Author: James T. Walker
Date: 07:50:16 05/15/02
Go up one level in this thread
On May 15, 2002 at 09:35:12, Jorge Pichard wrote:
>On May 15, 2002 at 08:00:09, James T. Walker wrote:
>
>>On May 15, 2002 at 01:41:49, Harald Faber wrote:
>>
>>>On May 14, 2002 at 20:03:56, James T. Walker wrote:
>>>
>>>>Because of the quick win by Fritz I suspect there is something wrong with this
>>>>opening.
>>>>
>>>>[Event "Blitz:5'/Athlon 1.4G/512M"]
>>>
>>>
>>>Wow, THAT is a level...
>>>
>>>
>>>>[Site "MyTown"]
>>>>[Date "2002.05.14"]
>>>>[Round "4"]
>>>>[White "Fritz 7"]
>>>>[Black "Hiarcs 8"]
>>>>[Result "1-0"]
>>>>[ECO "B65"]
>>>>[PlyCount "47"]
>>>>[EventDate "2002.05.14"]
>>>>
>>>>{64MB, Hiarcs8.ctg, Athlon 1.4G/512M
>>>>} 1. e4 {0} 1... c5 {0} 2. Nf3 {0} 2... d6
>>>>{0} 3. d4 {(b3) 0} 3... cxd4 {0} 4. Nxd4 {0} 4... Nf6 {0} 5. Nc3 {0} 5... Nc6 {
>>>>0} 6. Bg5 {(Bc4) 0} 6... e6 {0} 7. Qd2 {0} 7... Be7 {0} 8. O-O-O {0} 8... O-O {
>>>>0} 9. f4 {0} 9... Nxd4 {0} 10. Qxd4 {0} 10... Qa5 {0} 11. Kb1 {(Bc4) 0} 11...
>>>>h6 {0} 12. h4 {0} 12... hxg5 {0} 13. hxg5 {0} 13... Ng4 {0} 14. Be2 {(Rh4) 0}
>>>>14... e5 {-0.62/9 20} 15. Nd5 {(Qd3) 0} 15... Bd8 {0.39/10 82} 16. Qd3 {0}
>>>>16... Nf2 {0.39/10 8} 17. Qg3 {0} 17... exf4 {4.45/9 62} 18. Qxf2 {(Qh4) 3}
>>>>18... Bxg5 {1.81/8 3} 19. g3 {(Rh5) 1} 19... f5 {3.74/8 23} 20. gxf4 {0} 20...
>>>>fxe4 {5.64/9 30} 21. Bc4 {(Qg2) 0} 21... Be6 {6.51/9 22} 22. Qg2 {0} 22... Rf5
>>>>{7.94/8 9} 23. Qxe4 {2} 23... Re8 {16.37/8 11} 24. fxg5 {Resigns 3} 1-0
>>>
>>>
>>>Again I do not get the point. You want to prove an opening line wrong by taking
>>>this BLITZ game into account???
>>>
>>>Sometimes...it is really helpful to take a longer time control. For example in
>>>the line you posted above. On my slow (!) 600MHz Athlon Hiarcs has
>>>
>>>14...e5 with -0.62/9/41s, -0.37/10/1min06, +0.52/10/2min07 and changes to
>>>14...f7-f5 after 5min17 (0.00/10), stays there at 0.00/11/9min56 and plays that
>>>move at TOURNAMENT timce control after 11min10.
>>>
>>>I don't know whether this saves the day, but it is definitely of other quality
>>>than your g/5... I call it SHIT and not test. Such games have no value of
>>>anything concerning CHESS. It is indeed counting games and numbers to see
>>>rankings in 5min games, but it has NOTHING to do with CHESS.
>>>
>>>If the base of your disappointing results and assumed crappy opening book are
>>>such games, well there is no need to say anything more, this speaks for itself.
>>>
>>>And it leaves a bitter taste of what you, Chris and some others try to do here.
>>
>>Hello Harald,
>>I never said the Hiarcs book was bad! I only tried to offer it as one of the
>>possibilities for my results. I offered this because the Hiarcs 7.32 book had
>>problems. In trying to find out if Hiarcs had problems with it's own book I
>>searched all the games looking for games in which it came out of book with a bad
>>score. The game above is one game I found. There are a few more. I realize
>>that generally the programs and books are "tuned" to win at 40/2 for SSDF
>>purposes but these programs are expected to play "Blitz" also.
>>Yes I'm counting games and numbers period. I believe that no matter what the
>>time control used that the results are to be considered as part of the overall
>>performance. I don't believe you can take one position and say "Look, my
>>program solves this faster than any other program and therfore it is great". I
>>also don't believe that you can find one time control where a program performs
>>well and say that this is the only time control you can play against this
>>program. As for my disappointing results, they are not contrived or made up. I
>>am searching desperately for a reason. It is possible I have found the reason
>>but I will not know for several days/weeks. I have to run my test all over
>>again. I am an avid Hiarcs 7.32 fan. I was very happy with Hiarcs 7.32. If
>>you have seen my posted results then you should be able to understand my
>>disappointment with Hiarcs 8. I have seen you post good results here and so I
>>understand why you are pleased with Hiarcs 8. I have asked for help/solutions
>>from anyone with any ideas. Some people simply want to attack my methods or my
>>hardware. They don't seem to realize that the opponents of Hiarcs 8 are running
>>on the exact same machines which are sometimes switched.
>>Jim
>
>
>Jim I didn't meant to attack your hardware, sorry if you took it that way, since
>you and the the rest of us who have Hiarcs 8 are trying to desperately find how
>Hiarcs 8 perform at different time control. I appreciate what you are doing
>testing at different blitz time control. I believe that certain programs simply
>perform better at slower time control than other, for instance Fritz 7 and Chess
>Tiger are very well known to play good at all time control. But Hiarcs 8 with
>Nimzo8.ctg has perform well for me at longer time control.
>
>PS: I also noticed that with time control shorter than G/60 with Hashtable of 64
>MB Hiarcs 8 performs much better than with bigger hashtable, which I reserve for
>Longer time control greater than G/90.
>
>Pichard.
Hello Jorge,
You are exactly right. Some of us are actually trying to figure out WHY I'm
getting bad results and I appreciate any and all help. I have also played
Hiarcs 8 at Blitz with 64 Meg hash tables. I'm not sure that is the best but
I'll try different settings to see if it makes any difference. I have run most
of my G/60 games with 128M for Hiarcs 8. Maybe 64 Meg is best for that time
also. I was very happy to get good results at G/5 vs Shredder 6. I hope I get
more good results.
Regards,
Jim
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.