Author: Ralph Patriquin
Date: 10:38:50 06/10/03
Go up one level in this thread
On June 10, 2003 at 06:41:17, Jorge Pichard wrote: >On June 10, 2003 at 06:31:05, Jorge Pichard wrote: > >>On June 10, 2003 at 01:06:43, Ralph Patriquin wrote: >> >>>On June 09, 2003 at 03:25:29, Jorge Pichard wrote: >>> >>>>http://talkchess.com/forums/1/message.html?299865 >>>> >>>>http://talkchess.com/forums/1/message.html?299919 >>> >>> >>>Have you looked at all the games? They seem to be a little short on average >>>compared to the typical computer game. I haven't looked at all of them, but in >>>particluar game 3 doesn't look right. Shredder marches his king right out into a >>>mating net which is easy for a computer to see. My Shredder Paderborn simply >>>will not play those moves. Maybe a problem with your engine-engine match setup? >>> >>>Ralph >> >>I noticed the same same specially game three, the only explanation that I have >>for this is that Fritz8.ctg is too deep and since both programs are using >>Fritz8.ctg or opening book if they follow the cook line it is hard to walk out >>of a mating net, any other program like fritz 8 can avoid getting out of the >>mating net, but only to continue with an inferior game. The setting that i used >>is both used the same fritz8.ctg, both programs 64 MB Hash and pondenring off. >> >>PS: I also set the same position of game 3 using Shredder 6 to see if the move >>chosen was different but using the same time control shredder 6 also walked into >>a mating net. I don't see any reason to continue an inferior game out of the >>opening if the endgame will give you the same result. But yes this was something >>that caught my attention too. >> >>Pichard. > >You are welcome to match shredder 6 against SmarThink using the same time >control to see if shredder 6 performs better than Shredder Paderborn. I was >going to do that, but i decided to match SmarThink against Fritz 8 version 8.008 >under the same condiction and SmarThink is playing a la par after 4 games so >far, I will post them later. If I increase the size of the Hash to 128 it might >affect SmarThink, but i will keep testing it under the same condiction for >comparison reason. I beileve that later on Fritz 8 will take the lead being much >stronger than shredder Paderborn or Hiarcs. > >Pichard. Well, I ran some experiments. These on P4 2.26 GHz 256MB RAM 40'/40+40'/40+40': Shredder 6.02 32MB Hash 21... Rd8 22.Rxd8+ 2:07 Kxd8 23.Qd3+ 3:05 Shredder 6.02 64MB Hash 21... Rd8 22.Rxd8+ 2:58 Kxd8 23.Qd3+ 2:11 Shredder 6.02 128MB Hash 21... Rd8 22.Rxd8+ 2:54 Kxd8 23.Qd3+ 3:23 I also forced the game continuation: Shredder 6.02 32MB Hash 22.Kd2 Qa5+ 23.Kc1 2:58 Shredder 6.02 64MB Hash 22.Kd2 Qa5+ 23.Kc1 2:15 Shredder 6.02 128MB Hash 22.Kd2 Qa5+ 23.Kc1 5:10 Although 22.Kd2 is not a good move, the horrible blunder is 23.Kd3??. As you can see, I cannot reproduce it. My Shredder score quickl goes to ~ -6.00 after this move. Ralph
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.