Author: Albert Silver
Date: 16:01:20 04/14/01
Go up one level in this thread
On April 14, 2001 at 17:49:49, Rajen Gupta wrote:
>On April 14, 2001 at 17:25:10, Uri Blass wrote:
>
>>On April 14, 2001 at 17:08:43, Rajen Gupta wrote:
>>
>>>i've noticed ever since i've been following computer chess that there is a
>>>programming cycle at the end of which a particular programme gets exhausted and
>>>then reaches a plateau.presumably the programmer has milked everything possible
>>>from the programming technique he uses for his programme, after which unless he
>>>tries something new, there is a flat line.this is esp true for the
>>>champions-once a champion has lost his crown, he never makes a comeback; remains
>>>a has been-genius; mcpro; rebel; hiarcs,
>> now fritz is struggling to retain its
>>>top spot with the twin assault from the 2 tigers!
>>
>>I do not see it.
>>Genius was top programs for years and I believe that the reason it stopped to be
>>a top program is the fact that richard lang stopped to develop his program.
>HI URI:
>
>DID IT STOP IMPROVING BECAUSE HE STOPPED DEVELOPING IT OR IS IT THE OTHER WAY
>ROUND AS MOST PEOPLE THINK? HE CONTINUES TO DEVELOP GENIUS FOR HAND HELDS
>THOUGH,
It was my understanding that the handheld version was basically the Roma
program, which is a little over a decade old.
> WHERE THERE ISN'T MUCH HEAT ! GENIUS 3 WAS THE BEST BUT HE SUBSEQUENTLY
>RELEASED 4, 5. 6 AND 6.5 ALL OF WHICH WERE SAME OR SLIGHTLY WEAKER THAN G3
I think he got sick of it. And you are forgetting that Genius 3 was pretty much
a super evolved version of his Mephisto Amsterdam program. When you look at it
that way, you realize that he was number one for a decade and not a version or
two. As to being weaker, I don't think they were weaker.
>
>REBEL 8 WAS NUMBER 1 AND ALL SUBSEQUENT REBELS HAVE REMAINED ROUGHLY SAME
>STRENGHTH.
Pffff...... I don't think the latest Rebel is roughly the same strength as Rebel
8. Not even by a small margin. You are assuming that the only yardstick is the
SSDF which tests how well programs do against other programs. Rebel was
developed and has been developing to present the best opponent for humans. Doing
a pretty good job too.
>
>MCP MAY NOT HAVE BEEN 1 BUT BUT IT WAS 2 AND IT FAILED TO IMPROVE AFTER NUMBER
>6.IT HAS NOW CEASED PRODUCTION.
I don't know about number 8, but MC Pro 7.1 was better than number 6. No doubt
in my mind about it.
>
>NIMZO 7.32 AND 8 ARE EXACTLY EQUAL. CHESSMASTER 5000 ONWARDS I GUESS ARE
>BASICALLY UNIMPROVED.
>
>FRTIZ HAS CERTAINLY NOT REACHED A WALL AND MAY CONTINUE AT NUMBER 1 FOR A LONG
>TIME BUT LIKE ALL OTHERS WILL EVENTUALLY REACH A WALL-MY PREDICTION.
Eventually? That's awfully vague. That's like saying we all die one day.
>
>ONE GOOD THING I LIKE ABOUT HIARCS IS THAT IT RELEASES NEWER VERSIONS ONLY IF
>THEY ARE STRONGER THAN THE PLDER ONES UNLIKE SOME OF THE ''TOP PROGRAMMERS'' WHO
>KEEP RELEASING THE SAME OLD WINE IN NEW BOTTLES EVERY YEAR.
Who do you have in mind? Nimzo 8 isn't clearly better than Nimzo 7, and that's
quite possible. CM needs more testing as CM8000 is the first new version since
CM6000 and neither has been tested sufficiently to support your accusation.
Other than that, specifics would be nice.
Albert
>
>RAJEN
>>
>>Mchesspro or Rebel were not number 1 in computer chess.
>>
>>After Genius came Fritz and it is basically number 1 for years.
>>
>>Chessmaster,Hiarcs and tiger were in the top of the ssdf list for a short time
>>but most of the time Fritz is number 1.
>>
>>Fritz is often number 1 with a significant difference in rating relative to
>>other programs when other programs became number 1 with only a small difference.
>>
>>I also expect Deep fritz to fight for the top with the tigers.
>>The tigers are better at blitz but based on results that I saw in the rebel site
>>the difference is smaller at longer time control and it is not clear if the
>>tigers are better than Deep Fritz at tournament time control(we need
>>to wait for the ssdf)
>>
>>I do not see the wall of Fritz and I expect it to continue to get better,
>>>
>>>wonder who's going to displace chris theron in 2-3 yrs time?
>>
>>
>>I wonder if Tiger is going to be better than Fritz.
>>Even if tiger is better than Deep Fritz on one processor tiger is newer than
>>Fritz so it has the advantage.
>>
>>Tiger12 also topped the ssdf list because the version that was tested at that
>>time was Fritz5.32 and not Fritz6a.
>>
>>Program that is newer by some monthes have the advantage so even if Tiger is
>>better than Deep Fritz it does not mean that tiger is better than Fritz engine
>>of 2.4.2001
>>
>>Uri
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.