Computer Chess Club Archives




Subject: Re: What is Botvinnik's legacy to computer chess?

Author: leonid

Date: 17:27:03 02/21/00

Go up one level in this thread

On February 21, 2000 at 13:41:01, Christophe Theron wrote:

>On February 20, 2000 at 20:59:06, leonid wrote:
>>On February 20, 2000 at 14:39:24, Christophe Theron wrote:
>>>On February 20, 2000 at 01:39:09, Drazen Marovic wrote:
>>>>What is Botvinnik's legacy to computer chess?
>>>That to write a good chess program it's better not to be a strong chess player.
>>>Strong players have too much prejudices about how to write a chess program.
>>>In order to write a good chess program you must be ready to forget all you know
>>>about chess, and re-discover it completely. Strong chess players have spent
>>>years in learning the game and improving their play, so they are obviously not
>>>ready to get into the process of forgetting everything.
>>>In the future, it will happen to me too. Chess programming will certainly
>>>evolve, but as I have spent nearly 20 years in programming Chess Tiger, I'll
>>>probably be unable to adapt myself to totally new techniques. I'll become a
>>>dinosaur and young programmers will take over me - and all my peers.
>>>I hope this is not going to happen tomorrow, though.
>>>    Christophe
>>I believe that reality is even more simple that what you have said. Problem is
>>not that much with the "big secret" to descover, in order to write the good
>>chess game from scrach, problem is in the hardware speed. 20 or 30 years ago it
>>was a shame and now it is only slightly better. Good game of today is able to
>>see, without any artificial tricks, only 6 plies in few seconds in mid game. All
>>the beauty of the best games is that they provide us with some decent chess game
>>even before our hardware will permit it. Good game make all kind of sofisticated
>>acrobatics in order to glance some next 4 or 6 plies deep into game, to make it
>>enjoyable right now. But when the hardware will privide us with all those 10, 12
>>plies brute search to start with? Then writing the new good chess game from
>>scratch will become as simple as doing our three years of college.
>Check your numbers. Most top commercial programs will compute 10 plies in the
>middlegame in a few seconds (in say, less than 10 seconds) on current fast PCs.
>Most of them compute 13-15 plies at 40 moves in 2 hours in the middlegame.
>    Christophe

What computer and how many average number of nodes have each ply? When the
number goes considerably down it is possible for 8 plies to be executed in few
seconds. No hash tables and nothing like null moves. I said about the most
demanding part of the game.

I know the normal speed of the top comercial games since I bought few ot them
recently in order of finding where is my speed. I hope that I know what is the
speed even if all those numbers are hidden by strange "extension" effect. If
what I see is true, my numbers for 8 plies is on the level of the top games and
for 6 plies is even much better.


This page took 0 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.