c4 pawn. I suppose after 8. ... d6!? the best continuation is 9.Ne2 e5 10.e4 steering the game back into mainstream Saemisch. 

Leonid Nikolaevich Yurtaev  - Gennady Tunik  - 
Vladivostok Zaitsev 1995 (1)    
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bxc3+ 5.bxc3 c5 6.e3 Nc6 7.Bd3 O-O 8.Rb1 b6 9.e4 Ne8 10.Nf3 cxd4 11.cxd4 Ba6 12.e5 f5 13.Bd2 Rc8 14.Qa4 Na5 15.Bxa5 bxa5 16.O-O Qc7 17.Rfc1 Bb7 18.Qd1 g6 19.d5 Ng7 20.Bf1 Ba8 21.d6 Qd8 22.c5 f4 23.Ba6 Rb8 24.Rxb8 Qxb8 25.Nd4 Nf5 26.c6 Qb6 27.Nxf5 gxf5 28.cxd7 Qxa6 29.Qh5 Qb6 30.Qg5+ Kh8 31.Qe7 Kg8 32.Qxe6+ Kg7 33.Qe7+ Kg8 34.e6 1-0

Most of these lines have a cheapo feel about them...not that that's neccessarily a bad thing...

Yes Dink, 8.Rb1 looks like a beginners move. This may be an advantage in an OTB game, because black may feel he's better already. Actually it's much more subtle than that.

It's not only Rb1; it's all of these lines, and I say this as someone who has a soft spot for the Saemisch. Now that Black's best paths have been largely worked out, the whole opening feels like a high-grade swindle.

It seems you Never Guve Up!..  :)
If you want to play this "catchy" 8.Rb1!? look at this line:
1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. a3 Bxc3 5. bxc3 c5 6. e3 Nc6 7. Bd3 O-O 8. Rb1 d6! 9. Ne2 b6! 10. O-O Ba6! (The signs "!" here are to emphasize the black's plan..). Another continuation for black can be even  10....e5!? 11. Ng3 Re8 12. d5 e4!?..
Some posts ago (in a similar thread) I suggested you a line for white (easy to study, btw) which could give you some chances for +=. But..you're still trying to dig some diamonds out here where they seem to be completely missing.  :-?

Please do remind me Vass - what's your recommended line against the Nimzo?

I do take your (and BPaulsen's and others) advice very seriously.

I saw you recommended Rubinstein with Ne2 (Bronstein) somewhere - its that your recommended line?
I have played this is the past but got poor results which put me off. The primary reason - it is a very fashionable variation which blacks know very well and they are usually very well prepared. I don't know the present state of affairs but with precise play black is fine I think.

I think I'm going to concentrate now on: (a) trying to become a better tactictian (b) trying to become a better attacking player, (c) getting a better understanding of the nimzo middlegames, rather than hunting for the pot of (opening) gold at the end of the rainbow. Any advice from anyone on the best way of achieving these objectives would be most welcome!

Yes, Rubinstein with Ne2 was my suggestion (with Qd3 as an improvement). Very subtle line! Of course, it is popular because it's one of the very few lines where white can pretend for more than equality.
You already mentioned you were to read carefully Sokolov's "Winning Chess Middlegames".. He tried several setups for white against Nimzo too.. Maybe you can choose one of these...which can help you to find some chances for your beloved attacking chess.  ;)

Thanks for this Vass. What do you consider a good source for this interesting line?

NeverGiveUp wrote on 09/19/11 at 16:09:23:

Yes Dink, 8.Rb1 looks like a beginners move. This may be an advantage in an OTB game, because black may feel he's better already. Actually it's much more subtle than that.

Rb1 is total crap.  Black wants to play ...b6 anyway, for crying out loud. 

Markovich wrote on 09/20/11 at 16:49:16:

NeverGiveUp wrote on 09/19/11 at 16:09:23: Yes Dink, 8.Rb1 looks like a beginners move. This may be an advantage in an OTB game, because black may feel he's better already. Actually it's much more subtle than that. Rb1 is total crap.  Black wants to play ...b6 anyway, for crying out loud.

You may be right.  But, I have wonder about the following very common sequence:

1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 e3 c5 5 Bd3 Nc6 6 Nf3 Bxc3 7 bxc3 d6 8 0-0 e5 9 Nd2 0-0 and now 10 Rb1 b6 is pretty much the main line.

NeverGiveUp's line has lost a tempo with a3, but on the plus side he can happily play Ne2 instead of Nf3 which is supposed to be good vs the Huebner.  But I confess this line isn't in my repertoire for either side, so I don't know all the subtleties...

Vass wrote on 09/20/11 at 08:49:05:

Yes, Rubinstein with Ne2 was my suggestion (with Qd3 as an improvement). Very subtle line! Of course, it is popular because it's one of the very few lines where white can pretend for more than equality. You already mentioned you were to read carefully Sokolov's "Winning Chess Middlegames".. He tried several setups for white against Nimzo too.. Maybe you can choose one of these...which can help you to find some chances for your beloved attacking chess.  ;)

In that Saidy-Unzicker line, what do you do after 7..Nxd5?

Lee Roth wrote:
[/quote]
In that Saidy-Unzicker line, what do you do after 7..Nxd5?
[/quote]

1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. e3 O-O 5. Nge2 d5 6. a3 Be7 7. cxd5..
Good question! This 7...Nxd5!? is a very serious try for equality. I myself played it with black in four or five of my OTB games. And I never played it with white.. Anyway, it seems one of my opponents knowing that I play Nimzo has been prepared a line in which I struggled very much. This game which is not pleasant for me to remember continued: 8.Bd2! Nd7 9.Qc2! (my marks) not trying to expand with e3-e4 or playing g2-g3 and Bf1-g2 right away. In my analysis after the game I've found I made a dubious move 9...N5f6?! which was the main move at that time. There are several moves for black here, of course.. I suppose 9...N7f6!? is the most appealing. But my feeling that something is not quite good for black here stayed for a long time. After that game I never played 7...Nd5 anymore. Not because it's not good, but because it's difficult to handle in an OTB game. All I can say for this line is that after 8.Bd2, 9.Qc2, Rc1 or Rd1 (if c7-c5) and e3-e4 in the right time white is having at least a minimal edge. One thing I'm sure of is that 9....N5f6 is not good.  ;)
Edit: I remember Angus Dunnington in his excellent "The Nimzo-Indian Rubinstein - The Ever-Popular Main Lines With 4.e3" (2003) explained very well all the problems that white and black can face after 7....Nxd5. I think he recomended this 8.Bd2, too..

NeverGiveUp wrote on 09/20/11 at 10:10:07:

Thanks for this Vass. What do you consider a good source for this interesting line?

The best books for this variation seem to be:
Angus Dunnington "The Nimzo-Indian Rubinstein, The Ever-Popular Main Lines With 4.e3" (2003) and
Carsten Hansen "The Nimzo-Indian, 4.e3" (2002)
But my advice is: Don't rely on books only!.. A good database and a silicon brain will be of invaluable help. My experience is that relying on statistics is wrong, too.. And sometimes even players with ratings below 2500 can hit the best moves (by accident).  ;)
Aaah...and one more thing - even when the silicon tells you 0.00 don't trust it completely and prolong the line if you have a feeling that something has to be there!  In fact, the most exciting "novelties" come that way. 8-)
Edit: I forgot to recommend Reinaldo Vera's "The Nimzo-Indian" (2008). And all because of my age.. Eeeh!

According to Amazon, coming in March 2012:

The Strategic Nimzo-Indian:  A Complete Guide to the Rubinstein and Saemisch Variation by Ivan Sokolov

You just need to be patient.   :D

I can hardly wait!  :D

Ivan Sokolov should write more books because he is a great writer and a great player.

NeverGiveUp wrote on 09/19/11 at 09:49:41:

One more line for you guys which I've been looking at - strange looking but very interesting and challenging: 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3:+ 5.bc3: c5 6.e3 Nc6 7.Bd3 0-0 8.Rb1!? This looks like a beginner's move but is mentioned by Emms in his easy guide. He says "This stange move has some hidden points and is not to be taken lightly." The point of Rb1 becomes clear in the main line 8. ... b6 9.e4! Now 9. ... cd4: 10.cd4: Nd4:?? loses a piece to 11.e5 Ne8 12.Be4. So black will probably go 9. ... Ne8 after which white can go 10.Nf3 Ba6 11.e5 (threatening Bh7:+) and may be slightly better because the knight on f3 (rather than e2) gives him other possibilities not present in the main line. The knight may go to g5, and to e5 or d2 to support the c4 pawn. I suppose after 8. ... d6!? the best continuation is 9.Ne2 e5 10.e4 steering the game back into mainstream Saemisch.

Further on this guys - a club mate of mine has now suggested the clever 8.Ra2!? instead of Rb1. At first sight this looks like an even more nonsense move than Rb1 - but consider that Ra2 is a perfectly normal move in the Saemisch: to mobilse the Ra1 and to be able to transfer it to the center or kings side. Again after the natural 9. ...b6 white continues 10.e4! Ne8 (cd4: 11.cd4: Nd4:?? 12.e5 Ne8 13.Be4 +-) 11.Nf3! and white is well placed with the knight on f3. After, say 11. ... d6 play may continue either with 12.e5!? or 12.Be3 e5 13.d5, when the natural Na5 can be met with Nd2.

Hi, NeverGiveUp! I'm glad you're back after a long absense.  :)
And it seems you Never Give Up..  ;D
As for 8.Ra2 (after 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3+ 5.bxc3 c5 6.e3 Nc6 7.Bd3 0-0) here..I won't play "the natural" 8....b6. The somehow unnatural 8.Ra2 will be met by the unnatural 8....e5!? with the idea of e5-e4 (and even d7-d5) or b7-b6 if d4xc5..  ;)

Spot on again Vass - 8. ... e5! is a very good move. In several variations the rook an a2 is seriously exposed after this. So my conclusion is that this variation is not very good for white.

I have played Saemisch with f3 for a long time. People know this now and prepare against me: that's why I've been looking at other variations. In general I've done good shop with the f3 line since quite a lot of blacks go 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.a3 Bc3:+ 5.bc3: c5 6.f3 d5 7.cd5: Nd5: 8.dc5: (Keres) which is a variation where theoretically black should be OK but practically white is doing fine. Some blacks go 7. ... ed5: which is the Botvinnik variation and clearly better for white. The critical lines as far as I'm concerned are:
(a) 6. ... Nc6 7.e4 d6 8.Be3
(b) 5. ... b6 6.e4 Nc6 followed by Ba6 (very popular)
I do wonder about these lines though - are things really bleak for white? What's the current theoretical verdict? 

as for (b), I don't know if a theoretical verdict can be found, but it's popular for good reason, since White gives a pawn and (more importantly) Black's  ...Bxc4 attacks the bishop and sometimes allows for Bxh7+ Kxh7 Ng5+ Kg8 Qh5 Bd3 (a funny theme). Still it's pawn vs attack, perhaps easier to play for Black between same-strength players, but not easy to draw a conclusion.
If you want to play with f3 and a3, isn't it better to start with f3 ? you typically wait for Black to castle or play d5, and then you can play a3, now Bg5 is a threat and Black often has to play Nh5 or Ne8, in fact Black's ...0-0 is good for your attack. Just practical tips, certainly Vass or other people will have more accurate views.

wrote on 01/17/12 at 22:29:28:

..... and sometimes allows for Bxh7+ Kxh7 Ng5+ Kg8 Qh5 Bd3 (a funny theme).

What line exactly are you thinking of? If there is a pawn on f3 then Qh5 is impossible. If there is a pawn on f4 then ...Bd3 is answered with f4-f5. If the pawn is on f2 then it's not the 4.f3 or 6.f3 variation.

I had this theme as a variation in a game, indeed White must have played f3-f4-e4-e5 and Black b6-Ba6-Nc6-Na5.
The quickest route may be 1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.f3 0-0 5.a3 Bxc3 6.bxc3 Ne8 7.e4 b6 8.Bd3 Ba6 9.f4 Nc6 10.Nf3 Na5 11.e5 (see diagram). 
[FEN]r2qnrk1/p1pp1ppp/bp2p3/n3P3/2PP1P2/P1PB1N2/6PP/R1BQK2R b KQ - 0 11[/FEN]
Now 11...Bxc4 12.Bxh7+ Kxh7 13.Ng5+ Kg8 14.Qh5 Bd3 is the theme. Here it should work, as Black seems to defend after 15.f5 Bxf5 16.0-0 Bg6 17.Qh4 f6 18.Rf3 fxg5 19.Rxf8+ Kxf8 20.Bxg5 Qc8 21.Rf1+ Kg8 22.Be7 Bf5 23.g4 d6 (by following chessok.com online opening tree, Black lost horribly in Brasket-Bisguier, Puerto Rico 1968, by playing 20...Qxg5, perhaps somebody can post the game).
All in all, a risky but useful theme to remember, methinks.

Ah, thanks, I reached that position with the moves ...c5 and O-O already played. That makes a huge difference. In that case Black simply can avoid the combination by answering a timely ...h6.
In your sequence 12.O-O might be better, claiming compensation in the form of central dominance after Bxd3 13.Qxd3 h6 14.a4.

@ NeverGiveUp
Having mentioned Victor Moskalenko's "Revolutionize Your Chess - A Brand-New System To Become A Better Player" (New In Chess, 2009) in another thread, it came to my mind that this book could be of your help about studying your favourite Nimzo Indian's Saemish/Kmoch variations.
More than 50 pages of good analyses sharing many ideas for white from his own experience.  ;)