Author: Frederic Friedel
Date: 02:09:27 04/22/00
Go up one level in this thread
On April 21, 2000 at 10:32:34, guy haworth wrote: > >Turing actually played a game against someone, manually calculating the >position-scores and emulating the 'program' himself. [ He lost quite quickly. ] > The game was against Alick Glennie in Manchester 1952. The paper engine with Turing as the CPU was searching three ply and lost in 29 moves. >Yes, it would be fun to have this 'program' functionally recreated: it must be >quite trivial with the Crafty infrastructure. > We have reconstructed the engine, which I have on my hard disk as turing.eng for Fritz and ChessBase. It was done by Ken Thompson and Mathias Feist. We are still in the process of checking it. Apparently the Turing CPU made some errors during the game, most remarkably by playing 1.e4. If you look at the rules carefully (they are posted somewhere below) you will see that the first move should be 1.e3. At least that is what we think. The engine plays an interesting amateur game. If you allow it to go more than three ply (which was the artificial limit set by Turing) it would easily beat Glennie. If anyone is interested it taking an indepth look at the reconstruction of the Turing engine please contact me.
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.