Author: Uri Blass
Date: 08:14:45 06/17/03
Go up one level in this thread
On June 17, 2003 at 11:00:02, georges alain wrote: >On June 17, 2003 at 04:53:07, Uri Blass wrote: > >>On June 17, 2003 at 04:33:29, georges alain wrote: >> >>>On June 17, 2003 at 04:27:13, Uri Blass wrote: >>> >>>>On June 17, 2003 at 01:12:05, Jouni Uski wrote: >>>> >>>>>Finally Michael has put new version of this NPS beast with tablebase support to >>>>>download! But where are all test results? I have so far run it through all my >>>>>test suites. And honestly results are not so stunning: In endgame test it scores >>>>>mere 32/43, when most other engines score 38-41!? And in tactical suite it can't >>>>>match Fritz, Shredder or Ruffian despite stunning node speed. May be it's now >>>>>optimised for playing? >>>>> >>>>>Jouni >>>> >>>>I think that more nodes per second may be a disadvantage. >>>> >>>>It is better to be slower in nodes per second thanks to rules of searching the >>>>right nodes. >>>> >>>>Uri >>> >>>Junior 8 do not confirm this point >>>Philippe >> >>Junior8 proves nothing. >>I know no evidence that is the best program. >> >>Shredder is leading the ssdf list and Junior8 is not better than shredder7.04 >>based on the results that I read(I read that in direct match between Junior and >>shredder Junior did well but shredder does better against other programs). >> >>Uri > > >Shredder better than Junior ? , well junior in on stage for 2 or 3 weeks only >so.. and even if you are right and if Shredder is better the difference is too >small to be a conclusive proof >Philippe I agree that the difference between Junior8 and shredder7.04 is small but it is hard to learn from comparison between different programs what is better. If we look in the same program there are cases when programs became better thanks to searching less nodes per seconds. Fritz8 search less nodes per seconds than old Fritz. I believe that a lot of productive changes that are not of the kind of doing the same thing faster means searching less nodes per second. There may be also changes of doing the same thing faster but I believe that the changes of having a better algorithm are going to dominate. 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.