"Colle System: Zukertort Variation 5 b3" by GM Andrew Soltis,
1998 Chess Digest, Softcover, English Algebraic Notation, 102pp.,
$16.50

"Colle System: Koltanowski Variation 5 c3" by GM Andrew
Soltis, 1998 Chess Digest, Softcover, English Algebraic Notation,
99pp., $16.50

Reviewed by Glenn Budzinski

Soltis' Colle: Yesterday's Lunch

The Colle System, 1 d4 d5 2 e3 Nf6 3 Bd3 e6 4 Nf3 c5, is the
subject of two recent monographs by Andy Soltis. One focuses on
the Koltanowski Variation, which occurs after 5 c3, and named
after the man referred to as the "Dean of American Chess", while
the other examines 5 b3, known as the Zukertort Variation, and
named in honor of the 19th Century player who once challenged
for the World Championship.

Each book is about 100 pages in length and is similarly formatted
with a brief introduction, followed by five chapters of material.
Unusual for an opening book, theory is not discussed until the last
chapter.

Taking a look first at the "Colle System Koltanowski Variation, 5
c3", the line that is probably most recognizable as the Colle System
proper, one finds chapters titled "How White Wins; How White
Loses; Pawn Planning in the Middlegame; The Colle Endgame,
and Early Main Lines", the latter of which provides a theoretical
discussion of the main variations. Chapter One, "How White
Wins", includes several complete games that cover such strategems
as the White bishop sacrifice at h7 against the castled Black king.
"How White Loses" examines games where White goes awry by
selecting the wrong plan, "dilly-dallies", or plays the Colle against
a system designed specifically to render it ineffective (i.e., Black
places pawns on d5, e6 and f5). Chapter Three, "Pawn Planning in
the Middlegame", looks at the pros and cons of playing for the
isolated pawn as White. "The Colle Endgame" discusses positions
where White uses his queenside pawn majority to play for an
advantage in the ending. The two principal opening variations,
Black's responses 5...Nc6 and 5...Nbd7, are presented in Chapter
Five. 

Since the book is primarily intended to be opening literature, it is
the material in the last chapter that this writer will examine in
detail. The main lines, as given in "Encyclopedia of Chess
Openings, Volume D" (ECO) are 1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3
c5 5 c3 Nc6 and 5...Nbd7, identical to those identified by Soltis.  

In the 5...Nc6 line, both Soltis and ECO cite the same game,
Koltanowski-Soultanbeieff, 1935, and provide similar assessments
of equality after 20 Rad1 (Soltis) or 20...Bc7 (ECO). 

After 5...Nbd7, Soltis follows ECO's recommendation of 6 Nbd2
Be7 7 0-0 0-0 but deviates from 8 Qe2 with 8 e4. In a footnote,
ECO mirrors Soltis' sub- variation 8 e4 de4 9 Ne4 cd4 10 cd4 Ne4
11 Be4 Nf6 12 Bc2 b6 and evaluates the resulting position as
equal, citing the game Konstantinopolsky-Tolus, 1948 USSR
Championship. (The game Welling-Speelman, 1993 Tilburg,
omitted by Soltis, reached a similar position after 8 e4 de4 9 Ne4
Ne4 10 Be4 Nf6 11 Bc2 b6 12 Bg5 Ba6 13 Re1 Rc8 14 Qd2 cd4
15 Qd4 Qd4 16 Nd4. It was eventually drawn.) Soltis, however,
continues with 12 Qe2 cd4 13 Rd1 d3 14 Bd3 Qc7 15 Bg5 Bb7 16
Ne5, noting that White had a "slight but solid" advantage in
Colle-L. Steiner, Budapest 1929 (page 99).

Current White 8th move alternatives found in this writer's CD
ROM game database include 8 Re1, 8 Qe2 and 8 Ne5, none of
which are discussed by Soltis. 

There is also no coverage accorded to the Black kingside fianchetto
or King's Indian setup, cited by Tim Harding in his 1979 work the
"Colle, London and Blackmar-Diemer Systems" as "a good reason
why White should not go 3 e3 unless Black has played ...e6." After
the moves 1 d4 Nf6 2 Nf3 d5 3 e3, a line given by Harding goes
3...g6 4 Bd3 Bg7 5 Nbd2 0-0 6 c3 c5 7 0-0 b6 8 e4 cd 9 e5 Nh5 10
cd Nf4 11 Bb1 Qd7 12 g3 Nh3+ 13 Kg2 Qg4 14 Nb3 f6 "and
White was under pressure".

Additionally, ECO under variation D04, offers 6 0-0 Bf5 7 Qe2
Nbd7 8 e4 de4 9 Ne4 Ne4 10 Be4 Be4 11 Qe4 c6 12 Bg5 Nf6 with
equality in Pedersson-Cobo, 1953 Helsinki Olympiad.

While Soltis does not discuss the g6/Bg7 line, he includes the
game Franklin - Wade, 1961 British Championship, where Black
also fianchettoed his king bishop   after 1 d4 d5 2 Nf3 Nf6 3 e3
Nbd7 4 Nbd2 c5 5 c3 Qc7 6 Qa4 g6, and eventually won (page 28).

The first five chapters of the "Colle System Zukertort Variation,
5.b3" cover topics similar to the Koltanowski book. Chapter One
compares and contrasts 5.c3 and 5.b3; White winning strategies,
which usually involve kingside attacks made potent by the
fianchetto of the White queen bishop, are discussed in Chapter
Two; White's losses, which can occur by ignoring the strength of
Black's queenside play, are examined in Chapter Three; Chapter
Four looks at strategical mistakes by both sides, and the final
chapter consists of a theoretical overview of two principal
variations, 1 d4 d5 2 e3 Nf6 3 Bd3 e6 4 Nf3 c5 5 b3 Nc6 and
5...Nbd7.

A primary strategical difference between the Koltanowski and
Zukertort Variations, according to Soltis, is that the White center
might not be quite as solid in the Zukertort as it is in the
Koltanowski, since it is only bolstered by the pawn at e3 rather
than pawns at e3 and c3. On the other hand, in the Zukertort, Soltis
mentions that White's "queen bishop will probably have much
greater reach and flexibility. It also means that he will have holes
at c3 and a3 that Black can exploit. Usually the advantages of the
reach outweigh the holes" (page 14).

Taking a detailed look at Solits' opening theory on the Zukertort,
one of the main lines follows the 1923 game Euwe - Rubinstein,
which went 5 b3 Nc6 6 0-0 Bd6 7 Bb2 0-0 8 a3 b6 9 Ne5 Bb7 10
Nd2 Qe7 11 f4 Rfd8 12 Rf3 Ne4 13 Rh3 f5 14 Bxe4 dxe4 15 Qh5.
While all three sources, Soltis, ECO and Harding, include this
game, only Soltis clearly states that the resulting position is
advantageous for White. Harding provides the complete game, won
by White, but does not indicate his evaluation of the line, while
ECO considers the position after 15...Be5 16 Qh7 Kf7 17 fe5 Rh8
to be unclear (which is, perhaps, somewhat gratuitous in light of
the game continuation of 18 Qxh8 Rxh8 19 Rxh8, when White
went on to score the full point.) However, ECO and Harding
diverge on move 10 and consider Duz-Hotimirski-Nimzovich,
1907 Karlovy Vary (Carlsbad) to be the main line after 10...a6 11
f4 b5 12 dc5 Bc5 13 Qf3 Rc8 14 Qg3 Ne5, assessed as even by
both sources. In any event, one can't help but agree with Soltis
when he writes that "White's chances certainly look good here"
(page 98), although 10...a6 suggested by ECO and Harding may be
a Black improvement. (Soltis brushes this move aside with the
comment that "If Black bides his time with moves like 10...a6 and
11...b5, White starts his attack with 11 f4 and Qf3- h3" [page 96].)

Soltis' second principal variation in the Zukertort is 5...Nbd7.
Concurrence between the three sources exists through the moves 6
Bb2 b6 7 0-0 Bb7, when Harding continues by following the game
Spielmann-Stoltz through 8 Nbd2 Be7 9 dc Bc5 10 Qe2 0-0 11 e4
Re8 12 e5 Nh5 13 g3 g6 14 Nd4 Rc8 15 f4 "with a double-edged
game" won by White. ECO also follows Spielmann - Stoltz but
stops after 14 Nd4, assessing White a solid advantage.

Soltis, however, continues with 8 Ne5 and comments that "This is
most accurate since it prevents Black from blocking the attack
lines with 8...Ne4 ..." (page 99) According to him, White is for
choice after 8 Ne5 a6 9 Nd2 b5 10 Nxd7 Qxd7 11 dxc5 Bxc5 12
Qf3 Qe7 13 Qg3 0-0 14 a3. Although Soltis attempts to debunk
ECO's suggested improvement of 8...Be7 with the comment that
White can transpose into the line 6...Be7 7 0-0 0-0 8 Nbd2 b6 9
Ne5 Bb7 10 f4 and should obtain an advantage shortly thereafter
(pages 99 and 100), this writer would have liked to have seen the
1994 game Ubilava-San Segundo included, which contained
8...Be7 and resulted in a Black win. (The game went as follows:
1.d4 d5 2.Nf3 Nf6 3.e3 e6 4.Bd3 c5 5.b3 Nbd7 6.0-0 b6 7.Bb2 Bb7
8.Ne5 Be7 9.Nd2 0-0 10.Qe2 Ne4 11.Bxe4 dxe4 12.Nxd7 Qxd7
13.dxc5 Bxc5 14.Rfd1 Qe7 15.Nf1 a5 16.Ng3 a4 17.Qg4 f6
18.Nh5 Kh8 19.a3 Rf7 20.b4 Bd6 21.Rd2 b5 22.Qe2 Qc7 23.Rad1
Bd5 24.Nf4 Bxf4 25.exf4 Qxf4 26.Qxb5 e3 27.fxe3 Qxe3+ 28.Kh1
Rc7 29.Bd4 Qxa3 30.Bb6 Rc3 31.Bd4 Rc7 32.Bb6 Rc3 33.Bd4
Rc4 34.Bc5 h6 35.Qd7 Rxc5 36.bxc5 Qxc5 37.c4 Qxc4 38.Rd4
Qe2 39.R4d2 Qe4 40.Rf1 a3 41.Qc7 Qb4 42.Qc1 a2, White
resigned.)

Regardless of whether White chooses to meet the 5...Nbd7 line of
the Zukertort with either 8 Ne5 or 8 Nbd2, all three sources appear
to agree that White should be able to reach the middle game with at
least a small advantage.

A curiosity is the date of the Spielmann-Stoltz game cited above.
This writer found it in four separate sources: Soltis, ECO, Harding
and a large CD ROM database, and each identified it as having
been played in a different year. Soltis gives the year as 1933 (page
77), ECO offers 1932, Harding-1943, and the CD ROM database,
while agreeing with ECO on the year, gives an entirely different
game score: 8 Nbd2 Qe7 9 c4 Rd8 10 a3 e5 11 cxd5 Nxd5 12 dxe5
Bxe5 13 Nxe5 Nxe5 14 Bc2 Bg4 15 Qe1 Nb4 16 axb4 Rxd2 17
Qxd2 Nf3+ 18 gxf3 Bxf3 19 e4 1:0

While both Soltis books do a reasonably good job of highlighting
certain recurring strategical themes in each variation, unfortunately
this information comes at the cost of reduced theoretical coverage
and praxis. Even if the games serve no other purpose than
entertainment value, the reader should be assured that if he
purchases a book on a specialized opening or variation, he will
receive a relatively large selection of games played in that opening.
There are only about 20 complete games presented in each book.
Given the fact that this writer's database turned up dozens upon
dozens of Colle games from the 1980s and 1990s alone, to include
only 20 for each variation is insufficient for proper coverage in a
specialized opening book, at least in this writer's opinion. This
inadequacy shows itself in the lack of discussion of certain lines,
such as the King's Indian configuration, cited by Tim Harding back
in 1979 as a critical line against the Colle, 8th move alternatives
for White in the Koltanowski, etc. 

Following are two interesting miniatures omitted from the
Zukertort work. Olesen,M.-Shirazi,K. Chicago, 1992  1.d4 e6
2.Nf3 c5 3.e3 Nf6 4.Bd3 d5 5.b3 Nc6 6.0-0 Bd6 7.Bb2 0-0 8.Nbd2
Bd7 9.a3 Rc8 10.Ne5 Be8 11.f4 Qb6 12.dxc5 Qxc5 13.Nxc6
Qxe3+ 14.Kh1 Bxc6 15.Bxf6 Bxf4 16.Qg4 Bh6 17.Rae1 Qxd2
18.Re2 Qa5 19.Re3 Kh8 20.Qh5  1:0 Shereshevsky - Kupreichik
Minsk, 1976   1.d4 Nf6 2.Nf3 e6 3.e3 d5 4.Bd3 c5 5.b3 Nc6 6.Bb2
cxd4 7.exd4 Bd6 8.a3 0-0 9.Nbd2 Nd7 10.0-0 e5 11.dxe5 Bc7
12.c4 d4 13.Be4 Ndxe5 14.Nxe5 Nxe5 15.Nf3 Qd6 16.Nxd4 Nd3
17.Bxh7+ Kh8 18.Qh5 Nf4 19.Qh4 Nh3+  0:1

The reader should also be aware of the nontraditional format
employed by Soltis in both books. A standard structure for an
opening monograph is the dedication of a chapter for each major
variation and the inclusion of a sufficient amount of material
within that chapter to thoroughly cover the line. Soltis' method, on
the other hand, is to focus primarily on certain recurring themes
and, only in the last chapter, present a cursory review of theory.
The effectiveness of this approach remains to be seen.

Neither book contains a bibliography nor an index of
players/games. The lack of the latter can be especially frustrating
given the format of the books, which does not sort material by
variation.

Both of these books left this writer wanting. Regardless of whether
it be a want of more theory, more analysis, more games, or more
ideas by the author, in the case of these two works, more of
everything is needed. Either book may be marginally acceptable
for a less experienced player who wishes to get his feet wet with
the Colle System, but neither should be depended upon as a
primary opening reference. Nor is either a particularly good value,
even at $16.50 each, a relatively bargain basement price for an
opening book these days. Occasionally, even good writers have off
days. Perhaps the best way to view these books is as skeletons in
Soltis' closet.