Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: shredder 5 not the best

Author: Enrique Irazoqui

Date: 04:29:13 11/09/00

Go up one level in this thread


On November 09, 2000 at 04:20:41, pavel wrote:

>On November 09, 2000 at 04:08:33, Uri Blass wrote:
>
>>On November 09, 2000 at 03:58:47, pavel wrote:
>>
>>>On November 09, 2000 at 00:56:15, Uri Blass wrote:
>>>
>>>>On November 08, 2000 at 23:39:25, pavel wrote:
>>>>
>>>>>On November 08, 2000 at 17:34:00, Marcus Kaestner wrote:
>>>>>
>>>>>>On November 08, 2000 at 07:59:09, walter irvin wrote:
>>>>>>
>>>>>>>shredder 5 is not the best program , that would be deep shredder,followed by
>>>>>>>deep junior.a multi proccessor program is going to be the best.
>>>>>>
>>>>>>not at all.
>>>>>>in the chessbits-tournament rating list is deep junior (though on double-fast
>>>>>>hardware) not the first. see:
>>>>>>
>>>>>>shredder x 2717 (40 games)
>>>>>>century 3 beta 2656 (26 games)
>>>>>>tiger 11.9 2624 (12 games)
>>>>>>shredder 4s(chessbits-style) 2620 (28 games)
>>>>>>deep junior 6 2616 (130 games)
>>>>>>tiger 12 (new style) 2612 (31 games)
>>>>>>fritz 6 2600 (231 games)
>>>>>>tiger 12 2582 (120 games)
>>>>>>shredder 4 2581 (221 games)
>>>>>>
>>>>>>even there are not enough games for some programs you can see, that deep junior
>>>>>>hardly is the best.
>>>>>>
>>>>>>but i admit, that a dual-shredder 5 would have the best chances to become number
>>>>>>1.
>>>>>>
>>>>>>marcus
>>>>>
>>>>> I dont get this.
>>>>>
>>>>>In your earlier posts you said something like "shredder5 is far from being the
>>>>>best",
>>>>>this is based on the fact, according to you, that you have shredder5.
>>>>>but, in your tests shredder5 clearly seems to be more than 100elo better than
>>>>>fritz6, and well ahead of others. (that is if you believe your test results).
>>>>>
>>>>>i am kinda confused, do you mind clearing it up?
>>>>>
>>>>>thanks
>>>>>pavs.
>>>>
>>>>It is very clear.
>>>>Tiger13 and gambittiger are not in this list and Marcus claimed that they are
>>>>not worse than shredder5.
>>>>
>>>>Uri
>>>
>>>even if I consider that, ( and also rely on this list) tiger 13 has to be more
>>>than 100 elo better than previous version (tiger12) to be better than shredder5,
>>>I "seriously" doubt that.
>>>Not to mention I believe (not sure) I have seen it stated in the rebel webpage
>>>that the latest version is 40-50elo better than the previous one. and gambit
>>>tiger is supposed to be on the same leage.
>>>
>>>so still it's not clear.... (as to what he meant)
>>>
>>>pavel
>>
>>It is clear to me.
>>
>>I read at the rebel site that tiger13 and gambittiger are only 50-70 elo better.
>>
>>The results that I see suuport this and I asked marcus about it.
>>
>>He replied that tiger13 is really only 50-70 elo better at short time control
>>but the improvement is clearly bigger at longer time control.
>>He claims that the new tiger earns more from time than the old tiger.
>>
>>We need to wait to the ssdf to see if he is right about it.
>>
>>Uri
>
>
>its probably clear to the sense what marcus "wanted to say",
>but it is, in no way, clear that "shredder5 is far from #1" or "chess tiger is
>better than shredder5"
>even if I believe in what you say, chesstiger still has to be +100elo better
>than the previous version in "longer time control".
>But again with all "ifs" and "buts" aside, I agree with you that the only way to
>find out is to make SSDF play these fine engines.
>and really dont rely much on marcus's list...
>
>pavs

After what I have seen in my own games and testing, Fritz 6a/b, Shredder 4.22,
Tiger 13.0 and Gambit 1.0 are roughly equal in strength, say +- 20 points at
slow time controls. Any claims about a new program being 100 points stronger
than F6a/b, which in some cases means 170 over the prior version of the same
program, doesn't deserve any credibility.

Enrique



This page took 0.01 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.