Author: Jouni Uski
Date: 05:49:25 02/03/04
Go up one level in this thread
On February 03, 2004 at 08:25:35, William Penn wrote: >On February 03, 2004 at 01:25:17, Jouni Uski wrote: > >>On February 02, 2004 at 23:23:06, William Penn wrote: >> >>>On February 02, 2004 at 11:25:35, Chessfun wrote: >>> >>>>On February 02, 2004 at 11:04:03, Dmitri wrote: >>>> >>>>>Final result: >>>>> >>>>>Shredder 8 65.5/100 >>>>>Fritz 8 34.5/100 >>>>> >>>>>Rating differance of +105 elo for Shredder 8. >>>>> >>>>>Conditions of the test: >>>>>P4 2.4Ghz, 1Gb ram, 3-4-5 tb on, learning on, books own, book learn on, 128Mb >>>>>hash, 60 min for the game all other param. set to default. >>>>> >>>>>Comments: >>>>>Shredder seems to like closed positions more the Fritz and they are very much >>>>>equal in open types of positions. Shredder is better at creating space and >>>>>opening files then Fritz. The only weakness I can spot is it when it comes to >>>>>the endgame phase. Fritz is a lot better in the endgame. Just try the position >>>>>from the last game of human vs computer in corr. chess from www.chessfriend.com. >>>>>Shredder does not like Qxb7 with perp check where as Fritz sees that it is the >>>>>only way. >>>>> >>>>>I will not post games here so not to use up space, but you can email me for >>>>>them. >>>>> >>>>>With regards, >>>>>Dmitri >>>> >>>>This Shredder 8 is indeed clearly a very strong program currently I'm running it >>>>at 40'40 and have the following scores after 48/150 >>>> >>>>v Shredder 7.04 +3-0=1 >>>>v Deep Fritz 8 +2-0=2 >>>>v Fritz 8 +3-1=0 >>>>v CM9K +1-0=3 >>>>v Tiger 14 +1-0=3 >>>>v Junior 7 +2-1=1 >>>>v Junior 8 +2-0=1 >>>>v Crafty 1906 +1-0=2 >>>>v Fritz 7 +2-1=0 >>>>v Ruff 2.0 +1-0=2 >>>>v Rebel 1201 +2-0=1 >>>>v Deep Sjeng1.5 +1-0=2 >>>>v Hiarcs 9 +2-0=1 >>>>v Hiarcs Bar +3-0=0 >>>> >>>>Sarah. >>> >>>Thanks to both for the good information. I haven't run any scientific tests but >>>have spent several hundred hours analyzing a wide variety of positions with >>>Shredder 8. I am noticing (subjective) what appears to be a strong horizon >>>effect with Shredder 8 in infinite analysis mode, which is the only way I use >>>it. Under those conditions the moves don't seem particularly "intelligent" >>>unless it is let go for several hours, a minimum of 4-5 billion nodes >>>(positions), then it does indeed probably find the best possible move in most >>>positions - except it doesn't seem very smart in endgames. I get the feeling >>>that humans of average strength could probably beat it easily in some endgames. >>>That's with an XP 2400+ computer at 2.0GHz and 1GB RAM (512MB hash), using the >>>native CB Shredder 8 GUI on Windows XP Home. >>>WP >> >>I disagree! S8 is just excellent in endgames and reaches stunning depths = >>easily 20 - 30 ply in endgames. >> >>Jouni > >Is this a stunning endgame analysis? >5+ hours hours runtime (05:35:47) in infinite analysis mode, all 3-4-5 piece >tablebases, XP 2400+ at 2.0GHz, 1GB RAM, 512MB hash, native CB GUI with all >default settings > >[D] 3r2k1/5pp1/2RP3p/pp1p4/8/4P3/PP4PP/6K1 w - - 0 1 > >Analysis by Shredder 8: > >[snip early lines] > >28.Kf2 f5 29.Ke1 a4 30.Ke2 Kf7 31.g4 fxg4 32.Rc4 dxc4 33.b3 axb3 34.Kd2 bxa2 >35.d7 a1Q 36.h3 gxh3 37.Ke2 > ² (0.67) Depth: 30/57 05:35:47 7919282kN, tb=472425 > >This is purely "giveaway chess"! >WP Shredder has tendency to show anything nonsense in main line, but it don't affect palying strength at all! Only first move is important. Jouni
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.