Author: Ed Schröder
Date: 14:45:31 08/15/00
Go up one level in this thread
On August 15, 2000 at 15:06:58, Peter Skinner wrote: >On August 15, 2000 at 11:17:02, Vincent Diepeveen wrote: > >>On August 15, 2000 at 05:42:26, Uri Blass wrote: >> >>>On August 15, 2000 at 05:28:47, Marcus Kaestner wrote: >>> >>>>On August 15, 2000 at 01:15:31, Jouni Uski wrote: >>>> >>>>>Nice, that all TOP programs are there. With 4/15 programs >>>>>(Fritz,Junior,Nimzo,SOS) probability for Chessbase to win is quite high! >>>>>Can anybody count it? And specially when all these are very strong... >>>>> >>>>>Jouni >>>> >>>>no way. you will see it. >>>> >>>>marcus >>> >>>I do not know the future. >>>Chessbase can win and can lose. >>> >>>I am sure shredder,Rebel and tiger are better but the same is also for some >>>chessbase engines. >>> >>>You also cannot be sure that the best program is going to win. >>>In WMCCC of 1997 Junior won convincingly and Junior was not the best program. >>> >>>How do you know before the games that chessbase is not going to win? >>> >>>Uri >> >>I don't understand why are so many excited about tiger at CCC? >> >>I only see it play positional incredible bad, missing all kind of >>knowledge. It's really anti-computer style. Nowadays programs have a lot >>less problems with that. >> >>I see crafty even at a dual getting only half speed at which it >>will run at WMCC blowing away tiger in a positional way at the >>icc regurarly. >> >>Some games Tiger sees lands and sails off to hit the land real soon >>thereby playing well, others it doesn't see the land in the >>see and sails to unknown destiny and loses pathetic. >> >>To win a championship like this the first priority is a program that >>doesn't lose at all. >> >>I'm not impressed by Shredder's insight in the middlegame. >>Especially putting pieces into pins and other dangerous anti-positional >>things is something Shredder happily does, only its deep search and the >>fact that at these 13+ ply depths there are not many tricks left, >>allow usually such moves in a tactical way. >>Yet the bottom line is: shredder hardly LOSES a game. >>Therefore it always has a good chance to win a championship. >>Same is valid for junior. It just doesn't blow things quickly. It's missing >>some fundamental knowledge, but no matter whether it makes mistakes, >>it doesn't blow it soon. It is not bad in endgame either, one of the >>big reasons it won 1997 title. Shredder is real good in endgame. >> >>Tiger is in a few endgames not bad, but in the average endgame it loses >>a lot in very weird ways at the icc servers. That's basically blitz of >>course, so no big deal usual, but the slower games show the same problems >>also. >> >>Crafty might get killed real soon in its book against Kure book who had >>the chance to study automatically generated books real well as he's >>a lot on the internet. Against certain other programs crafty will also >>lose bigtime because of book. >> >>So i don't see crafty win either, but it will play in a reliable way >>of course. Very good tested. >> >>If you have a big weak point in your engine, you can shake it winning >>a heavy champ like this. >> >>Finally considering tiger and rebel will have nearly the same >>book, i definitely would be on the last if my choice was limited >>to those 2, at least rebel knows something about sacraficing pawns... >> >>Vincent > >I have never seen Vincent impressed with anyhting but his own engine, so I take >this with a grain of salt. One thing I know for sure, Diep won't win. I will bet >millions on that fact. Maybe Vincent is impressed with the "knowledge" that he >has incorporated into Diep, but every other program trashes in on ICS, so.... > >I am impressed with the advancements other programs have made. And I find it >just as impressive that Shredder doesn't lose many games. > >Just a side note.. how many games did Diep lose at last years WCCC? > >3 of 7, and finished with 3.0/7.0... impressive... > >How did Shredder finish? Oh that is right 1st... he must be doing something >wrong. Maybe the following is true and in particular for Vincent: finding (some) weak spots in particular engines sometimes makes you forget about the (more) strong points. Ed
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.