Author: Dann Corbit
Date: 10:49:03 01/28/00
Go up one level in this thread
On January 28, 2000 at 11:34:44, Tom Kerrigan wrote: >On January 28, 2000 at 08:09:49, Robert Hyatt wrote: >>As I said... he _believes_ it is the best, because of the test games he played >>vs micros to test this feeling. But in writing, it is logical to be >>conservative so that you don't have to retract something once someone proves >>your statement was "too much"... > >I was thinking about this the other day... >If MChess Pro runs at ~2,500 NPS on a P5/133, then it was almost certainly >running at < 5k NPS when Hsu ran his "micro experiment." >So if Hsu was running his chip at 100k NPS, he was outsearching MChess by a >factor of 20. It's not a huge surprise that DB won. > >The reason I like MChess is because its evaluation function is obviously >extremely sophisticated, and it's proven itself over and over and over, over the >course of ~2 (?) decades. > >I think it's pretty presumptuous to say that function X is better than MChess's >function, unless possibly function X was written by Mark Uniakle or Johann de >Koening. (Whose functions are also sophisticated and well-proven.) I think it a bit odd to compare them at all. They use a very different technology. In other words, "If we could speed up PC's by 1000 times, they would compete effectively with Deep Blue." is a nonesense statement. We might as well say, "If bears could fly, they would be faster than eagles." PC's don't run 1000 times faster and there is no forseeable technology to allow them to do it anytime soon. So we are arguing about a completely untestable hypothesis. Since we have access to NEITHER Deep Blue NOR 1000x speed PC's, it is an exercise in metaphysics rather than science. I therefore hold the position that you can take any position that you like and it is equally valid. In fact, my cat, walking over the chessboard (if I should happen to buy a cat) would bump the pieces to better locations than either of them. Absurd? Maybe. But prove me wrong. You can't. That's because the hypothesis is totally untestable.
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.