Author: Dann Corbit
Date: 14:53:01 08/15/01
Go up one level in this thread
On August 15, 2001 at 17:37:10, Roy Eassa wrote: >On August 15, 2001 at 17:05:22, Dann Corbit wrote: > >>A very entertaining contest. >> >>The comments of the IM's should be worth their weight in gold for the engine >>programmers. Nothing like a good public scathing to make someone perk up and >>take notice. >>;-) > > >I read all the IMs' comments. They are indeed VERY entertaining, but I wonder. >They are talking about the programs as if they (the programs) were people. How >Fritz and Junior are brothers and play the same. How they are disappointed in >Hiarcs because they think it plays so well normally. How Rebel Tiger is unsound >and relatively weak, whereas Gambit Tiger is sounder (!). How CM8000 is a good >tutor for kids, but they (the IMs) are surprised it's doing well and shocked >that it could ever beat Hiarcs. How Genius plays "the same" as the Tiger >programs. > >As we all know, these characteristics are bogus. It's all statistics. A single >game between programs can result in almost any style of game and any result. >They'd have to look at a 20-game match between two programs before any comments >about styles and relative strengths they (the IMs) make would have any validity. > Failing that, it's kind of like reading your horoscope: it may sound plausible, >but it's all hokum. > >My $0.02. I agree with you on the personifications. The stuff useful for the programmers will be stuff more like: "Look at this idiotic move..." "Doesn't this program know anything about pawn structure?" "I'd have to give that move a ??" etc. The general impressions will be far less valuable. It is the specific details that merit keen attention.
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.