Author: Uri Blass
Date: 00:22:06 12/21/01
Go up one level in this thread
On December 21, 2001 at 01:31:46, Will Singleton wrote: >On December 21, 2001 at 00:57:21, Uri Blass wrote: > >>On December 20, 2001 at 21:14:53, Christophe Theron wrote: >> >>>On December 20, 2001 at 19:54:48, Thomas Mayer wrote: >>> >>>>Hi Christophe, >>>> >>>>> 2) My program as well as other commercial >>>>> engines are using search techniques >>>>> that have never been published. >>>> >>>>well, I think that is the main reason why they are stronger (among others, >>>>better testing, more time etc.) and will stay stronger... >>>> >>>>But I think new techniques could be also discovered by Amateurs - not so long >>>>ago you was also Amateur - and I am totally sure that you have VERY special >>>>search techniques and pruning methods - else I do not understand the depths the >>>>Tigers are reaching... It's really amazing to watch... :) >>>> >>>>>> Perhaps some are using forward pruning techniques that are not >>>>>> published anywhere. The degree to which this affects their playing strength >> is debatable. >>>> >>>>>No it's not. It makes commercial programs clearly stronger. >>>> >>>>In my opinion, "correct" prunning is they best way to get stronger and >>>>stronger... As you said - nullmove can't be everything... believe me, I am >>>>searching, maybe I will find someday something... And I am totally sure all the >>>>others are also searching... futility pruning for example is an idea, but I >>>>don't use it myself because it don't seems to be a win for my engine... and >>>>chess is complicate enough that there could be many many many ..... many more >>>>ideas... prunning is always risky, but if you find a good and robust way, it is >>>>overall a win... and I believe, YOU have found something and the others for sure >>>>also... As long as chess programs evaluate 99,999% totally idiotic positions >>>>there is a way to prune something away... The question is: HOW... :) >>> >>> >>> >>>It's not that a have a special technique. It's that I have many of small >>>techniques, each adding value to the engine. >>> >>>There is no "big thing" to find I think. But there are many small improvements >>>to find. In the end you it makes a big difference. >>> >>>But there is no miracle. It represents years of work. >>> >>> >>> >>> Christophe >> >>I think that the time is dependent in the programmer and it is possible that a >>programmer with more talent can do it in a few months. >> >>I agree that there are many things to find but I also tend to believe that there >>is a big thing to find. >> >>Uri > >The big thing to find is faster hardware. One can freeze his current prog, and >in a few years it will beat every human on the planet. > >Will No I mean only to software improvement I also do not think that beating every human in the planet is the final target and most of the buyers of new engines already lose against the programs of today. I think that there will be always buyers for an engine that is 50 elo stronger than the previous engine in comp-comp games. 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.