Author: Albert Silver
Date: 16:09:16 01/21/99
I find Nunn's use of computers and his opinions somewhat intriguing. While
it is clear he understands the uses of them, and certainly makes use, as has
been shown in his updated versions of certain classic books, the upcoming NCO,
and the series on simplified endgames.
I have a copy of his recently published Secrets of Practical Chess (which I
have not read bar one chapter though it is on the top of my reading list) in
which there is a final chapter called "Using a Computer". As the whole book is
devoted to teaching what he believes to be essential knowledge to all players,
one would expect this last chapter to reveal techniques or methods to maximize
the use of machines. Not so. Instead he only tells us some of his frustrating
experiences with programs, not realizing that many of the problems encountered
had already been very much resolved 'before' his book came out. Very curiously,
he only mentions Fritz 4 (?!) in the chapter. I found this to be absolutely
astonishing and in my opinion completely negates the value of any opinions he
may have on programs. Obviously, he tells us that they understand nothing of
anything (he presumes ALL programs to be like Fritz 4) and furthermore plays
down any value they might have as a partner. The one positive comment he has is
that they can be useful (glad to know they have 'some' use) to play out certain
positions (as recommended by Mark Dvoretsky). He also shares with us a test in
which he took the 15 most important theoretical novelties from Informant 68 and
put ol' Fritz to the test. The computer only found a lousy three! Which is not
all bad when you think of it he adds. What on earth this is supposed to prove is
beyond me. Moreover, he presents two successes and one failure. The failure was
absurd to say the least. He begins by noting the machine did not choose the
stupendous combination, and furthermore, refuted it by finding a defense making
the whole line unclear at best. Yes, it was clearly a stinking failure....
Before I go on, I'd just like to note for the record that I also think Nunn one
of the BEST writers on chess, so this should be seen as a somewhat
tongue-in-cheek attitude to show that clearly many of the 'expert' GMs still
don't know how to make the most of their programs.
On databases, he says practically nothing on how to use them, merely what
features they should include. He also goes on and on on how bad the available
databases are. In many ways he is quite right when commenting on the general
state of the million game databases available: multiple names for the same
players, multiple names for events, wrong moves, etc.... But, and this is
important when considering that he uses Chessbase, the two highest quality
databases in these regards are without a doubt those issued by Chessbase itself
and Chess Assistant. Instead of just telling us how bad they all are, why not
give us a few ideas on where to get a good one (or doesn't he know?)?
I have just acquired the newest updated database of Chessbase called BIG '99 (I
realize that when he wrote his material, he probably only had access to the
smaller 875,000 game BIG '98). I will make a small comparison to a popular
database with serious header problems, Ultimate Game Collection II (I don't have
the 3rd but imagine it has the same problems):
UGC II (with 1,005,474 games) has 151,607 players listed as well as no less than
41,722 events!
Big '99 (with 1,114,429 games) has 73,050 players and a mere 11,249 events.
Even Chess Assistant 3.0's 800,000 database only had 78,000 players (number of
events are harder to determine in it because of the way it is sorted). I haven't
received the newer 4.0 yet so cannot comment on the 1,000,000 database that
accompanies it.
All in all, I can only say it is strange to read this from someone whom I'd
assumed to be somewhat of an authority on the matter (mostly based on all of the
computer based material he has released).
Albert Silver
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.