Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: rebel 10~!! super strong on amd k62 500

Author: blass uri

Date: 13:21:32 07/28/00

Go up one level in this thread


On July 28, 2000 at 15:17:26, Ratko V Tomic wrote:

>> The question which program is best may be dependent in the
>> time control. What is the hardware and time control of your games?
>
>I play on either Celeron 400Mhz or another Celeron 466Mhz, 64Mb for hash tables
>(on machines with 128Mb total RAM). Depending on how much time I have, most of
>the time I set programs to 15 or 30 seconds average per move (my time is usually
>similar to program's although I don't pay attention to how long I think on a
>move).

Hiarcs is good at this time control.
I think that the learning bug of Hiarcs becomes a problem in longer time control

I played one game with hiarcs7.32 against Crafty17.11 with time control of 3
hours/40 moves and no permanant brain and Hiarcs lost and analyzing the game
proved that there was a learning bug problem(hiarcs could play better without
learning from previous search).

 I also use infinite time control or analysis mode when I occasionally
>look through a database game. In the first weeks of trying out a program I also
>play a few tournament level games.
>
>As to using programs one version back from their current ones, from games and
>results I have seen, it doesn't seem it would change very much in terms of
>playing pleasure or general impression. I still think a loss to CM, Fritz or
>Junior would leave the usual after-taste of a 'cheap shot'. Over the almost two
>decades, there were only few times where going to the new version or a new
>program brought something really new. The appearance of Super Constelation, AI
>chess (later M-Chess), Rebel 8 and Hiarcs 6/7 are the few cases where after the
>first game I thought 'hey, this is something interesting and new.' The rest of
>upgrades or tries 'new' programs over years didn't bring a sense of something
>essentially new, ahead of the pack.

I know that the difference between Junior6 and Junior5 or Fritz6 and Fritz5.32
is a big difference in ssdf rating.

The difference in other cases(like Rebel8-Rebel9) is usually smaller.

In the case of Junior the programmer almost did not work about search and the
main difference is the evaluation function.

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.