Last but not least, we have 1.d4 Nf6 2.c4 Nc6 3.Nf3 e6 4.Nc3 Bb4 5.Qc2 (5.e3 is possible, but 5...d6 followed by a timely …e6-e5 gives Black a solid game).  At the master level, this is probably the most common response to the Tango. The Qc2 Nimzo has been popular for many years and is in the repertoire of many players. Because this position arises from a mainstream opening so often, there are many examples to examine.

E Bareev -J Timman [E33]
Wijk aan Zee 2004
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.Qc2 Nc6 5.Nf3 d6 6.Bd2

For a while 6.Bg5 h6 7.Bd2 (not 7.Bh4? g5 8.Bg3 g4 winning a pawn…more on this below) was favored after the game M. Gurevich - Benjamin New York 1989: 7...e5 8.a3! Bxc3 9.Bxc3 Qe7 10.d5 Nb8 11.e4 0–0 12.Be2 Nh7 13.g3! Bh3 14.Rg1 Bd7 15.Nh4 c6 16.f4! Na6 17.dxc6!  bxc6 18.0–0–0 and White developed an advantage.  But if the center remains fluid for a bit, White is unlikely to castle queenside. Then …h7-h6 can be a useful move, rather than a target for attack, e.g.   7...0–0 8.a3 Bxc3 9.Bxc3 Qe7 10.g3 e5 11.d5 e4 and the knight cannot go to g5 (M. Gurevich - E. Torre Baltic ch., 1996).

White has a completely different approach in 6.a3. After 6…Bxc3+ 7.Qxc3, I like 7...a5, in order to deter White's thematic b2-b4 advance. My only concern is 8.d5, but Black came out okay in Ward - Quinn Monarch Assurance Open, 2000: 8...exd5 9.cxd5 Ne7 10.e4 Nxe4 11.Qxg7 Rg8 12.Qd4 Nc5 13.Bg5 Nf5 14.Bxd8 Nxd4 15.Nxd4 Kxd8 and Black went on to win.

Now it is confession time for me. My recollection of trapping Yasser Seirawan (see Tango II) is not confirmed in the database. Peerless leader Silman interjected that he had actually caught the venerable Yaz the same way. Seirawan replied that we were both crazy. Well, it turns out I wasn't crazy about beating Yasser in this variation, but I was off on the details. Readers may take for granted that Silman is getting old, if not senile.  Harder to take is the thought that Yasser is getting old. But I'm afraid, dear readers, that I am getting old, too! Forty years old in March, my memory is now just a jumble. For the record, this is the real game, Seirawan - Benjamin Philadelphia, 1989: 8.Bg5 h6 9.Bh4 Qe7 10.e3 e5 11.dxe5 dxe5 12.Be2 Bf5 13.0–0? g5 14.Bg3 Ne4 15.Qc1 h5 16.h3 h4 17.Bh2 g4 18.hxg4 Bxg4 19.Kh1 h3 20.gxh3 Rxh3 21.Ng1 Rxh2+ 22.Kxh2 Qh4+ 23.Kg2 0–0–0

There's no quit in Seirawan, but he only plays this out because it's a blitz tournament.

24.Qd1 Rxd1 25.Raxd1 Bxe2 26.Nxe2 Qg4+ 27.Ng3 Nxg3 28.fxg3 Qe2+ 29.Kh3 f5 30.Rde1 Qg4+ 31.Kh2 Qxc4 32.Rf2 Qe4 33.Rd2 b6 34.Rde2 Ne7 35.Rf2 Nd5 36.Rfe2 Nf6 37.Kh3 Ng4 38.Kh4 Nh2+ 39.Kh5 Qg4+ 40.Kh6 Nf3 41.Rd1 Ng5 42.Red2 Ne6 43.Rd7 Qg5+ 44.Kh7 Nf8+, 0–1.

Finally, if White answers 7...a5 with a modest move like 8.e3 or 8. b3, 8...0-0, 9...Re8 (or 9...Qe7), and 10...e5 should do Black nicely.

6...0–0

At first I tried 6...e5, but I find it flawed now, not so much because of 7.d5 Bxc3 (7...Ne7?? 8.Qa4+ – don't fall for that one!) 8.Bxc3 Ne7, but rather 7.a3! Bxc3 (7...exd4 8.axb4 dxc3 9.Bxc3 is worse) 8.Bxc3. Now 8...Qe7 is necessary to defend the e5-pawn, but then 9.d5 forces the knight back to b8.

7.a3 Bxc3 8.Bxc3 Re8

I've never played this move, but it seems to be most in favor now. 

9.Rd1

White delays a decision on his kingside development while forcing Black to place his queen on e7. Lautier - Piket Groningen, 1995 continued 9.e3 e5 10.d5 Ne7 11.Bd3 c6 12.dxc6 bxc613.0–0 Ng6 14.h3 Qe7 15.Bxg6 hxg6 16.c5 Bf5 17.cxd6 with a slight edge for White, but perhaps 12...Nxc6 is better. After 13.Ng5 (13.0–0 e4 14.Bxf6 exd3 15.Bxd8 dxc2 16.Bc7 Bf5 17.Bxd6 Na5 18.Nd2 Rad8 19.c5 b6 is good for Black) 13…h6 14.Ne4 Be6 15.Nxf6+ Qxf6 looks okay for Black, despite the bishop pair.

9...Qe7 10.b4!?

Timman had seen Black's possibilities from the other side in an earlier game.  Timman - Morozevich Amsterdam, 1995 continued 10.g3 (?! Timman) 10...e5 11.d5?! e4! 12.Ng5 and now 12...Ne5 13.Bg2 Kf8! gives Black the edge, according to Timman (14.Nxe4? Bf5).

10...e5 11.d5 Nb8 12.e4 Bg4

Black eases his cramp by ensuring all his pieces get out.

13.Be2 Nbd7 14.Nh4 Bxe2 15.Nf5 Qf8 16.Kxe2 Nh5 17.Bd2 Nf4+ 18.Kf1 g6 19.Ne3 c5 20.g3 Nh5 21.Kg2 Qe7 22.Rb1 Rf8 23.f3 Kh8 24.Qd3 Ng7 25.Nd1 f5

The game has turned into a King's Indian with an unusual distribution of pieces. White might be a shade better, but Timman handled the opening of the position better than his esteemed opponent.

26.Nf2 Rf7 27.exf5 gxf5 28.Rhe1 Rg8 29.f4 Nh5 30.Qf3 Qh4 31.fxe5 Rfg7 32.Rb3 Nxe5 33.Rxe5 dxe5 34.Qxf5 Qxc4 35.Rf3 Qxd5 36.Ne4 Rg4 37.Qxh5 Rxe4 38.Bh6 Re2+ 39.Kh3 Qe6+ 40.Rf5 Rf2, 0–1.

The last word on the Tango will be a return to a variation from Tango I. After 1.d4 Nf6 2.c4 Nc6 3.Nc3 e5 4.d5 Ne7 5.e4 Ng6, Silman wants to know about 6.h4 – what do I have to say about that?

I may well have underestimated this move. First of all, Black needs to play 6...h5. 

6...Bb4?! 7.h5 Ne7 (Prokupchuk – Bologan, Krasnodar 1997) and now 8.h6! g6 9.Qf3 Nfg8 10.Bg5 with a clear edge for White, according to Bologan. No argument here.

The best player to have the position after 6…h5 on record, Vadim Milov, played 7.g3 here, but didn't impress: 7...Bc5 8.Bd3 a6 9.Qe2 d6 10.Bg5 c6 11.a3 Ne7 12.Rc1 cxd5 13.cxd5 Bd7 14.Kf1 Ng4 15.Nh3 Qb6 16.Nd1 f6 17.Bd2 f5 18.Kg2 0-0 19.Bg5 fxe4 20.Bxe4 Nf5 21.b4 Bd4 22.f3 Nf6 23.Bxf6 Rxf6 24.Ng5 a5 25.Rb1 axb4, 1/2-1/2, Milov -Ramesh Biel, 2001.

More troublesome is 7.Bg5. 

If Black develops his king bishop outside the pawn chain, he cannot break the pin by normal means. There is little to go on in the database. 

7...Bc5 looks logical enough and was recommended by Orlov. White tried a reasonable plan in Tomescu - Iordachescu Bucharest, 1993:

8.g3 d6 9.Bh3 Ne7 10.Bxf6 gxf6 11.Qd3

Now 11…c6 led to a hard fought victory for White: 12.Bxc8 Qxc8 13.Nge2 f5 14.f3 Bb4 15.0–0–0 fxe4 16.fxe4 cxd5 17.cxd5 a6 18.Rdf1 Qg4 19.Rf6 Ng8 20.Rf5 f6 21.Kb1 Qg6 22.Rhf1 Ba5 23.Qc4 Kd8 24.Qa4 Bc7 25.R5f3 Rh7 26.Qb4 Rb8 27.Nc1 Rf7 28.Nd3 Nh6 29.Rc1 b5 30.Nf2 Rd7 31.a4 a5 32.Qb3 b4 33.Nb5 Bb6 34.Qd3 Bc5 35.Ka2 Ng4 36.Nxg4 hxg4 37.Rf5 Ke7 38.h5 Qh6 39.Rcf1 Rf8 40.Qe2 Qg7 41.Rh1 Kd8 42.Rh4 Rg8 43.h6 Qg6 44.h7 Rxh7 45.Rxh7 Qxh7 46.Rxf6 Rg6 47.Rf8+ Ke7 48.Qf1 Qg7 49.Rc8 Qf6 50.Qc4 Qf2 51.Nxd6 Bxd6 52.Qc6 b3+ 53.Kxb3 Qxg3+ 54.Ka2 Kf7 55.Qd7+ Be7 56.Qe8+ Kg7 57.Qxe7+, 1–0.

Fair enough, but Black looked solid there – if he hadn't forgotten about the castling rule, he would have come out alright.  Let's take another look at the position after 11.Qd3

The simplest continuation for Black is 11...Bxh3 followed by 12...Qd7 and 13...0-0-0. I think this general plan (Bc5, d6, Ne7 and eventual queenside castling) works for Black against most White approaches from move eight. So, it looks like Black is okay, but I would like to see more strong players test this 6.h4 variation.

Thank you all for following this long, if not completely exhaustive investigation of a half fringe, half mainstream defense, the Black Knights Tango. Next up: I'll offer some anti-Sicilian ideas for White.