Author: Uri Blass
Date: 17:11:57 10/20/02
Go up one level in this thread
On October 20, 2002 at 19:37:18, Johan Hutting wrote: >On October 20, 2002 at 15:22:41, Uri Blass wrote: > >>The >_released_ >>winboard version of Ant is clearly better >>than the >_released_ >>winboard version of Celes. > >which is over a year old anyway. Ant winboard version is also old. What is your estimate for the level of Celes relative to the winboard engines in Leo's tournament? Is it better than the public Ant? >> >>It lost against the baron and celes with > >so baron is also clearly worse then ant? Baron is clearly better than the winboard Ant but It was not clear for me if it was better than the last Ant. The point is that they repeated the same opening after losing and in both games the queen got from d8 to d5 and d6 and d8. I know that there is an opening 1.e4 d5 2.exd5 Qxd5 3.Nc3 Qd6 but even if this opening is good it seemed from the games that Ant did not know to play it so it is a clearly bad choice for opening book. >Ant only scored 1 point vs Djenghis and a very lucky half point vs ZZZZZ. > >>1.e4 d5 2.exd5 Qxd5 3.Nc3 Qd6 and in both games the queen came later to d8. >> >>I do not understand this opening choice. > >Their opening choice was 'fine' as far as I could tell. Hans Secelle (operator >Ant) mentioned he checked all booklines with fritz and didn't find any >weaknesses. Their search however had some major instabilities. They outsearched >Celes by 6-7 ply on some moves with a too small branching factor. > >Also, after 15...Qa7 their position went downhill (Qd5, Qc5 or Qd7 seems better. >Celes expected Wc5, b4, Qd5 IIRC), Celes had a constant +2.xx after 18.Qg3 while >Ant's score fluctuated between -1.00 and -3.00 every couple of moves. > >Their major problem seemed to be testing new settings, which 'might' be caused >by people not playing with ancient winboard versions. The programmer of Ant was >already attempting to figure out what caused the search inconsitencies during >the game. He'll quite probably show up with a much stronger playing version of >Ant next weekend, at least, I hope so for him. > >Perhaps you should participate in a tournament yourself to figure out the >importance of good books, well tested engines and why to use unreleased >versions. If I participate I participate with unreleased version to prevent preperations but I do not plan to participate with untested version. I do not believe in the importance of books and the success of List with no book convince me that a book is not very important. I believe that something with similiar level to list4.61 that is unknown with no book or a very small book has good chances to get 50%. My latest movei(not public) has book and pondering but my tests are almost always with no book when I always change the first move so I can get different games. Uri
This page took 0.01 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.