How the Kolt Beats a Horse
(Early …Ne4 Plans in the Colle System)

Whoa there!  Colle-Who and Colle-Which?

Oh, hi.  Glad I’m not talking to myself.

Given the prevalence of the Colle System, it is surprising how many players forget it has two completely separate systems.  They both typically begin 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 c5, but they discourage …c4 in different ways:

Incidentally, most people who only know of one version think 5.c3 is the Colle System. This is rather peculiar since at anything above Expert level, 5.b3 is more common and performs much better in practice.

There is a widely circulated myth that the Colle does not perform well in 2000+ play, yet the 5.b3 option boasts an impressive 58% score!  That statistic, taken from the 800 or so games I could find between players rated above 2000, becomes more impressive when one realizes that most of the time Black is the stronger player in games where White plays the Colle-Zukertort.

Then why the myth?

Honestly, the Colle provokes a certain degree of disgruntlement because it is hard to beat.  Palliser, Watson, and Schiller have all pointed out how hard it is for Black to win against the Colle, and it is easy for this to beget resentment that finds its fruit in constant belittling.

There is another reason though — one of the Colle’s greatest assets can turn out to be an obstacle.  The Colle can be played against almost any setup without getting crushed.  It makes a good choice for casual and improving players who do not want to spend a great deal of time studying lines.

Unfortunately, when those “improving” players improve, they sometimes fail to update their repertoire.  At some point you have to stop treating the Colle as one-stop-shopping and work it into a repertoire with other openings.  Just as a 1.e4 player cannot play the Ruy Lopez if Black plays 1…c5, the Colle player has to flesh out his arsenal.

For example, after 1.d4 Nf6 2.Nf3 d5 3.e3 Bf5!, White can do okay with 4.Bd3, but to play for the greatest advantage he really has to play 4.c4, transposing into a Slav sideline that (based on analysis I did for Zuke ‘Em) is better than people realize.

Okay, now that I’m up to speed on the two Colle systems,
why is …Ne4 so much more difficult for the c3-Colle player?

To put a finer point on it:

• The C-K player typically relies on e4 to free his dark-squared bishop.  Obviously, e4 is unlikely to happen any time soon after Black plays …Ne4.  The C-Z player develops this bishop to b2, so the inability to play e4 is irrelevant.

• After …Ne4, the C-Z player can often target the g7-pawn (which the knight used to be shielding) with his bishop on b2.

• Playing c4 and Nc3 is a standard response to an early …Ne4, both attacking the d5-pawn that the knight used to protect but now is protected by.  Unfortunately, the former would be a loss of tempo after White has already played c3, and the latter is impossible for the C-K player because he tends to play …Nbd2 very early.

So, what is the Koltanowski player to do?

Well, if Black is going to get …Ne4 in, he pretty much has to play an early …b6 & Bb7 because C-K players play Bd3 and Nbd2 very early.  So, dealing with the threat of …Ne4 really comes down to responding to an early …b6.

Koltanowski himself met such a Q-side fianchetto by playing Ne5.  Palliser suggested the same in his book Starting Out: The Colle.  I could not find anything better when I wrote Zuke ‘Em, so I suggested this move for C-Z players.  Unfortunately, further analysis has shown it is not sufficient for an advantage.  White typically has a hard time getting anything more than equality using a Stonewall when Black has not played …Bd6, and that is true here as well.

In fact, Black has two comfortable routes to equality (or better) against a C-K player using Ne5:

Black already has the advantage after 11…Rad8!, 11…Rac8!, or (my recommendation) 11…a5!, when White’s g-pawn is just a liability.  He cannot push it with advantage because exchanges on e4 and e5 will undermine it, as Nick Defirmian illustrated in 1999 against Berg.  For example, 11…Rad8! 12.g5 Nxe5 13.fxe5 Ne4 14.Nxe4 dxe4 15.Bxe4 Bxe4, and the g-pawn drops off the board next move.

To determine a solution to this problem, we have to go back to the root of it.  Let’s ask ourselves What gives Black the right to make such an ugly, arrogant move so early?  Playing a knight to such an advanced post (as Black no less!) so early seems somehow unjust, and we think White should have some way of punishing Black’s brashness.

To understand why Black can make these moves safely, we must look at what weaknesses his play introduces and ask ourselves why White is unable to target those weaknesses.  Then the answer becomes rather clear.

The weaknesses arising in Black’s camp are concentrated on the a4-e8 diagonal (you know, the one Black’s King is on…since Black has not taken time to castle). In particular, playing …b6 has weakened c6 and the Knight that jumped to e5 used to cover d7.

Unfortunately, normal Colle play is particularly ill-suited to taking advantage of this weakness.

White would like to play Ne5, Bb5(+), and perhaps Qa4 to cause problems along this diagonal, but such plays cut against the grain of the Koltanowski variation. Obviously, White has already played his Bishop to d3, so now Bb5+ would represent a wasted tempo in some sense.

Secondly, the Colle is quiet enough that Black may delay …c5, meaning he has the …c6 “biff” in his pocket as a response to Bb5+.

Lastly, the cruel, tragic truth is that White generally needs to play Qe2 to push e4 (to prevent …Ne4), but it is this exact move that often leaves him unable to meet …Ne4 because once the queen moves to e2, she cannot go to a4 (or even c2 to at least pressure the knight after it lands on e4).

Fashioning a satisfactory involves making use of a few key observations:

• White’s opportunity for play along the a4-e8 diagonal increases significantly once Black plays …c5.

• Black is unlikely to play …Nc6 if he is trying to make use of an early fianchetto, this means there is less pressure on d4 and no chance of …Nb4, hitting White’s bishop.

• If Black plays …b6 and later needs to play …b5, it represents a loss of tempo.

Now, what kind of stew can we make from the three observations above?

Good Question!  My pot is empty.

Don’t feel bad.  You have to knock down a few walls and rip up some flooring to find the solution, which involves two major principles:

1.      When possible, we will disrupt Black’s Q-side by playing Bb5+ immediately after …b6.  To get the most from this play, we switch up the Colle move order and play Nbd2 before Bd3.

2.      If Black plays …c5 after already having played ...b6, we do not play c3 in response.  We welcome …c4 because we will now be a tempo up on those lines where Black had not played …b6.

Explain that second one.

After …c4, Black will have to play …b5 soon to defend the pawn.  However, he has already played …b6.  The tempo lost in moving the b-pawn twice turns out to be rather critical.

Doesn’t playing Nbd2 so soon have some other problems?

I don’t think so, not for the C-K player at least.  Indeed, it allows them greater flexibility in meeting deviations because the extra defense given to the knight on f3 can be useful against …Bg4 deviations and the option of playing c4 and retaking with the knight (rather than retaking with the bishop that has already moved) should be quite useful.

You said “not for the C-K player at least,” does that mean
C-Z players should stick with their normal move order?

Yes.  Playing an early Nbd2 is a significantly more delicate matter for the C-Z player.

The biggest issue is that if he plays main-line moves he will find himself in the jaws of Prie’s trap after, say, 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Nbd2 c5 5.b3 Nc6 6.Bd3 Bd6 7.Bb2 0-0 8.0-0 Qe7 9.Ne5 Qc7!, when 10.f4?! cxd4 11.exd4 Nb4 forces White to part with his bishop.

In an earlier Chessville article [Reviving the Colle-Zukertort Main Line] I gave the improvement on this line, but it gets ruined if White plays an early Nbd2.

On the other hand, as I mentioned earlier, C-Z players are not in nearly as much of a pickle when it comes to meeting …b6. Even better, C-Z players can transpose into one of the lines I give here without changing their move order with 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 b6 5.Nbd2! Bb7 6.0-0 c5 7.Ne5.

Oh, you have some lines?  Show me.

I’m about to, but first let’s discuss how the general plan unfolds and why it works.

Keep in mind that the goal here is not necessarily to get a great position.  The goal in these lines is to dodge a bullet and make Black think and sweat a bit for equality.  Even when Black gets an equal position, it will be the kind of position that C-K players are used to.  Hopefully, he will also have a good deal more time on his clock too!

White’s strategy is based on a few points:

• After Black has played …b6, White does not need to castle before playing e4 if either of the two conditions below are true:

  • Black has also played …Bb7.

  • White has not played c3.

• With a proper move order, White can provoke …c5 after …b6 by Bb5+. Afterward, White need not worry about …c4.

• Black must either play …Nbd7, …c5 or …Be7/d6 before he can safely play …Ne4 due to the threat of Ne5+Bb5

The above, when taken together, mean that White can always safely play e4 before Black can play …Ne4.

I see how developing his bishop helps him get away with …Ne4
(he can castle quickly),  and how …Nbd7 helps (he can lop off the
knight that comes to e5), but how does …c5 help?

For those who cannot believe Black is okay after the text, keep in mind that after 10.Nxe4 dxe4 11.f3 a6 12.Bxc6+ Bxc6 13.fxe4 Bxe4 14.O-O f5, the standard assault 15.Qh5+ is not the knock-out blow it often is.  After 15…g6 16.Nxg6 Qf7 17.Nf4 Rg8, White might be wondering if that pawn was really worth it.

And how does …Bb7 help White to castle safely?

And how does c3 make it harder for White to play e4 safely?

If White is in a position where he has to rely on e5! as an answer to …cxd4, he ends up losing the pawn-marauding war because Black’s pawn has a clear path to White’s king via c3 and d2.  If there is no pawn on c3, Black’s pawn becomes stranded on d4 in this continuation.

Thus, White will end up doing one of three things:

1. Play Ne5 with advantage, either because Black played …Ne4 at a bad time or because Black has been coaxed into playing …Bd7.

2. Play e4 after preparing it with Nbd2, Bd3, and (perhaps) Qe2. [Note that in this case White will delay castling to make sure he can get e4 in.]

3. Obtain a very good position after Black plays a misconceived …c4.

Okay, enough generalities.  Show me those lines!

Fine.  We really only need to look at how the Bb5+ finesse unfolds.

This finesse either gains a move (by provoking …Nbd7 or a c-pawn advance that can then be ignored) or forces Black to compromise his position.  White should play this any time Black plays …b6 before developing his queen’s knight.

If Black plays his queen’s knight out before …b6, the moves for White are pretty easy to find up until at least e4.

Thus, we will look at two lines:

1. 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Nbd2 b6

2. 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Nbd2 c5 5.c3 b6

The first is the only line that is relevant for C-Z players, who can transpose to it without changing their move order.  That transpositional line is 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 b6 5.Nbd2! Bb7 6.0-0 c5 7.Ne5.  See notes to move 7 of the Line A commentary below.

Line A

1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Nbd2 b6 5.Bb5+!?

5…c6

Anything else is fine for White.  5…Nbd7 allows the annoying 6.Bc6 Rb8 7.Qe2!, when Black should know better than to rid himself of his “thorn in the flesh” with 7…Bb7?!, for after 8.Bxb7 Rxb7 9.e4, the undefended rook on b7 adds extra color to exchanges on e4.

More circumspect lines are also pleasant for White — not a huge surprise given that both pawn breaks (c4 and e4) are available to him.

5…Bd7?! 6.Bd3! might look odd, but is actually quite intelligent.  After 6…c5, White can either advance his e-pawn immediately or revert to a typical C-K setup.  To see why White would want to do the latter, see Line B.

6.Bd3 c5

Of course, other plays are possible, but they are unlikely to pose White any difficulties.

White’s plan is to ignore this move, which means he has fewer options (for example, Qe2 is more or less impossible now).  If White finds himself “out of moves” in this line, he should play Ne5.  He will likely find himself up one move on a more normal variation by not having played c3.

Remember, however, that White can often play e4 safely before castling as long as he has not played c3.

7.0-0 c4

The other way of holding back e4 (7…Bb7 8.Ne5 c4) should lead to something similar to what we will see here.  Note that this is where Zukertort players will find themselves after 1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Bd3 b6 5.Nbd2! Bb7 6.0-0 c5 7.Ne5.

White’s plan is to threaten Black’s Q-side pawns and then either lock them up or grab an advantage there.  Either way he is will likely trade off bishops with Ba3.  He should try to get in Ne5 before Black covers that square with a knight.

If Black develops with …Nc6 or …Nbd7 before playing …c4, White should immediately play Ne5 (not waiting for …c4), and he’ll be fine.

8. Be2 Bb7

We will soon see that Black runs into difficulty because he is still two moves away from castling, so it might seem like …Be7 or …Bd6 is the smarter play, however Black can get into big trouble because his a-rook lacks protection (e.g. 8…Bd6?! 9.b3 b5?! 10.a4 a6? 11.axb5).

8…Qc7 is really Black’s only other option here, as it requires White to play 9.c3 before playing b3.  However, Black would still have to play 9…Bb7 to give his rook protection, and play would be similar to the line given here (and likely will transpose) because 8…Qc7 has not stopped White from playing Ne5.

8…Nbd7 would also require 9.c3 and would put the brakes on Ne5, but it doesn’t matter because Black needs to have a piece supporting a possible …Bc6 to prevent a future Qa4 should White take on c4.  Black would really prefer to retake with his b-pawn, but would not be able to after 8…Nbd7 9.c3 Bb7 10.b3 b5 11.bxc4.  White’s position after 11…bxc4?! 12.Qa4 is excellent.

9.b3 b5 10.a4 a6 11.Ne5!

This last move might seem a bit optimistic, but Black does not have time to play …c3 due to his under-defended king.  (11.c3?! Ndf3 12.b4? Ng5!).  Black is not in a position to take those pawns on the Q-side because White can put significant pressure down the a- and b-files.

The diagramed position [above] is quite rich in possibilities due to threats on both sides of the board.  If nothing else, White can lock up the Q-side, play f4, and use his space advantage on the K-side to mount an assault on Black’s king.

Line B

1.d4 Nf6 2.Nf3 d5 3.e3 e6 4.Nbd2 c5 5.c3 b6 6.Bb5

6…Nbd7

Should Black play 6…Bd7 instead, White has already won a small victory.  A bishop on d7 not only gets in the way of Black’s forces but also gives White an easy plan of retreating the bishop to d3, Stonewalling with Ne5+f4, and looking for a g4! break.

7.e4! dxe4

7…cxd4 8.Nxd4 will allow White to play e5 with advantage

8.Nxe4! Nxe4 9.Bc6 Rb8 10.Bxe4 Qc7!

The only move.  Otherwise Bf4 makes Black’s rook cry.

11.Qe2

White threatens d5.

The problem for Black from this position is that he would like to play both …Bd6 and …Nf6 soon.  The knight move helps unwind his position.  In addition to putting his knight on a natural square, it shoos White’s bishop from e4, allowing Bb7.

A bishop on d6 allows Black to continue contesting the e5 square and stops his position from being generally passive.

However, trying to get both of those moves in the near future allows White to sacrifice the minor exchange for excellent control of the center and play against Black’s K-side.  The Black player's going for a win might start down that road, but it is unlikely to take them where they want to go.

The rest of this line is devoted to an example of how such play might unfold.

11…Bd6

This makes d5 less of a concern. Note that Black cannot get around what is coming by playing 11…h6?! — then White can play 12.d5!

12.Bg5 0-0 13.0-0 Nf6?!

There is nothing that says Black has to play this, but otherwise White has a pretty free hand.

14.Bxf6! gxf6 15.d5! f5

Clearly, this is the way to refute White’s play if such a refutation is to be found.

16.Bc2 exd5 17.Rfe1 Qd8

Covering g5 to allow his next move.

18.Rad1 Be6 19.g3!

Black cannot enjoy the prospect of seeing Nh4 come down.

A final note from the author:

I've created 12 training exercises to help you practice the motifs of this lesson.  You can find them here and solutions can be downloaded from here.

Thanks, David Rudel

Read these other articles on David Rudel's favorite opening:

• Four Keys to the Colle-Zukertort - A Car for the Future

• Reviving the Colle-Zukertort Main Line

More Opening Analysis

Chessville
logo by
ChessPrints
 

Ranked #1 by Google
"largest chess database''

The
Chessville
Chess Store

The
Chessville
Weekly
Newsletter

Subscribe
Today -
It's Free!!

The
Chessville
Weekly
Archives
 

Advertise
with
Chessville!!

Advertise to
thousands
of chess
fans for
as little
as $25.

Single insert:
$35
x4 insert:
@ $25 each