Author: stuart taylor
Date: 11:40:16 02/28/05
Go up one level in this thread
On February 28, 2005 at 11:47:32, Uri Blass wrote: >On February 28, 2005 at 11:27:45, stuart taylor wrote: > >>On February 28, 2005 at 10:56:42, Uri Blass wrote: >> >>>On February 28, 2005 at 10:40:30, Peter Skinner wrote: >>> >>>>On February 28, 2005 at 05:41:23, stuart taylor wrote: >>>> >>>>>On February 28, 2005 at 04:11:01, Tony Hedlund wrote: >>>>> >>>>>After being roundly beaten by Shredder 7.04, in every test, Junior 9 has now >>>>>shown it's worth, and has slapped today's Shredder roundly in the face, >>>>>repeatedly! >>>>>S.Taylor >>>> >>>>I _could_ still lose the match by a slight margin, but no where near what >>>>happened with Shredder 7.04. >>>> >>>>I still think 7.04 is the strongest version of Shredder out. We will have to see >>>>how Shredder 9 does. >>>> >>>>Peter >>> >>>Shredder9 is leading 15-11 against Shredder7.04 in another match. >>>You cannot decide which version of shredder is the stronger based on one match. >>> >>But Shredder 9 must have new weaknesses somewhere, which Junior attacks. > >I am not sure. >It is possible that Shredder9 could also beat Junior9 even 30-10 in case of >having luck in the first game. What does the first game cause? (that junior doesn't have a weak game of Shredder's to repeat over and over again? Why? is this what has been been happening the whole time? every time the colors are the same, junior does the same thing to Shredder?) > >The games are not independent events and Junior9 may get 20-20 or lose 30-10 >against the same program dependent on luck. > >>Anyway, I'm not interested in which engine is stronger, on average. I want to >>know that te engine I use, represents the best moves I will get from computer >>programs, up till 2005. > >In that case use no engine because for every 2 programs that you use there are >positions that A play better than B and there are positions when B plays better >than A. > > And that is NOT the case with Shredder 9. >>(I can understand an occasional better move with another program, in one or two >>positions which might come up after many games, but not that you should be able >>to feel comfortable that Shredder 7.04 when playing Junior 9, has PROBABLY >>chosen the better move [than Shredder 9 has, when given the same position to >>analyze] ). > >I am not sure if shredder7.04 got the same positions that shredder9 got(I did >not analyze the games) but even if it did it does not prove that shredder7.04 >played better. > >getting better result is no proof of playing better. > >Suppose that position A is a draw by perpetual check but if you do not search >deep enough you may blunder and get a lost position but the opponent can win >only if the opponent see deep enough. > >Suppose >1)Shredder7.04 blundered and got a lost position(first mistake) >2)Junior9 did the second mistake and lost instead of winning. > >Suppose >Shredder9 chose to draw by perpetual check. > >In that case >Shredder7.04 got better result but shredder9 played the better move. > >Uri I don't think these are the common cases over here. And anyway, it has already been seen more than once than Shredder 7.04 did great against Junior 9, and more than once that Shr. 9.0 hardly saved itself against Junior 9. I mean, tested by different people and different specifications (I don't remember it all, but I've seen a few such results which were recently posted on ccc). The chances seem very unlikely to me, that all this is only chance.
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.