Lets Take A

Look...
Nigel Davies



We invite you to submit games to be considered by Nigel in this column. For
all games submitted, please provide the following information: (1) Names of
both players; (2) Ratings of both players; (3) When and where the game was
played; (4) The time control used in the game; and (5) Any other information
you think would be helpful for us to know. Please submit the games (in PGN
or CBV format if possible) to: nigeldavies@chesscafe.com. Who knows,
perhaps you will see the game in an upcoming column, as Nigel says to you,
Lets take a look...
Computers as Shrinks
In an attempt to overcome limitations inherent in conventional
computer-aided diagnosis, investigators have created programs that
simulate expert human reasoning. Hopes that such a strategy would
lead to clinically useful programs have not been fulfilled, but many of
the problems impeding creation of effective artificial intelligence
programs have been solved. Strategies have been developed to limit the
number of hypotheses that a program must consider and to incorporate
pathophysiologic reasoning. The latter innovation permits a program
to analyze cases in which one disorder influences the presentation of
another. Prototypes embodying such reasoning can explain their
conclusions in medical terms that can be reviewed by the user. Despite
these advances, further major research and developmental efforts will
be necessary before expert performance by the computer becomes a
reality. Abstract from a paper by Peter Szolovits, Ph.D.; Ramesh S.
Patil, Ph.D.; and William B. Schwartz, M.D.; Cambridge and Boston,
Massachusetts
Medical science has been investigating the concept of computer diagnostics
for a while, theres quite a way to go. The problem for computers is that their
reasoning may be insufficiently subtle to unravel more complex human
ailments, and the problems will be much greater when dealing with the
complex and often contradictory features of the human mind.
Little wonder then that computer aided chess annotations are often
disappointing. The computer will spot tactical opportunities, but computer
generated verbal comments are usually quite inappropriate. The computer
doesnt understand whats happening in a game, it merely computes
variations and finds those that are most favourable according to its algorithm.
As artificial intelligence stands at the moment, the best the programmers can
do is find things that are fairly bland and hopefully amusing.
One particular area in which computers cannot help is in understanding
patterns of mistakes and their origin within the human psyche. In my own
work as a coach Ive found that discovering the cause of such patterns is
exceptionally useful to a player as they can then go about improving. When
errors seem to occur randomly and for no particular reason its very difficult
to know how to address them.
Perhaps worse still is when the computer detects no error at all, but then
suddenly pronounces your position as lost! This months game, sent to me by
Valence Jordan, was a good example of this:
I recently lost an important game in my pet Caro-Kann. Chessmaster
10th Edition found no fault with the play on either side until I was
already lost! I made a commitment not to lose a game without learning
from my mistakes. What are they please?
Well, on the basis of this game there is a common thread to the errors, that of
excessive passivity. But where exactly does such a tendency come from?
Whilst Id need to see more games for confirmation, Ive tended to find that
players learn to keep their heads down when theyve been beaten too many
times in tactical battles, so they try to avoid this happening by playing
solidly. But the weakness is still there just waiting to be found, and
meanwhile many of the strategic battles will also be lost because the player
concerned will retreat rather than have the pieces come into conflict.
Curing this isnt easy because the weakness may well be there because of a
plain lack of tactical talent. Without that its very difficult to help people
improve, you can try to work around it, but progress will be slow.
A N Other Jordan,V
Internet Game
Caro-Kann [B19]
1 e4 c6
The Caros not a bad idea for the club player, especially those who are less
tactically inclined. One line that Black will tend to get in practice is 2 d4 d5 3
e5 c5 4 c3 Nc6 5 Nf3 Bg4 6 Be2 e6, with White often losing the d4-pawn
within a few moves. Club chess is not the same as world championship chess,
a factoid that is often lost on both club players and those who claim to write
books for them.
2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5
With 4...Nd7 having been all the rage in recent years, the text is almost an
oddity. Which probably means its a good time to start playing it again.
5 Ng3 Bg6 6 h4 h6 7 Nf3 Nd7 8 Bd3
8 h5 Bh7 9 Bd3 is more usual, with similar positions to the game, just with
Whites h-pawn on h5.
8...Bxd3 9 Qxd3 e6
At one time 9Qc7 was considered more
accurate because of Whites reply, but
then they discovered that it probably
transposed back anyway.
10 Bf4 Qa5+ 11 Bd2 Qc7 12 0-0-0 Ngf6
13 c4 Bd6
This natural move is arguably a bit
passive. 13...b5 is a more incisive
continuation, fighting for the d5-square. A
correspondence game Meyer Shakarov,
USSR 1979 continued 14 cxb5 (14 c5 is probably better) 14...cxb5+ 15 Kb1
Bd6 16 Ne4 Nxe4 17 Qxe4 0-0 18 Ng5 and now 18...Nf6 19 Qf3 Rac8 would
have been the simplest line, with rather the better game for Black.
14 Ne4 Nxe4
The immediate 14...Bf4 may well be better, leading to simplifications though
not necessarily equality.
15 Qxe4 Nf6 16 Qe2 Bf4?!
This exchanges another pair of pieces, but it leaves Black quite passively
placed. The game Dembo Werner, Budapest 2003 witnessed a somewhat
more active treatment for Black in 16...0-0 17 Bc3 b5, when 18 c5 Bf4+ 19
Kb1 Qd8 20 g3 Bc7 21 g4 Qd5 22 g5 Qe4+ 23 Qc2 Qxc2+ 24 Kxc2 Nd5
gave him quite a good position.
17 Ne5 Bxd2+ 18 Rxd2 Nd7 19 Rh3 Rd8?
This should have led to even worse
trouble than it did in the game. Black
should try to simplify here with 19...Nxe5
20 dxe5 0-0, though admittedly this is far
from pleasant after 21 Rd6.
20 f4
Not bad, but not the best either. White had
a much stronger move in 20 Qg4!, the
point being that 20...0-0? loses to 21 Rg3
g6 22 Nxg6, etc.
20...Nxe5?!
Black will find it difficult to get counterplay after this exchange, one of the
main problems being that a later ...c6-c5 will be met by d4-d5. He should
play the cold-blooded 20...0-0, after which 21 Rg3 Nf6 doesnt seem to be
fatal. Black is ready to play 22...c5.
21 fxe5 g6?
And this is very bad, not just weakening Blacks kingside but ruling out
counterplay based on ...f7-f6. For better or worse Black had to play 21...Qe7,
intending to castle and then play ...f7-f6. Hes worse, but not decisively so.
22 Qf1 Qe7 23 g4?
Letting Black off the hook for a second time. 23 Qf4 is much stronger,
stopping Black from castling by hitting h6 and after 23...h5 waiting for Black
to commit his king to the kingside before playing g2-g4. It all looks very
unpleasant for Black.
23...h5! 24 g5 Rf8??
As there was no indication in Blacks notes that this was a mouse-slip, I will
work on the assumption that it was deliberate. In this case one has to ask why
Black should have preferred it to 24...0-0, still protecting the f7-pawn, but
with the added bonus of getting the king to safety and connecting the rooks?
Whatever the answer it turns a reasonable position into a very bad one, which
is why I have awarded it two question marks.
25 a3
Another good move would have been 25 Kb1.
25...Rd7 26 Rhd3 Rg8?
The proverbial fiddling whilst Rome
burns. Black has to try getting his king to
the queenside with 26...Kd8, when 27 d5
doesnt quite seem to put the ball in the
net after 27...exd5 28.cxd5 cxd5 29.Rxd5
Rxd5 30.Rxd5+ Kc7, etc.
27 Kb1?!
White should probably have shown a little
more urgency with 27 d5; for example,
27...exd5 28 cxd5 cxd5 29 Rxd5 Rxd5 30
Rxd5 Kf8 31 Qf6, etc. Now Blacks king gets a second chance to head for the
hills on the queenside.
27...Kf8
27...Kd8 was still the best attempt to get his rooks connected. Now his
position is deeply horrid.
28 Qf6 Ke8 29 d5 cxd5?
29...Qxf6 was probably the best try, but in any case its bad for Black after 30
exf6 exd5 31 cxd5 c5 32 Rc3 b6 33.b4.
30 cxd5 Qxf6 31 exf6
And not 31 gxf6? because of 31...Rxd5 32 Rxd5 exd5 33 Rxd5 g5!, finally
getting his rook on g8 back into the game.
31...e5
The e5-pawn drops after this, but its
probably the best try in relative terms. 31...
exd5 32.Rxd5 Rxd5 33.Rxd5 leaves Black
helpless against an invasion by Whites
king, moving the king to f8 allows Rd8
mate.
32 Re2
32 d6 is also very good, threatening 33
Rc3.
32...Kd8 33 Rxe5 Kc7
Or if 33...Re8, there might have followed 34 Rxe8+ Kxe8 35 d6, etc.
34 d6+ Kc6 35 Re7 Rgd8 36 Kc2 b5 37 Kc3 Kc5 38 b4+ Kc6 39 Kb3 a6
40 a4 bxa4+ 41 Kxa4 Rxd6 42 Rc3+
42 Rxd6+ Kxd6 43 Rxf7 is also good.
42...Kb6 43 Rxf7 Rd4 44 Rfc7 R8d6 45 R7c4 Rxc4 46 Rxc4 Rd8 47 Rf4
Rf8 48 f7 Kc7 49 Ka5 Kb7 50 Rf6 1-0
Recommended Reading
Rapid Chess Improvement by Michael de la Maza (Everyman Chess, 2002)
Copyright 2007 Nigel Davies. All rights reserved.

[ChessCafe Home Page] [Book Review] [Columnists]
[Endgame Study] [Skittles Room] [Archives]
[Links] [Online Bookstore] [About ChessCafe] [Contact Us]
Copyright 2007 CyberCafes, LLC. All Rights Reserved.
"ChessCafe" is a registered trademark of Russell Enterprises, Inc.