Author: Mark Young
Date: 15:01:54 09/28/99
Go up one level in this thread
On September 28, 1999 at 13:14:49, Dann Corbit wrote: >On September 28, 1999 at 11:02:43, eric guttenberg wrote: > >>If my memory is correct, in the first 10 games H7.32 failed to win a single >>game and was behind 2.5-7.5; in the next 15 games Hiarcs has won 7 or 8 and >>has scored 10.5-4.5! Maybe Junior will go on a streak soon...or maybe not. >> >>I have never seen a clearer demonstration of the futility of basing an >>opinion of program playing strength on a handful of games. > >That is another tremendous value of having tests of this nature. How often have >we seen a post like: >Title: "Program foobar stinks!" >Content: Program foobar lost miserably to program barfoo! What a waste of >magnetic domains on a plastic substrate! Look at the foolish move it made on >move #<val>! I wish I had never wasted the money on that heap of rubbish!!!" >{Game score goes here} > >If the SSDF did not exist, we would have to invent them. >IMO-YMMV. As predicted Hiarcs 7.32 came back against Junior. I found the same up down results when I was testing Hiarcs 7.32. I found the book to be the cause of most of Hiarcs 7.32 bad results. When I made a new book for it out of the Junior 5 book with some GM games added. This nonsense stopped and Hiarcs 7.32 because a world beater in my testing.
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.