Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: More Shredder 4.2x games

Author: Enrique Irazoqui

Date: 08:16:30 08/26/00

Go up one level in this thread


On August 26, 2000 at 10:27:51, Uri Blass wrote:

>On August 26, 2000 at 09:58:43, Enrique Irazoqui wrote:
>
>>On August 26, 2000 at 09:45:41, Uri Blass wrote:
>>
>>>On August 26, 2000 at 09:30:33, Enrique Irazoqui wrote:
>>>
>>>>On August 26, 2000 at 09:07:23, Uri Blass wrote:
>>>>
>>>>>On August 26, 2000 at 08:53:45, Enrique Irazoqui wrote:
>>>>>
>>>>>>Shredder won the WMCCC, and even if there is an important luck factor in any 9
>>>>>>rounds event, the fact that it won for the second consecutive time, together
>>>>>>with the quality of its games, makes quite clear that sheer luck was not the
>>>>>>reason for Shredder's success. In my opinion, Shredder is the strongest in the
>>>>>>endgame and, together with Junior, has probably the best evaluation functions of
>>>>>>all programs. On the other hand, it lacks the search speed of Fritz, Tiger,
>>>>>>Nimzo and The King.
>>>>>
>>>>>Are you sure about it?
>>>>>Stefan told me that one of the things that he worked about was tactics.
>>>>
>>>>The new Shredder has improved a lot in tactics, among other things, and Shredder
>>>>5 will be improved further, but it is still far in this regard from the very
>>>>fast searchers.
>>>>
>>>>>My chessmaster6000(ss=10) needed some hours to avoid Qxb4
>>>>>r3rb1k/6p1/p2p3p/1ppb4/PB6/5NRP/1q3PP1/1B1QR1K1 b - - 0 1
>>>>>
>>>>>I think that Fritz will also have problems with this position and my
>>>>>only program that can avoid Qxb4 is Junior5.9 (and I guess that also Junior6a).
>>>>>
>>>>><snipped>
>>>>>>Shortly before London, I played these games with Shredder 4.22 on 2 P600E
>>>>>>machines with 256MB RAM and the same WMCCC time controls. S4.22 beat neatly the
>>>>>>other 2 betas I made it play against, by 6-0 in one of the matches. The overall
>>>>>>results were great, and the final Shredder 5 will be even better, I'm sure of
>>>>>>it. Enjoy.
>>>>>
>>>>>I do not see 6-0 in the games that you posted but only draws against Fritz6b and
>>>>>Junior6a.
>>>>
>>>>I don't post beta games, unless the programmer gives his permission.
>>>>
>>>>>It is a disappointing result after the impressive result in WMCCC.
>>>>
>>>>It is not disappointing at all to tie against Fritz and Junior. If you add up
>>>>the great results against the 2 betas I mentioned, then you have that Shredder
>>>>did very well in my games. I have been not surpised at all at Shredder's
>>>>performance in London. I expected it to do very well in there and I'm glad it
>>>>did.
>>>
>>>After knowing that Shredder won the WMCCC people could expect Shredder to win
>>>Junior6a and Fritz6b because My guess was that Fritz6 beta is better than
>>>Fritz6b(I understood that Fritz6b is available with the young talents)
>>
>>It is.
>>
>>>I also remember the prediction by Marcus Kästner that chessbase is not going to
>>>win because other programs(probably Shredder) are at least 50 elo better than
>>>Fritz6a.
>>
>>The X program of Marcus wasn't Shredder.
>
>I do not know if X played in the tournament and I did not talk about X.
>
>Marcus said that Fritz needs to be at least 50 elo better in order to win the
>tournament and I guess that he meant that Shredder is at least 50 elo better
>than Fritz6a.

Again: Marcus' X was not Shredder.

More: none of us had enough time to be sure about any program being 50 points
better than Fritz. And even if we did, a 9 games event has a margin of error
much, much bigger than 50 points. So ir was all guessing, educated guessing, but
guessing.

>I also remember that he agreed with the prediction of thorsten:
>Rebel,Shredder,Fritz.
>
>I guess that he menat that both Rebel and shredder are at least 50 elo better
>than Fritz6a.

I would disagree, strongly.

>>>after reading it everything that is less than a win for Shredder is
>>>disappointing.
>>
>>That's irrational. Shredder is great, but not invincible. These top programs are
>>very close to each other.
>
>Shredder should be better than Fritz6b and Junior6a in order to have equal
>chances to Fritz(london)
>
>The draws against Fritz6b and Junior6a do not prove that it is not better(there
>are not enough games) but they are disappointing if you assume that shredder and
>Fritz6(london) are in the same level.
>
>Uri

Uri: put things together. S4.22 tied against F6b and J6a, crushed Z 6-0 and beat
Y neatly. Altogether, I find it impressive.

Enrique



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.