Blackmar-Diemer Gambit O'Kelly & Ziegler Defence

Hi all- here is the first entry of my chess blog.  I have a chess openings site at http://tws27.50webs.com/chess/introduction.html which covers numerous gambit systems, and for the benefit of Chess.com readers I will be offering some insights into these in my Chess.com blog.

The Blackmar-Diemer Gambit (1.d4 d5 2.e4 dxe4 3.Nc3 Nf6 4.f3) is a controversial gambit line in which White sacrifices a pawn in return for open lines and development, particularly utilising the half-open e and f-files.  Objectively White may well fall marginally short of full compensation for the pawn with best play, but the gambit provides many club-level players with a lot of fun.  I got into this opening partly because it is a mirror-image of the Göring Gambit (1.e4 e5 2.d4 exd4 3.Nf3 Nc6 4.c3) which is also a lot of fun.

White generally has two main approaches in the accepted lines of the gambit.  White can castle short, bringing a rook to the f-file immediately, and putting pressure on f7, and often plays Qe1-h4 to add to the attack.  Alternatively White can castle queenside, which helps to support a d4-d5 push and allows White to take more liberties in advancing the g and h-pawns, but White often takes longer to get a rook on the f-file then.

To begin with I will examine the most critical response from Black, the Ziegler Defence (4...exf3 5.Nxf3 c6) and the related O'Kelly Defence (4...c6).

Ziegler Defence

1.e4 d5 2.d4 dxe4 3.Nc3 Nf6 4.f3 exf3 5.Nxf3 c6

This is currently considered the most critical defence to the Blackmar-Diemer.  Black aims to set up a solid Caro-Kann formation and waits for White to commit the f1-bishop before developing the c8-bishop. 

6.Bc4

Also playable is 6.Bd3.  Black normally meets this with 6...Bg4, placing the f3-knight in an irritating pin.  White may then continue with 7.Be3 defending d4, although 7.h3!? Bxf3 8.Qxf3 Qxd4 9.Be3 is an interesting and speculative sacrifice of a second pawn.  Following 7.Be3, a possible continuation runs 7...e6 8.Qe2 Nbd7 9.h3 Bh5 10.g4 Bg6.

White has a fair amount of development and a space advantage as compensation for the sacrificed pawn, though objectively, it isn't clear if it is enough.

After that digression, we return to the most important continuation, 6.Bc4, eyeing f7.

6...Bf5

I have had the pleasure of playing the white side of 6...Bg4? 7.Ne5.

Note that 7.Bxf7+ Kxf7 8.Nxe5+, regaining the pawn and keeping an attack going, is also good.  After the more spectacular 7.Ne5, Black cannot take the queen because of 8.Bxf7 mate.  If Black plays 7...Bh5?, then 8.Qxh5! follows, and Black still cannot take the queen, but even the relatively best 7...Be6 8.Bxe6 fxe6 is poor for Black due to the doubled and isolated e-pawns.

So Black must play 6...Bf5.

7.Bg5

Less effective are the standard attacks with 7.0-0 e6 8.Ne5 Bg6!, or 8.Ng5 Bg6!.  Stefan Bücker and others looked into ways of making the 8.Ng5 line work at the Chesspublishing.com forum, but it seems that Black is better after 8...Bg6 9.Ne2 (intending Nf4) 9...Bd6 10.Nf4 Bxf4 11.Bxf4 0-0.

7...Nbd7

Lev Gutman and Stefan Bücker have demonstrated that after 7...e6 (intending 8.Qe2 Bb4, preventing White from safely castling to the queenside) White gets a reasonable game with the well-timed attack on the f5-bishop with 8.Nh4! Bg6 9.Nxg6 hxg6 10.Qd3.  White will tie Black to the defence of the f-pawn and threatens sacrifices on e6 and g6.

8.Qe2

Here 8.Nh4 is less effective because of 8...Bg4.

8...e6 9.0-0-0 Be7

With 9.0-0-0 White reinforces the d4-pawn and threatens to push d4-d5, taking advantage of the pin on the f6-knight.  Thus, 9...Bb4 is inaccurate here because of 10.d5!.

10.Rhf1 The rook goes to its usual spot on the f-file.

10...Nd5 11.Bd2 White doesn't want to let Black shatter the queenside pawn structure with ...Nxc3, or exchange off a lot of pieces, after which the extra pawn would become more of an important factor.  Thus this retreat is probably necessary.

11...Qc7

As is so often the case in the Blackmar-Diemer, White has a fair amount of compensation for the pawn due to the open lines and development, but is it enough?  It's hard to say.  However, 12.h3 followed by g4 harassing the f5-bishop is a good plan from here.  If Black keeps the king in the centre or goes long, then White should put pressure on Black using the e and f-files, while if Black castles short (probably objectively best, but also risky), the g and h-pawns will come racing forward at Black's king.

O'Kelly Defence

1.e4 d5 2.d4 dxe4 3.Nc3 Nf6 4.f3 c6!?

An important alternative way for Black to try to get into a Ziegler Defence is with the O'Kelly Defence, 4...c6.  After this move, 5.fxe4 e5! is annoying for White (6.Nf3 exd4 7.Nxd4 Bb4, or 7.Qxd4 Qxd4 8.Nxd4 leaves White with an isolated pawn in the endgame). 

5.Nxe4 (which transposes to a line of the Caro-Kann Defence, 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nf6 5.f3?!) doesn't give White anything as long as Black avoids the tempting line, 5...Nxe4 6.fxe4 e5 7.Nf3 exd4?! 8.Bc4!.  Instead simply 5...Nbd7 followed by simple development is at least equal for Black.

Therefore White's best response is probably 5.Bc4.

Here Black can get into normal Ziegler Defence channels by playing 5...exf3 6.Nxf3, transposing to the line 4...exf3 5.Nxf3 c6 6.Bc4, while avoiding the deviation 6.Bd3.  However, Evgeny Bareev came up with a different idea in his 2000 game against Nigel Short.

5...b5!?

Another important idea for Black is to defend e4 with 5...Bf5.  White should attack the bishop with 6.g4 Bg6, and then displace the f6-knight with 7.g5 Nd5 8.fxe4 Nxc3 9.bxc3, and White has a fully acceptable position.  Black can grab the pawn on e4, but White has plenty of open lines and development to compensate.

6.Bb3 e6

6...exf3 is also an important option, with the idea 7.Nxf3 b4.  In that line I like 8.Ng5!?, since Black cannot take on c3 due to the threats to f7, and White has various sacrifices on e6 and f7, e.g. 8...e6 9.Nce4 Be7 (9...Nxe4 10.Nxe4 Qh4+ 11.Nf2 is safer) 10.Nxf7!? Kxf7 11.Ng5+ Kg8 12.Nxe6.

7.fxe4 b4 8.Na4

White prepares Qf3 on the next move.  This is an improvement over the 8.Ne2?! of Short-Bareev, Sarajevo 2000, which left White with scant compensation for a pawn after 8...Nxe4 9.Nf3 Ba6 10.0-0 Bd6 11.c4 bxc3 12.bxc3 Nd7.

8...Nxe4 9.Qf3 Nf6 10.Ne2 Ba6 11.Bg5

And once again, the eternal question... does White have enough for the pawn?  I can't see many attack-minded club-level players losing much sleep over that question though- White's practical chances in this sort of position are pretty good, unless you are playing at a very high level.