Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: CCL results 08-06-98

Author: Mark Young

Date: 07:58:52 06/12/98

Go up one level in this thread


On June 12, 1998 at 07:10:17, blass uri wrote:

>
>On June 11, 1998 at 18:27:25, Mark Young wrote:
>
>
>>>>Just for fun I tried to answer this. My data shows about a 45 point gain
>>>>in rating if you double the speed of a program running on a P II 266.
>>>
>>>what is the fast time control that you doubled in your data?
>>>>
>>>>You would need a P II 93000
>>>>to get a 3.125 point elo gain, if you doubled its speed. Running todays
>>>>software.
>>>
>>>how did you get 93000?
>>>what is the formula that you are based on?
>>>>
>>I took my data that I had  on NPS matches. Noted the change in rating
>>performance for the data I had. What it showed was as computers get
>>faster the change in rating from a doubling get less and less. So from
>>my data I calculated how fast a P II would have to be in MHz to only
>>gain 2 to 5 elo points, if its speed were doubled.
>
>I do not understand what is the formula doubling get less
>you say f(266)=45(you gain 45 elo on aPII266)
>f(93000)=3.125
>but what is the function f you are based on?
>f(x)=c/(x^d) when c and d are constants can be a function.
>
:) I'm not saying. but you should be about to figure it out from the
info given.

>>Based on my limited
>>data.  Now remember this is for a computer playing chess at 3 min a
>>move. If a P II 93000 only played at one sec a move then a doubling
>>would gain more points.
>>The same way as it does today. The faster the time control the more
>>rating gain you will get playing a slower computer. Example PII 300 Vs
>>PII 266 playing the same program. At 3 min a move avg. The ratings gain
>>from a P II 266 to a P II 300 is next to nothing. If you played them at
>>1 sec a move avg. The gain in ratings because much more.
>>I don't claim this will be accurate. As the data was made with many
>>chess programs old and new. But I think it might be somewhat close.
>>
>>Don’t take this too seriously as it was done in fun.
>>
>>The formula was based on the data. Noting the change in rating to the
>>change in speed of the computer. On a doubling at 3 min a move.
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>>>It may not be right. But I gave it a try.:)
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>>My speculation is 10-20 as the current top programs at such fast
>>>>>machines are already so good that they can't be slaughtered with
>>>>>your suggested 5-25.
>>>>>
>>>>>- Ed -
>>>>>
>>>>>
>>>>>>This just will tell us nothing. I suppose you think differently. So can
>>>>>>you please explain what kind of lesson you are expecting from this?
>>>>>
>>>>>>But OK, as I said, this has to be done at least one time.
>>>>>
>>>>>>And BTW I hope to be wrong. I like to be surprised!
>>>>>
>>>>>>    Christophe



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.