"Test, Evaluate and Improve Your Chess: A Knowledge-Based
Approach" by Danny Kopec and Hal Terrie, 1997 Hypermodern
Press, English Algebraic Notation, Paperback, 235pp., $17.95

Reviewed by Glenn Budzinski
                         
As the title implies, Kopec and Terrie's "Test, Evaluate and
Improve Your Chess" is a book of chess tests. In fact, there are six
examinations included, each selected with the purpose of helping
to identify a player's knowledge of the game as expressed in terms
of playing strength, or rating.

According to Kopec, an International Master and a computer
science academician who runs an annual summer chess camp, the
book contains "an appropriate test for nearly everyone: there are
test positions suitable for players of all levels covering all phases
and most aspects of chess play - and after you take these tests,
study the answers, and review your results, your chess will
improve!" (page 3).

The six tests examine the student's knowledge of a total of 158
different positions. The first of the six chapters covers the 24
position, "Bratko-Kopec Test", which originated in 1980 and was
used principally "to demonstrate that computer programs lacked
certain knowledge which is critical to strong chess players" (page
2), along with the "New Position" test, which added an additional
24 positions. Chapter Two consists of the "Camp Test 1995" used,
as the name implies, at Kopec's 1995 chess camp. A 35-position
rook and pawn test can be found in Chapter Three; Chapter Four
provides a 27-position test of "Other Endings" and Chapter Five is
a 24-position "Novice Test", meant for players with a USCF rating
of 1500 or less. The final chapter is a one page conclusion. There
are also four appendices, a bibliography, a listing that briefly
describes each test position, bios of the authors and an
introduction.

A reasonable way to validate the authors' claim that their tests
accurately identify a player's knowledge (or lack thereof) is to
actually take the tests. This writer, who has maintained a relatively
stable USCF 2100 tournament rating for more than ten years (with
a peak in the 2170s to a low of 2057), appeared to be a good
subject.

Three tests were taken: the Bratko-Kopec, Camp Test 1995 and the
Rook and Pawn Endings Test. In each case, the two minutes per
move time limit was adhered to as closely as possible and answers
were recorded in the manner suggested.

On the Bratko-Kopec test, which consists of a variety of complex
middlegame positions taken mostly from older master games and
books, a score of 19 out of 24 was achieved, equivalent to a strong
master rating.

The Camp Test 1995 covered a diverse range of positions from the
early opening to the late endgame. Five of the twenty-four
positions came from the Pergamon book "Mastering Chess: A
Course in 21 Lessons", co-authored by Kopec. This writer found
this test to be considerably more challenging than the
Bratko-Kopec, and it was reflected in his score of 14.8 which did,
however, fall in the range of the other two 2100 players cited in
Appendix 3 who took the test. (One scored 13.83 and the other,
16.08. It should be noted, however, that in most statistical circles, a
sampling of only three players, as is the case here, would not be
considered large enough from which to deduce any firm
conclusions about identification of knowledge/rating.)

Unfortunately, this writer's copy of the book contained six blank
pages in the Rook and Pawn Endings section, which meant that
nine positions were omitted altogether and three answers were
unavailable. Since no explanation of scoring could be found
(which may appear on one of the missing pages), there was no way
to form any generalizations about the 10 out of 23 score, other than
the fact that this writer's rook and pawn ending skills could use a
little work. Reviewing the Levenfish and Smyslov classic "Rook
Endings", cited in Kopec and Terrie's bibliography, it should be
pointed out that positions on the test did seem to address many of
Levenfish and Smyslov's 22 concluding maxims.

Similar to other tests that one encounters in life, chess tests also
contain an element of controversy. Kudos to the authors for
highlighting some potential problems and criticisms at the end of
Chapter One (pp. 26-28). For instance, a factor that affects results
is the two minute per move time limit. As K & T mention
(comparing the thought processes of humans versus machines),
"During actual games human players tend to non-uniformly
allocate their time to individual moves. Thus a chess master
typically spends ten or twenty more minutes in a position which is
identified as 'critical'for finding a key move or a key plan..."

K & T additionally advise us that "Certainly, a few positions in the
experiment are not ideal, and a few are even controversial as to
what the best move is, but this will not significantly invalidate a
human or machine subject's overall score." Perhaps, but why not
just avoid controversial positions altogether? One such position is
#57 from Camp Test 1995, a variation of the Schliemann Defense
to the Ruy which occurs after the moves 1 e4 e5 2 Nf3 Nc6 3 Bb5
f5 4 d4 fxe4 5 Nxe5 Nxe5 6 dxe5 c6. Although White's best move
may be the piece sacrifice 7 Nc3, chosen by Kopec and Terrie, the
outcome up the road is not as clear as the authors seem to indicate.
They provide the continuation of 7...cxb5 8 Nxe4 d5 9 exd6 Nf6 10
Qd4 and conclude "with a dangerous attack" for White in their
solution (page 84), which flies in the face of two respected sources
on the Schliemann, Mikhail Tseitlin, in his 1991 book "Winning
With the Schliemann" and Jimmy Adams' 1982 monograph
"Schliemann/Jaenisch Gambit". Both Tseitlin and Adams continue
the 10 Qd4 line by following games: Tseitlin cites Black as having
the advantage after move 17 in Gonzales-Montalvo from 1978 and
Adams offers Semkov-Antonov from 1980, a 24 move draw which
mirrors Gonzales-Motalvo at least through move 17. On the other
hand, Tseitlin does concede that Kopec and Terrie's choice of 7
Nc3 is probably best in that particular position, although the
subsequent continuations hardly look dangerous for Black.

Another controversial position is #53 from Camp Test 1995, which
is reached after the moves 1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Qd7 5
a3 Bxc3+ 6 bxc3 b6 7 Qg4 f5 8 Qg3 Ba6 9 Bxa6 Nxa6 10 Ne2
Nb8 11 Nf4 Nc6 and now White plays 12 Nxe6 because, according
to the authors, 12...Qxe6 13 Qxg7 Qg6 14 Qxh8 "favors White
since Black is unable to trap the White queen." (page 83) Would
someone please show all of us non-USCF 2400 players, who have
not seen (or cannot remember) IMs Mike Valvo and John Watson
and GM Ian Rogers' 1989 analysis in "Inside Chess", why the
White queen can't be trapped?

While the tests do seem to do a reasonably good job of pinpointing
one's present playing strength (at least in the case of this writer),
how much a player's game actually improves after taking them and
studying the answers, is less than clear. Other than the rook and
pawn endings test, positions tend to be isolated examples from
which it is difficult to draw general inferences as to whether or not
(and to what degree) a certain type of position is understood. Thus,
what does it really mean if one incorrectly answers the question on
isolated pawns? In many cases, the sample of positions concerning
a specific topic is just too small to establish any definite conclusion
about one's knowledge (or lack thereof) of that concept.

Perhaps this writer should be accused of chess insanity, but to him,
Kopec and Terrie's tests actually seemed enjoyable to take. In
addition to the entertainment value, the reader will also get a good
idea of the breadth of his chess knowledge compared to others. As
long as too much more is not expected, this unusual book should
prove to be a worthwhile addition to most anyone's chess
library.