CHESS: A PSYCHIATRIST MATCHES WITS WITH FRITZ
A review by JAMES DuBOIS
Dr. Ernest F. Pecci, M.D. has written a physically formidable how to book on
the subject of defeating our beloved (most of the time) Fritz. As expect from
a lifelong psychiatrist this nicely bound book is quite academic in nature with
its gold lettered spine and burgundy covers. The dust jacket is littered with
praise and hyperbole from both ends of the ELO scale. A certain Mr. Wilmont
McCutchen is quoted as saying that Dr. Pecci's Bird's Strategy raised his rating
by 200 points, while Garry Kasparov proclaims the book as having "...a
far reaching influence on shaping new ways to approach the game of chess ...."
The four hundred plus pages are overflowing with strategy and diagrams with
some interesting personal insights by the author.
This reviewer must take at face value that every word in the foreword was actually
written by Garry Kasparov. He briefly mentions Deep Blue (of course), Advanced
Chess and Rapid Chess. There is a definite sense that he truly does believe
in Pecci's system and makes an almost spiritual connection with the "creative
intuition" aspect in the battle of man versus machine.
We learn from the author's preface that he began playing chess again after
Kasparov-Deep Blue, 1996 even though he hadn't played since the early 1980's.
He started with Chessmaster 3000 and progressed to Fritz 6. He writes that this
program is considered to be nearly unbeatable and was frustrated by it. With
some "slight adjustments" he was able to defeat it from a number of
openings in less than 40 moves. "I had developed strategies and tactics
that were considerably different from modern chess theory to combat ideas that
had strongly influenced the chess database [sic] programmers." A healthy
ego at work. At this point I was extremely anxious to test his theories, but
I also could not help from reading on when encountering statements such as,
"...White should be able to win or draw every time regardless of what black
does." Conversely it would have been as challenging for Pecci to finish
that sentence with, "...regardless of who or what Black is!" He does
confess, "In enjoying a game for pleasure, I might, sometimes take one
or two 'blunder' moves back...." "I would gladly offer the computer
the same courtesy, but why should a computer need it? Computers don't make mistakes."
We are introduced to Edward Sheppie, a well rounded chess player who discusses
his relationship with Dr. Pecci and how his chess was changed forever. Sheppie
writes, "There is a certain sense of exhilaration associated with beating
Fritz 6 in a blitz game without having to take any moves back." He tells
that Fritz had a more than adequate time allotment for the games played in the
book to allow a full search to arrive at the best move. The scope of this work
is seemingly as endless as the game itself. I must reiterate, that this is a
physically formidable book. That doesn't translate to dull by any means, especially
for devoted Fritz users. More short chapters follow with the emphasis on rethinking,
that our "surface minds" are no match against a computer. Agreed.
Pecci acknowledges several times that chess playing programs do come in an assortment
of strengths. "...the highest rated programs... put you on the defensive
very quickly." I believe for the most part he does know the subject matter
well enough to author a book of this stature and that he has a deep love for
chess, computers and the human mind. Our kind of guy. "...there is enjoyment
in often disagreeing with the evaluation it (Fritz) gives your move selection."
Sure, if you win.
I disliked the author's parallel of chess as a military game. It works for
the purpose of the book in the middle chapters as Pecci outlines "establishing
an attack trajectory to the castled King", but I think we (certainly ChessBase
users) are far more sophisticated to perpetuate this metaphor. This leads to
the main concepts of the "Barrage" position. It is essential to obtain
an inverted triangle of three pawns. White wants pawns on d4, e3 and f4. Ideal
is the Bird opening. Then push the Kingside pawns to the fifth rank as quickly
as possible while contemplating 0-0-0 or just playing the King up one square.
"A pawn chain obstructs the Black pieces from engaging in the defense of
their King." It's not quite that easy and the author warns that the moves
must be precise.
After demolishing the bulk of the Chessmaster series, Fritz 5 is "...approached
with some trepidation." Game one was drawn. The author's resolve was strengthened
to refine his system until he could routinely defeat Fritz in an average of
30 moves from a variety of openings. His second game actually brought Fritz
to its digital knees. The bulk of the book (300+ pages) consists of the author
illustrating how his barrage position system wins with a variety of openings.
He includes many standard openings with emphasis on Bird's Opening (because
of Fritz's consistent response) for the human, of course not omitting From's
gambit and with a lengthy chapter, "Fritz plays White." Fritz wouldn't
be the same without his personality. The author found the licking sound of Fritz's
chops "annoying" as he goes down two pawns in a game to ultimately
come out the victor in 36 moves. The book closes not with glorious praise or
condescending conclusions, but with one last game against Fritz 6 as the author
woke from sleep with some self doubt, "...despite hundreds of victories...."
"The result was one of my most satisfying games." This with the program
set at its highest level.
Ernest Pecci Fritz 6 [C26] 11.07.2001: 1.e4 e5 2.d3 Nc6 3.c4
Bc5 4.Nc3 Nf6 5.Be2 d6 6.f4 0-0 7.f5 Nd4 8.Nf3 c6 9.h4 Re8 10.Ng5 h6 11.Bh5
hxg5 12.hxg5 Nd7 13.Bxf7+ Kxf7 14.Qh5+ Ke7 15.f6+ Nxf6 16.gxf6+ gxf6 17.Qh7+
Ke6 18.Nd5 Re7 19.Qh3+ f5 20.Nxe7 Qxe7 21.Qh8 f4 22.Rb1 Nc2+ 23.Kf1 Nb4 24.Rh7
Nxd3 25.Rxe7+ Kxe7 26.g3 Kf7 27.Bd2 Nf2 28.Qh7+ Kf8 29.gxf4 1-0.
Replay the game here
With all this said, the acid test for the book comes down to one single question.
Does the method outlined by Dr. Pecci work for players other than him? The only
acceptable way to give credibility to his system is if any player of average
playing strength (or stronger) can assimilate the theories outlined well enough
to win at regular intervals.
I began with some play on the Internet Chess Club site. My rating for blitz
on this server generally stays around the 1700 mark. I sought out my online
computer nemesis JSBach. I enjoy playing this program because it does offer
chances instead of the elite (some rated 3000+) and snobbish programs that occupy
the ceiling of the seek graph. I confess that I did loose a fair amount of games
because of the learning curve involved in gaining a barrage position. Most of
those games were closer than usual, but the clock was an adversary since I needed
the few extra ticks to play more precise. This game was my first win. It is
also the first win in my life with Bird's Opening.
Author James DuBois
|
James DuBois (1628) JSBach (1876) [A03] ICC 3 3" [J. DuBois]:
1.f4 d5 2.d4 Qd6 3.e3 the inverted pawn triangle. 3...Bf5 Pecci sites
this a favorite line for computers. 4.Bd3 Bxd3 5.Qxd3 Nc6 6.c3 So far
I'm continuing with the soft moves as instructed. 6...Nf6 7.Nf3 Qe6 This
is why I like playing JSBach, it will play a Q move that is seemingly weak only
to have you chase her and wreck your own position. However, I do see some logic
here. 8.Nbd2 h5 9.Ne5 Double d pawns are okay according to Doctor P.
9...0–0–0 In the classic Pecci dictum I strive for 0–0–0 while my opponent
goes 0–0. 10.0–0 Still the opposite wing castling principle is
just as effective and follows the plan. 10...Rg8 11.b4 Rd6 12.b5 Nd7 I
think the program just didn't comprehend what was going on. Black is now on
the defensive no matter which alternate moves could of been played. At this
point it's pretty much over and I'm wondering how I'll fair against Fritz. Surely
it won't have a brain cramp like this.13.bxc6 bxc6 14.Qa6+ Kd8 15.Qxa7 Nb6
16.Qb8+ Nc8 17.c4 Qf5 18.c5 Rf6 19.Ndf3 g6 20.Ng5 Bh6 21.Bd2 Rf8 22.Ba5 Ke8
23.Qxc7 Nb6 24.Bxb6 Rd6 25.Qb8+ Qc8 26.Qxc8+ Rd8 27.Qxd8# Thank you Dr.
Pecci for the virtual rating points! 1–0.
Replay the game here
I can quite honestly say that I gave the Pecci system a more than fair examination
against Fritz 7 at a variety of time controls ranging from blitz (game in 5
minutes) to games of 40 moves in two hours. Running on a 1 GHz Pentium III machine
with 256 megabytes of RAM and assimilating a new chess paradigm made for excruciating
play on my part. I'm sure that I exhausted Fritz's insult repertoire.
DuBois,J Fritz 7 [A03] 01.04.2002
[DuBois,J]:
1.f4 d5 2.d4 Nf6 3.e3 Bf5 4.Bd3 e6 5.Nf3 I played BxB...my
first "take back" move. 5...c5 6.c3 Nc6 7.h3 N to e5 was tried
as seen many times in the book. 7...Ne4 -0.72/10 11
8.g4 Bg6 -1.04/9 10 9.h4 f, g, h pawns on the
fourth. So far I'm somewhat optimistic as Fritz asks if I switched off his opening
book. 9...c4 -0.85/9 9 10.Bc2 Ng3 -0.63/10
13 11.Rh3 Bxc2 -0.63/12 1 12.Qxc2 Ne4 -0.63/12
13 13.Nbd2 f5 -0.60/11 6 14.Nxe4 This
goes against the Doctor's orders. Keep the pieces on the board. Fritz's Knight
is too well placed for my liking. 14...fxe4 -0.63/10 1
15.Ne5 Be7 -0.38/12 18 16.g5 Nxe5 -0.57/11
4 17.dxe5 0–0 -0.72/11 2 18.Bd2 In many
of the games I found this to be the problem child. 18...Qa5 -0.47/12
16 19.0–0–0 You can supposedly play like this. The principle is
to keep your attack alive while not overly concerning yourself with what Fritz
is doing. 19...Qxa2 -1.88/11 2 20.Be1 Qa1+
-1.94/12 6 21.Kd2 Qa6 -2.07/12 9
22.Kc1 b5 -2.32/12 7 23.Qe2 Rab8 -2.44/10
2 24.Bd2 b4 -2.88/11 3 25.cxb4 Bxb4 -3.32/11
2 26.Kb1 Bxd2 -4.38/10 0 27.Rxd2 Rb5 -5.19/10
0 28.h5 Everything has fizzled as Fritz never came under fire.
But the book made it look so simple! 28...Ra5 -8.88/9 1 0–1.
Replay the game here
The book included over 4,000 diagrams illustrating how Fritz's king is to be
naked at the conclusion of the game. I find it extremely difficult to digest
that it is as simple as stated to accomplish said task. Returning to Mr. Wilmot
McCutchen (quoted earlier) I'd be very interested in seeing how he handles Fritz
7 on compatible hardware since his peak rating was almost identical to mine,
thus putting us both in the majority of the rating pool. This book seems to
be geared for the strong club player up to Expert. I believe that I understood
the concepts, but was it my own weaknesses as a player and/or my psychological
composition that lead to my failure with his approach? If so, then the system
itself fails since it is not universally adaptable. The book is definitely intriguing,
but perhaps more so would be a contest similar to the recently completed Ilya
Smirin match, replacing the GM with the doctor and have him square off against
the ChessBase stable of thoroughbred programs under the same playing conditions.
There is a conviction to the book; otherwise I'm sure Pecci would have never
written it. It is thought provoking, well produced and is of some worth to chess
computer aficionados (primarily Fritz users) if only as a curiosity. I don't
see it as a panacea for the man versus machine battle or as the "how to"
book on defeating Fritz. Perhaps solving the problems associated with love,
children and money will after all prove to be easier subject matters.
REVIEWER'S ADDENDUM
Just as the author questioned himself the night before sending the book to
the printer, so did I question myself as to whether I played to my full potential
regarding Dr. Pecci's system. Without comments or annotation, here is my last
game before pointing my cursor to send.
Fritz 7 DuBois,J [A40] 03.04.2002:
1.d4 e6 2.c4 g5 3.e4 0.66/10 8 h5 4.Nc3 0.85/10
11 Be7 5.Bd3 0.97/11 9 Nc6 6.Nge2 0.85/12
0 d6 7.0–0 1.04/10 7 e5 8.Be3 1.10/11
12 a6 9.Nd5 1.35/9 5 h4 10.Qa4 1.35/9
6 Bd7 11.Qb3 1.32/10 7 Rb8 12.Rad1 1.19/9
4 Bf8 13.dxe5 1.32/10 6 Nxe5 14.Ba7 1.60/10
5 Nf6 15.Nxf6+ 3.63/9 2 Qxf6 16.Bxb8 3.63/8
0 g4 17.Qxb7 6.72/10 7 h3 18.f3 7.60/10
3 [18.Bxc7 6.91/9 4 hxg2 19.Qa8+ #4/3 0
Ke7 20.Qd8+ #2/2 0 Ke6 21.Nd4# #1/2 1 ] 18...Nxf3+
19.gxf3 9.00/9 1 gxf3 20.Bxc7 9.00/9
1 Rg8+ 21.Ng3 10.10/9 3 [21.Kh1 10.75/9
2 Qg5 22.Qa8+ #6/6 0 Ke7 23.Bd8+ #5/5 1
Ke6 24.Nd4+ #4/4 1 Ke5 25.Nxf3+ #3/3 14 ] 1–0.
Replay the game here