Author: Anon
Date: 06:00:37 09/22/99
Go up one level in this thread
On September 21, 1999 at 19:42:45, Wayne Lowrance wrote: >Hello >First my comments are not to suggest my H performs better than yours but that it >plays very much different. >my computer is pentium 11 400 mhz and set H up to 64 meg hash time control at >40/1 hr. >One significantv difference, I use Fritz 5.32 General Book because I quickly >found out that H book is inferior ! None the less I carefully followed your >moves move by move. up to move 22 at which point I stopped the comparison as it >was very clear that your posted moves where very inferior to the moves that my H >was making. here is a summary. > >at: >8. My H wanted to 0-0, forced your move ! >9. My H wanted Be3, forced your move not none the less H score difference >between your H and Mine was very similiar. >18. My H wanted Bxh6 (1.73/8 your move on My H dropped score to (+1.06) >significant evaluation difference for just one move ! and your H seems to go >down hill from here on based on what evaluation by my H. >19 My H wanted h5 forced your move >20. My H wanted Rh-g4 forced your move >21. My H wanted Qc5 forced your move. by this time My H degraded indicated the >score had degraded to -0.52 where as my H was indicating 2.21 at move 20 , quite >a difference in play. There is a difference when playing the game at set time controls and allowing the computer to select it's move than there is in watch mode. Example: 18. My H when set at same time controls and playing the game over exactly as odell played it will play 18. h4 (1.67/8 72) However in watch mode looking at 4 or 5 moves seems to prefer Bxh6. 19. This was one move I never expected but at time control H played it changing it's eval at the very last second 19. g4 (1.07/9 211). 20. Again no problem here gxf5 (1.90/8 196) In watch mode seems that even when it reaches the same depth it can produce a different best move from what it played. I would like to know if anyone can repeat 21. h5 or 23. Kc1 in either watch mode or play mode at the depth indicated. > >I find this very interesting cause there seems to be a difference in how Hiarc's >plays based on posts and comments that I follow here. I dont have a answer, is >it hardware differences ? i dont's know. I am positive of one statement that can >be made, I do not like the Hiarcs default book based on games played against >other programs on a well known chess server. > >I would like to voice an impression concerning the play of Hiarc's verses time >control. Hiarc's plays at or near the the top in Blitz 5 min. Its superiority in >play against other programs on chess server drops as the time element is >increased. The reason may be that Hiarcs positional understanding is the plus >advantage when game moves are made rapid but the advantage diminishes at longer >time controls against deep fast searchers such as Fritz etc. In game at 30 min >Fritz5.32 and Hiarchs are very very close it seems to me > >I may be wrong in both statements, but that is my experience with the program >against other programs. >
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.