Chessville - by chessplayers, for chessplayers!

Reviving the Colle-Zukertort Main Line
by David I. Rudel, author of
Zuke 'Em - The Colle-Zukertort Revolutionized

Too Many Plates…

As described in an earlier article, the Colle-Zukertort opening excels at suppressing counter-play, or at least delaying it.  Indeed, the system owes much of its appeal to Black’s inability to target much of anything, let alone the White King, in the first twenty moves or so.

However, White’s vigilant constriction can take on the feel of spinning plates should Black choose a solid-yet-active course.  White can find himself trying to squeeze an egg to half its size without winding up with yoke on his nose.

The main line illustrates, understandably, the most important instance of this:

1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 c5 5.b3 Nc6 6.O-O Bd6 7.Bb2 O-O [diagram]

Black has three major tactical ideas in this position:

  • He can threaten …e5 by first playing either …Qc7 or …Qe7.  White can play 8.Ne5 to totally stop this option.

  • He can plant a Knight on e4.  White could play 8.Nbd2 if he wants to put this idea on ice.

  • He can play …Nb4, forking the c2-pawn and White’s prize Bishop on d3.  This move is only a mild annoyance unless he also plays his Queen to c7, at which point the c-pawn can come under double attack. (E.g., 8.Ne5 Qc7  9.f4 cxd4  10.exd4 Nb4)  White can completely ice this idea with 8.a3

As hard it might be to believe, for decades Colle players managed to stop all three of these ideas using the continuation 8.Nbd2 Qe7 9.Ne5 cxd4 10.exd4 Ba3.

Let’s examine how this magic is possible.
 

The Queen and the Lever

The saving grace that allowed White to keep all the plates spinning for many years is that Black’s threat of …Nb4 is not particularly troublesome unless his Queen is on c7.

Unfortunately (for Black), the Queen on c7 cannot “see” the c5 square due to the c6-Knight.  This means that if White exchanges on c5, the Bishop could be in danger.

For example 8.Nbd2 Qc7?! 9.dxc5! Bxc5 10.c4 e5? 11.cxd5! Nxd5, and White can win a pawn with 12.Bxh7+ Kxh7 13.Qc2+ [diagram]









From c7 the queen cannot
protect the bishop on c5.

Black is not advised to get around this problem by resolving the lever himself with an early …cxd4.  Doing so will just lead to a bad IQP position (e.g., 8.Nbd2 cxd4  9.exd4 Qc7  10.a3 e5?!  11.dxe5 Nxe5  12. Nxe5 Bxe5  13.Bxe5 Qxe5  14. Re1).
 

Luring the Knight

You might be thinking: Well gee!  If 8.Nbd2 Qe7 9.Ne5 cxd4 10.exd4 solves all of White’s problems, what’s the problem?

Prie.  Prie is the problem.  The resourceful French GM could not leave us C-Z players in peace.

Prie realized the black queen does not have to move immediately to c7, she just has to end up there eventually.  After 8.Nbd2 Qe7, White must play 9.Ne5 to stop the threatened e-pawn advance.  Prie realized that he could attack the knight that has been lured to e5 with 9…Qc7!!  White has to protect or exchange it, which means he cannot address the threat of …Nb4.

Trading off the Knight with 10.Nxc6 does not leave White with any advantage to speak of, but after 10.f4 cxd4 11.exd4 Nb4 [diagram] White will lose his prized d3-Bishop.









Other Options

After 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 c5 5.b3 Nc6 6.O-O Bd6 7.Bb2 O-O [diagram] White has tried a3, c4, or Ne5 99% of the time.








 8.a3 was Max Euwe’s choice, and he used it to beat Rubinstein with his own opening in 1923.  However, this does nothing to address the threats of …e5 and …Ne4. After 8…Qe7  9.Ne5 cxd4  10.exd4 Ne4, White has nothing.

[Note: I am inclined to think that White can do better in this line (more on that later) and Black’s most accurate play is 8…b6  9.Nbd2 Nd7!?, retreating the knight when White has no K-side attack to speak of.  White retains some advantage here, but is unlikely to get the game he wants since he cannot maintain his knight on e5.]

8. c4 side-steps all these tactics and brings us to an odd version of the Queen’s Indian Defense.

In this variation of the Queen’s Indian, White normally plays c4 and d4 very early, and Black eventually plays …c5 and …d5. At that point White immediately exchanges a central pawn (in pretty much all lines). Unfortunately, White cannot do that here because he was so late playing his c-pawn forward. Instead, Black will be the one initiating the first pawn exchange. Black is fine after both 8…cxd4 9.exd4 b6 or 8…dxc4 9.bxc4 b6.

I suggested 8. Ne5!? in Zuke ‘Em because the move shuts down any chance of …e5, is novel enough to catch many off guard, and even stops …Ne4 because the d1-h5 diagonal is now open.  (E.g., 8…Ne4?  9.Bxe4! dxe4 10.Nxc6 bxc6 11.dxc5 Bxc5 12.Qg4!)

In this 8. Ne5!? line, I found an attack that makes able use of the 3 moves Black must spend to bag the d3-Bishop (8. Ne5!? Qc7  9.f4 cxd4  10.exd4 Nb4  11.Nc3! Nxd3  12.Qxd3 [diagram])

Black has to defend against both Nb5 and the Rf3-h3+ Ng4 idea.  I could only find three instances of White using this plan, and Black got eaten alive each time.

While this line has significant practical merit, IM Silman and Colle-Zukertort practitioner John Dowling have pointed out that accurate play appears to allows Black to survive the attack and obtain a comfortable game.









White’s Attack is
Easy to Underestimate

So, what do we do? Do we use a less ambitious line and accede to dull equality when Black brings his A game? Do we just throw this opening away and go find another one that allows Black to equalize if he knows what he is doing? Do we give up on chess and take up knitting?

No. First we turn the board around.
 

The Solution

Given the problems White has preserving an advantage in this line, I would honestly be happy with one of those reversed-opening solutions where White is a nebulous “move up” on some respectful Black defense.  But that’s not the story here…not by a long shot.

Let’s take a look at the diagram on page one with colors, except we’ll pretend White (here Black) didn’t castle.

(We might arrive at this board after 1.d4 d5  2.c4 e6  3.Nf3 Nc6  4.e3 Nc3  5.Bd3 b6  6.Bd3 Bb7  7.O-O [diagram].)








Now, what opening does this look like?

The answer is:  Nothing.  It doesn’t look like anything…except perhaps an odd concoction of 2 parts Tartakower Defense to 1 part vodka.

Okay, so that didn’t help.  But what if we play 7….dxc4! 8.Bxc4 a6 [diagram]








Still not sure what I’m getting at?

Okay, let’s throw in 9.Bd3 b5 [diagram]








And now, perhaps, it has become more clear…the position looks a lot like a Meran variation of the Semi-Slav where the c-pawn has not moved [or a reversed QGA where Black hasn’t played …c5.]

Compare the position after an extremely important main line of the Semi-Slav (1.d4 d5  2.c4 c6  3.Nc3 Nf6  4.Nf3 e6  5.e3 Nbd2  6.Bd3 dxc4  7.Bxc4 b5  8.Bd3 Bb7  9.0-0 a6 [diagram, left]) to the position after we start with the first several moves of the Colle-Zukertort (1. d4 Nf6  2.Nf3 d5  3.e3 e6  4. Bd3 c5  5. b3 Nc6  6.O-O Bd6  7.Bb2 O-O) and then play 8.dxc5! Bxc5 9.a3 Bd6 10.b4 [diagram, right]






The differences are:

  • White has not blocked his Bishop with c3, as Black does in the Meran with c6.

  • White has developed his Bishop whereas Black has developed his Knight.

  • White has used his extra move to castle.

The more you know about the Meran system, the more important and exciting the differences become.

Most obviously, the open diagonal complicates Black’s life because the cardinal rule of the Meran system is that White should always, always, always advance the e-pawn immediately after Black plays …a6 (corresponding to the position in the diagram above).  However, with the Q-bishop eyeing the center, the e-pawn push has to be reconsidered.  [Note that White can play b5 to undermine the knight’s defense of e5.]

It is not merely the pressure on e5 that matters here.  The availability of Bxf6! if Black pushes his e-pawn to e4 can be critical (e.g., 1.d4 d5 2.Nf3 Nf6 3.e3 e6 4.Bd3 c5 5.b3 Nc6 6.O-O Be7 7.Bb2 O-O 8.dxc5 Bxc5 9.a3 Qe7 10.b4 Bd6?! 11.Nbd2 e5? 12.c4 e4 13.Bxf6! when …Qxf6 drops a pawn and 13…gxf6 is even worse, weakening the pawn structure, blocking the queen’s access to the K-side, and still not allowing Black to keep material equality: 13…gxf6 14.cxd5 exf3 (14…exd3 15.dxc6 bxc6 16.Nd4) 15.dxc6 fxg2 16.Kxg2 and Black cannot pause to recapture due to the threat of Qh5.)

Being able to develop the bishop instead of the Q-knight is useful in that White has the flexibility of playing Nc3-b5 or Nc3-a4 should it be needed.  In the normal Meran Black cannot bring out the bishop early as it incites an advantageous e-pawn advance.

While the above two points are probably more or less obvious, what is not obvious is the bone-crackling importance of castling.  Indeed, Black typically must leave his king stuck in the middle for quite a while, for the opening tends to demand more active play.  He just does not have the time to develop his bishop and castle.  This difference is truly profound.

The issue here is that b5 move I mentioned earlier, undermining the knight.  In several branches of the Meran, there is a standard theme of the b-pawn threatening to fly up the Q-side while the opponent’s e-pawn rips up the King-side.

In fact, the Meran variation has the only established opening line (according to Pedersen) with four queens on the board:

1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Nf3 e6 5.e3 Nbd7 6.Bd3 dxc4 7.Bxc4 b5 8.Be2 Bb7 9.e4 b4 10.e5 bxc3 11.exf6 cxb2 12.fxg7 bxa1=Q 13.gxh8=Q [diagram]









Four Queen Madness

If Black had been castled, this whole variation folds since the rook would be protected.  Indeed, in all other variations of the Meran, this threat remains just that—a threat.  If White knows what he is doing, Black cannot actually pull this off with advantage.  However, now that castling has occurred, the b-march becomes a viable refutation in several lines.
 

Analysis

[Note: Colle-Zukertort players can get away with only studying a fraction of the analysis here if they use a slightly less ambitious move order.  See Recommendation section at the end of this article.]

1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 c5 5.b3 Nc6 6.O-O Bd6 7.Bb2 O-O 8.dxc5 Bxc5 9.a3 [diagram]








 Black has a move to use before his bishop gets kicked in the face by b4.  This in itself poses a bit of a problem because it’s unclear how Black should develop.  9…b6 turns out to be, relatively speaking, one of the best options (which says something about Black’s position, for how often do you see White forced to fianchetto in the Meran!!)

The other main option is 9…Qe7, planning to retreat the bishop to b6.  Black has a small benefit in this line compared to the more normal Meran (where due to move-order differences …a6 (here a3) has not been played), for in the normal move order …Ba6 (here Ba3) is a powerful tool in that line.

Other developing plays should not turn out well for Black. 9…Bd7 cannot be right, and retreating the bishop early does not help Black since White would then not use a move on b4, instead getting a nice game by simply preparing c4 with Nbd2.  Black certainly cannot push through ...e5-e4 with such a quick retreat (9...Bd6?! 10.Nbd2 e5 11.c4 e4? 12.Bxf6).

However, retreating without playing …e4 is no picnic either.  For example 11…Bc7 leaves Black gasping to keep all his pieces and pawns safe after 12.cxd5 Qxd5 (…Nxd5 leaves Black even worse as his game resembles the waiter in Goodfellas who is forced to dance as Joe Pesci shoots at his feet.)  The bishop on c7 is just too much of a pain to defend while White feints at other targets (like the h7-pawn, the e5 pawn, and Black’s Queen).

White has an easy time attacking Black’s position with some permutation of Qc2, Rac1, Rfd1, b4, and either Ne4 (threatening Black’s critical f6-Knight) or Be4 (threatening the pinned Knight.)

One creative idea for Black is 9…Qc7, planning…Ne5: 10.Nbd2 Bd6 11.c4 Ne5!?  At the end of the day, though, the Queen’s exposed position on the c-file is too great a liability.

As for non-developing moves, …a6 and …a5 come to mind, as they stop the b-pawn from advancing.  9…a5 has to be bad as it gives up the b5 square: (9…a5 10.c4 with Nc3-b5 to follow when convenient).

The moves we will analyze, then, are the developing moves 9…Qe7 and 9…b6 as well as the prophylactic move 9…a6.

Since this line is nearly an utter novelty, we are more or less working from scratch here.

9…Qe7

9…Qe7  10.b4 Bb6 

As a rule for these variations, Black should not retreat the Bishop to d6.  In this particular case the problem comes from a possible fork on e4.  This fork allows White to push the c-pawn (rather than the b-pawn) with great advantage: 10…Bd6?! 11.Nbd2 e5 12.c4, and 12…e4 is met by 13.Bxf6! Qxf6 (…gxf6?! 14.dxc6) 14.Bxe4 dxe4 15.Nxe4.

11.Nbd2 e5

Other moves give White a great game after simply 12.c4.  11…Rd8 is Black’s best alternative, leading to sharp, wide-open play where it appears each of White’s minor pieces is just a little more useful than their Black counterparts. (e.g., 12.c4 e5 13.cxd5 Rxd5 (13…Nxd5  takes away defense of h7 and fails to even temporarily give more support to e4: 14.b5 Na5 (14…Nxe3?! does not yield adequate compensation) 15.Qc2 g6 16.Bxe5) 14.Qc2 Bg4 15.Bc4 Rdd8 16.Rac1 Rac8 17.Rfd1

12.b5

With the Bishop on b6, the fork in the note to Black’s10th does not work, so White plays in a manner similar to the 4-queen line, knowing his castled position means that Black has to blink first.

12…e4  13.bxc6 bxc6

Black blinks.  Playing either 13…exd3 or 13…exf3 is worse since White will just continue to mow toward a8 with 14.cxb7.

14. Bxf6!

White takes now when Black cannot afford to retake with the queen.

14…gxf6 (…Qxf6?  15.Nxe4 dxe4  16.Bxe4) 15.Be2 exf3  16.Nxf3 [diagram]








While Black has some significant weaknesses, the real issue is that he has trouble activating his pieces (in particular his queen) to get any compensation for his wrecked pawn structure.  White can either play for c4 or Nd4 to consolidate his advantage.

The most obvious thing to check is Black’s possible play down the open g-file:

16…Kh8 17.Qd3 (taking away e4 from the black queen and making room on d1 for the f1-rook, which in turn makes room for White’s king) 17…Rg8 18.Rfd1 Bg4 (…Bh3 19.g3 and Black has no way to make use of the weak light squares.) 19.c4 Rad8 (19…Qe4  20.Kf1) 20.Rac1 Bxf3 21.Bxf3 d4 22.exd4 Rxd4 (…Bxd4  23. Qf5) 23.Qc2

If Black is not going to try to make something of the g-file immediately, he might try stopping (or discouraging) the c4 play.  16…Be6 and 16…Rd8 are the most obvious tries at that, but they each end up putting the Q-bishop in an awkward position.  16…Be6 blocks the qQueen in even more than she already is, while 16…Rd8 can be met by 17.Nd4, threatening to fork the queen and rook.  17…Bd7 blocks the rook that was just moved to the d-file while 17…Bb7 looks dumb and in fact is rather a blunder due to 18.Bd3!  with the threats of Nf5 and Qg5.

An inventive try is to first feint with 16...Be6 and then after 17.c3 (allowing Qa4, adding a supporter to c4 while hitting c6) switching back to the g-file play: 17…Kh8 18.Qc2 Rg8 19.g6, and once again Black’s queen is hard to activate.

Given the problems Black has in activating his queen, he might try 16…Qe4 immediately, but the queen does not have many safe squares to go to, and I think in general poses little threat.  White should just prepare to push c4 with 17.Rc1.  If Black tries to discourage this with 17…Rd8, White can ignore him and play 18.Bd3 Qg4 19.c4 dxc4 20.Rxc4 Qxc4 21.Bxc4 Rxd1 22.Rxd1, and the two Bishops do not make up for Black’s pawn structure.  Note that picking on the a-pawn after 17.Rc1 backfires: 17…Bc4 18.Rb1! Bxa3?! 19.Nd4, and Black’s position begins to crumble due to a host of threats.

9…b6

9….b6  10.b4 Be7

As described in the last variation, the Bishop almost never belongs on d6.  In this case the problem is that it then blocks the queen from d5 and d3.  The first is significant because the queen might want to recapture on d5 to add support to the knight on c6 [which can become rather fragile in this variation].  The second is significant because playing the bishop to e7 will generally force White to give support to his own bishop by way of Qe2.  In addition to these points, the bishop is here vulnerable to a knight landing on either c4 or e4.  As White can push either of these pawns, such a landing is quite likely.

A sample line after 10…Bd6 is 11.Nbd2 Bb7 12.c4 Qe7 13.Rc1, with the basic plan of pressuring the knight on c6 with cxd5, Qa4, Bb5, and perhaps Nd4.  The knight doesn’t have many good places to run, and the queen cannot defend it without allowing a self-pin.  Black can play 13…Nd7 to give the knight an “out” square on e5, but after 14.cxd5 exd5 15.Qb3, the d5-pawn becomes high –maintenance.

11.Nbd2 Bb7 12.c4 [diagram]

And now we have a position that simply must be good for White because it corresponds to a particularly aggressive version of the Meran (with a move in hand) where White (here Black) has gone astray.

I’m referring to the line 1.d4 d5  2 c4 c6  3.Nc3 Nf6  4.Nf3 e6  5 e3 Nbd7  6.Bd3 dxc4 7.Bxc4 b5  8.Be2 Bb7  9. O-O a6, which Pedersen suggests is just a bit too ambitious because White can cause Black trouble with 10.e4 c5.

But now imagine that Caissa gave Black a free move, so that he could play …c5 before White played e4…then imagine that White responded with b3 (which he has a TERRIBLE record with (5 wins, 11 draws, 16 losses!!!), even without Black having the free move!

Then we would have, after 10. --- c5  11.b3 Bd6 12.Bb2 O-O the reversed version of the position shown.  Starting with a variation that already sees Black trouncing White and then giving Black a free move should certainly lead to more than an adequate position.

9…a6

If your opponent has a keen sense of the Meran and understands the importance of the b-pawn march, you might find yourself up against a plan involving …a6 and …b5.  This doesn’t do much for development, but does ice White’s chief methods of gaining an advantage through advancement of the b- and c-pawns.

9…a6  10.Nbd2 b5  11.Qe2 [diagram]

White connects his rooks, puts pressure on Black’s Q-side pawns, and most importantly protects his d3-Bishop in view of the possibility of the d-pawn being exchanged.