Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: next SSDF list...when?

Author: James T. Walker

Date: 10:59:15 10/14/04

Go up one level in this thread


On October 14, 2004 at 12:19:39, William Penn wrote:

>On October 14, 2004 at 09:32:37, James T. Walker wrote:
>
>>On October 14, 2004 at 02:29:53, Uri Blass wrote:
>>
>>>On October 13, 2004 at 20:36:53, James T. Walker wrote:
>>>
>>>>On October 13, 2004 at 08:03:36, Tony Hedlund wrote:
>>>>
>>>>>On October 13, 2004 at 07:21:53, William Penn wrote:
>>>>>
>>>>>>It's good to see these test results being posted here again!
>>>>>>Any idea when the next SSDF list will be published? Sorry if that has already
>>>>>>been mentioned. I was away from this board for awhile.
>>>>>>Thanks,
>>>>>>WP
>>>>>
>>>>>I don't know. It seem like my friends are losing interest. We are now discussing
>>>>>if/how we will continue.
>>>>>
>>>>>Tony
>>>>
>>>>I hope that before you decide to quit you will at least consider using shorter
>>>>time controls.  Nobody plays 40/2 anymore except SSDF.
>>>>Jim
>>>
>>>1)It is not nobody.
>>>As far as I know thorsten is using 40/120 in his tournaments and other people
>>>like Leo use 40/40 on faster hardware that is almost the same as 40/120.
>>>
>>>The time control in WCCC was 60/120+30 but considering the faster hardware it is
>>>even slower time control than SSDF.
>>>
>>>2)The fact that most people use shorter time control is a good reason to have
>>>40/120 because we have enough information about
>>>shorter time control.
>>>
>>>Uri
>>
>>"Nobody" means humans in tournaments/matches not computers.  The WCCC time
>>control is fine.  The fact that SSDF is using old/slow hardware in comparison is
>>not important at this time.  Most people are not using Quads at home now.  I was
>>suggesting the faster time controls to increase data in a given amount of time
>>since I believe that results will not vary greatly from 3 min/move to 2 min/move
>>or even faster.  Getting faster results is more interesting to most testers than
>>sticking to the old 40/2.  (My opinion from experience beta testing with others)
>>Jim
>
>I am only interested in correspondence chess, so the slower the better. For that
>purpose 40/2 is borderline acceptable as a testing time control, but certainly
>not anything faster.
>
>A minimum time control for correspondence play is 1 hour per move, but
>preferably at least 4 hours per move for each player. That comes to 80-320 hours
>of computer time needed for a 40 move game.
>
>Don't say it is impossible. This sort of long computer analysis is being used by
>thousands of correspondence chess players today!
>WP

Never said it was impossible.  You are welcome to run games/analysis at any time
control you wish.  If you think the SSDF or any other testing body will be happy
to do the same you are wrong.  I believe that nothing they do should be
interpreted as being relavent at 4 hours/move.  If you think 3 min/move is good
enough for 1-4 hours/move analysis you are in a small minority.  None of this
has anything to do with the direction the topic was drifting towards.  The SSDF
is losing interest in testing even at 40/2.  I was only expressing my wish that
they consider shorter time controls before quiting completely.  It seems the
SSDFs main purpose was to provide computer chess enthusiast with data on
different programs strength relative to each other.  They started with the 40/2
time control because it was fashionable and pretty much standard.  It is out of
fashion now and I have maintained for the past 3/4 years that they should
consider faster time controls to keep up with the times.
Jim



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.