Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: New CM9 super strong setting

Author: Dimitris Poulos

Date: 14:06:36 08/05/03

Go up one level in this thread


On August 05, 2003 at 16:07:56, Rajen Gupta wrote:

>On August 05, 2003 at 13:37:29, Dimitris Poulos wrote:
>
>>On August 05, 2003 at 12:35:45, Rajen Gupta wrote:
>>
>>>
>>>
>>>On August 05, 2003 at 11:22:07, Dimitris Poulos wrote:
>>>
>>>>Hi,
>>>>single computer, ponder off
>>>
>>>hi: pointless match: like asking sprinters to run 1 legged hopping race to
>>>decide the 100m champ!
>>
>>Hi,
>>I don't think you have a point here.
>
>i played fritz 8 vs junior 8 on a single comp over 50 gmes and the score was
>substantially in favour of fritz 8; vs both programmes over 50 games each; i was
>not doing  comparative study; i only had 1 comp at home and another one in my
>flat where i worked during the week in a different city. when i closed down this
>flat and moved back home and managed to get both comps next to each other(both
>identical comps) i found the results to be greatly different; junior8 and
>shredder 704 are both roughly level against fritz8:
>
>my entire database of eng vs eng matches on a single comp, i now consider
>worthless; the time mngement of shredder704 and even more so junior8 are
>completely different with ponder on; they (esp junior8)spends a huge amount of
>time in some important positions and much less in simple positions; wheres f8
>does not show such a great variation in time management;
>
>now you will ask how does this apply to cm 9000? and the answer is that all
>engines in order to be properly rated at whatever time control need ponder on!
>of course to amuse yourself you can play any nmber of these matches and make
>whatever comclusions you like, but it still remains a pointless exercise and
>without much value!

I think you need to think about indications in a research process. And then you
shall understand if these results are pointless or not.
What you discribe is a minor value probability. I expect the seting that plays
better with ponder off, to play still better with ponder on( compared to other
settings). And it does. I don't care if Fritz or maybe the King might have some
'bug' in time management when pondering. The bug will be there wheather the
setting is good or not. And there are good settings that find moves that others
cannot. The last comes from experience...




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.