Author: Uri Blass
Date: 06:00:30 10/28/04
Go up one level in this thread
On October 28, 2004 at 01:26:22, Daniel Shawul wrote: >On October 27, 2004 at 08:39:22, Uri Blass wrote: > >>On October 27, 2004 at 07:45:35, Daniel Shawul wrote: >> >>>On October 27, 2004 at 03:49:38, Jorge Pichard wrote: >>> >>>> >>>> >>>>Rank Engine Score Ba Da S-B >>>>1 DanChess 10.5/15 · ·· ·· ·· ·· ·· ·· ·· ·· 11=11=100101=11 42.75 >>>>2 Baron 1.5.0 4.5/15 00=00=011010=00 · ·· ·· ·· ·· ·· ·· ·· ·· 42.75 >>>> >>>> >>>>15 of 20 games played >>>>Tournament start: 2004.10.24, 22:52:30 >>>>Latest update: 2004.10.27, 07:13:03 >>>>Site/ Country: S0025128408, United States >>>>Level: Blitz 20/1 >>>>Hardware: AMD Athlon(tm) Processor 1202 MHz >>>>Operating system: Microsoft Windows 2000 Build 2195 >>>>PGN-File: C:\Program Files\Arena\Arena.pgn >>>>Table created with: Arena 1.092 >>>Hi Jorge >>>thanks for the games. >>>I don't think DanChess is as strong as the result implies. >>>May be it is better than Baron in blitz. DanChess itself is weak in blitz. >>>I am sure the difference will vanish at longer time control. >>>best >>>daniel >> >>I do not understand. >>If Danchess is weak in blitz then how can it get good results in blitz and not >>at longer time control? >> > >Where did you see that? DanChess won a blitz match and You replied "I am sure the difference will vanish at longer time control." It simply contradicted your claim that DanChess is weak at blitz. >AFAIK DanChess gets good results in longer time controls,which is expected >because it is slow and more knoledge based. I think that better order of moves and better branching factors and better pruning tricks is the main factor for being relatively better at long time control. I do not think that knowledge based program should have problems at blitz and Hiarcs is maybe the strongest blitzer when it is also slow searcher. I know that Ufim was considered to be slow searcher and it was relatively better at blitz because at long time control it was outsearched. AEGT and WBEC are good examples. My >tester Gerhard send me results of blitz games at 40/5 which indicate danchess >1.07 is not stronger than the previous version 1.06 in blitz. >But in other tournaments at long time controls it seem to improve. So i presume >it is weak in blitz. > >> >>Maybe you mean that Baron is weak in blitz or that Danchess is relatively weaker >>at longer time control. > >No. DanChess is infact stronger at longer time controls. > >About Baron Richard knows what it is good at. But my impression ,from what i >read here and from what i see of Baron 1.4.0 search, is it is more Knowledge >based and is better at longer time controls like DanChess. > >Daniel I agree that Richard probably worked more on the evaluation and less on the search but it does not mean that the baron is better at long time control. better branching factor or better order of moves can do program relatively stronger at longer time control. 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.