Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If programs are only IM strength, why bother to play Kramnik??

Author: Uri Blass

Date: 05:47:43 05/14/01

Go up one level in this thread


On May 14, 2001 at 08:20:45, Joshua Lee wrote:

>700Mhz x 5.9 = 4130Mhz  4130/450 = 9.177777778 hello!!! you are wrong, i am not
>just making up things that i want to believe. Ya know those ratings still are
>not acurate for 1 simple reason they are not FIDE or USCF rated so it is just
>like if i wanted to give them any rating i want and say Chessmaster 3000 is
>rated 2950!!! Sorry but it doesn't matter unless it's official besides the math
>just doesn't work there are some programs which are accurate and some that are
>NOT i will give you that but the difference between an 8way 700Mhz machine and a
>450Mhz K6-2 is 9+ times  like i said before so there is atleast 180points
>difference between the two weather you agree or not if not you better go get a
>new calculator . Thanks

I do not think that there are at least 180 points difference because of the
following reasons:

1)The difference against humans is not identical to the difference against
computers.
2)Even in comp-comp the difference may be less than 180 elo because of
diminishing returns from deeper search.

We do not know if there is significant diminishing returns from deeper search
because it was never tested in these speeds.

We know that at small depth there is not significant diminishing returns but
things may be different at high depthes and it is possible that the difference
in comp-comp games between 450 Mhz and 8*700 Mhz is only 100 elo.

The only way to know is testing.

Uri



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.