Author: Uri Blass
Date: 06:53:48 01/27/03
Go up one level in this thread
On January 27, 2003 at 09:36:03, Rolf Tueschen wrote: >On January 27, 2003 at 09:25:38, Uri Blass wrote: > >>On January 27, 2003 at 09:04:02, Rolf Tueschen wrote: >> >>>On January 27, 2003 at 02:18:36, Bruce Moreland wrote: >>> >>>>DJ lost. Aside from this, neither it nor its team has anything to be ashamed >>>>of. Chess is a zero-sum game. >>> >>>Objection! It is a business, Bruce. Of course the team and poor DJ are not to be >>>blamed, but the spin doctors who lied about the strength of today's progs, they >>>are to be blamed. That's why people are so deceived. How was Bob Hyatt treated >>>when he said that the advantage of DB2 was still a trump card? It's the CC >>>people themselves who on the one side have the knowledge but on the other side >>>drifted into daydreaming wishful thinking. >> >>I think that you forget that Kasparov could not train against deeper blue when >>he could train against Junior and play against the specific weaknesses of >>Junior. >> >>Kasparov knew based on his experience that Junior does not play well in the kind >>of position that he played for(unless the programmers changed it in 6 months not >>to have the weaknesses). >> >>He could not do it against deeper blue and this is the reason that deeper blue >>played better relative to Junior's first game. >> >>The opinion that deeper blue is not better than the commercial of today is based >>on analysis of the games and finding no impressive move of Deeper blue that >>programs of today need hours to find(it is not a proof that deeper blue was not >>better but the point is that I found no evidence that deeper blue was better). >> >>Deeper blue stopped to play chess so it is impossible to prove something about >>it and we do not know the speed of deeper blue. >> >>The claim that it was 100 times faster than the programs of today is not >>something that we can take as a clear fact. >> >>Nodes are defined different by different programmers and we simply have no data >>about Deeper blue so it is better to stop to discuss about deeper blue. >>> >>> >>>> Sometimes you lose. Losses are rarely pretty. >>>>When computers lose, it's even more likely that you'll see something ugly, >>>>because a computer will make the strongest move it can, not the strongest move >>>>that is not embarassing. >>> >>>You misunderstand the chess of this game. Sorry. But as I said often enough, the >>>playing down a book line without real understanding is a scandal to CC itself. >> >>I do not think that it is scandal. >> >>It is possible also to claim that it is a scandal that humans can buy programs >>and learn their weaknesses when they cannot do it against human opponents. >> >>These are the rules of the game and I see no problem with them. >> >>Uri > >Uri, > >Your arguments on DB2 are well thought and not in my league so I skip them. But >your comments about the actual game are a bit strange. > >First of all I do not think that Kasparov trained for 6 months now. That would >also be nonsense because, as I wrote, the GM know that a little change in >personalities would result in a different play. The same with a new book. So I >cannot accept your statements here. To know that in this line it's VERY >important to know all the analyses is crucial in GM chess. There are calm lines >and hot lines. This one is hot a) ooo vs oo and b) g5. It's all in the strengths >of progs. It's very tactical. But the key here in our debate is the influence of >books. I think the nonsense of books has been proven. I would throw then in the >trash. > >Rolf Tueschen I do not think that kasparov planned all the way to win the game but I believe that he knew from his experience that Junior often does mistakes in the type of position that happened in the game. Kasparov could not get this training against deeper blue. I also do not expect the use of books to help Junior. The Junior team has some options: 1)Using a big book and in that case the book will have mistakes and kasparov may get an advantage 2)Using a small book when every line was checked manually by a team of GM's and computers In that case kasparov may have an advantage thanks to the fact that he will know more moves in book. 3)Using a small book to play in 0 seconds and a big database to use after the moves of the small book when the big database is used for extensions(for example a program may extend by 1/2 ply every move that was played in the past and there may be other ideas to try) but unfortunately I know no program that use database when it is out of book in that way. Uri
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.