Author: Kevin Mulloy
Date: 17:51:53 01/19/99
Go up one level in this thread
On January 19, 1999 at 15:54:23, Martin wrote:
>I have read some of the articles/reports here on computer chess , and almost
>every time I see a comment like "We all loose 99% of the time"
>"need better teaching from the program " etc.
>What is wrong with you guys ??
>They are not that hard to beat ,they still have weaknesses like eating
>the pawn on b2 while their kingside is beeing run over and other horizon
>effects.Don`t give up but keep on playing till you find the program`s
>soft spots and the points will start coming in !
Dear Friend,
You have a great attitude, but beat these newer programs?? I assume that
you mean at tournament time 40/120? You are far better than I! I have played
with several of the "top" programs over the last 6 years and I can't beat them.
I don't keep all of the games that I have attempted, but I suppose that I have
tried 200+ games at 40/120 and I have not won 1. For the programs that allow me
to handicap by rating like the current Fritz 5.32 (Remember Socrates 3.0?) I can
win some points up to around 2000, but it gets tough to draw (for me)after that!
I was able on 2 occasions to beat 2 of the top programs (Genius 3 and Hiarcs 5
-- 2 of the top programs at the time I ran my little test) during 6 "simulated"
correspondence games. I used every chess library tool at my disposal (including
the "Master Books" for Genius3) and I won 2, and lost 4. I started running
professional programs on a Pentium 60 and I did keep records on some of the
advances in speed that these hardware/programs have made. Example:
MCP5 P60 8 megs ram = 400,000 positions per minute.
MCP5 P60 8 megs ram (With P120 upgrade processor) 540,000 positions per minute.
MCP5 P166 16 megs ram 1,000,000 positions per minute.
MCP7 P166 80 megs of ram (with 200 MMX upgrade processor)1,700,000
MCP7 PII 450 256 megs of ram 2,700,000 positions per minute...
The point is that in 6 years my rating has gone up from about 1850 to 2050 in
postal chess -- thats about 30 points a year or a 2% gain per year (and it is
far easier to avoid point dropping blunders in postal!) the programs and
hardware have increased in speed(in the same amount of time) by at least 600%!!
I realize that hardware speed is only one measure of performance -- but I also
realize that the professional programs are "out-running" me by a wider distance
with every release (program or processor). The net results in my opinion? I'm
as close as I'll ever get to beating one of the top programs at tournament time
controls. I am not complaining though! As a matter of fact, I will continue to
purchase the new programs to support the people that write them because I
believe that the programs and computers are the biggest reason that I can play
as well as I do. Look at it this way: If you could go down to your local gym
every day for a year and "shoot hoops" with the 6th grade basketball team or you
could play the recently retired #23MJ every day for the same year -- at the end
of the year, with whom would you have learned more? Which would have made you
the better player? I would never beat MJ either, but I would be a better player
if I practiced against him. This is just my opinion, I could be wrong...
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.