Author: Hristo
Date: 04:49:53 07/27/03
Go up one level in this thread
On July 27, 2003 at 07:23:00, George Tsavdaris wrote: >On July 27, 2003 at 07:01:22, Milorad Madjar wrote: > >>On July 27, 2003 at 06:34:04, Uri Blass wrote: >> >>>On July 27, 2003 at 06:27:48, Milorad Madjar wrote: >>> >>>>On July 27, 2003 at 06:18:52, George Tsavdaris wrote: >>>> >>>>>On July 27, 2003 at 05:56:23, Uri Blass wrote: >>>>> >>>>>>On July 27, 2003 at 05:20:37, Eduard Nemeth wrote: >>>>>> >>>>>>>Junior 8 can’t mates without TBs with B+N !!!!! >>>>>> >>>>>>No surprise. >>>>>>A lot of programmers do not consider that knowledge as important when tablebases >>>>>>can solve the problem. >> >>>>Of course, I think that is very stupid if one the best chess programs (or >>>>Kramnik, Kasparov, Fisher,Capablanka, ....) don't know mates without TBs with >>>>B+N than he must see or read to Tbs or different book for mates!!!!! >>>>Regards >>>>Milan >>> >>>I think that trying to compare top programs with humans is a stupid comparison. >>>Programs are supposed to solve some problems and the way that they solve it is >>>unimportant. >> >>I think it is veri important.Because If you go on this way, today or tomorrow >>you can have very big Tbs and very big opening book and two chess programs Will >>be play 20-30 >>moves with very big opening book and have 10 moves in the middlegame and that go >>to very big Tbs. >> I put to you one question : where is the chess ??? If the game have 60 >>moves in this game you have only 10 moves of really chess !!!! > >What is "really chess"? In an endgame for example: "really chess" is >non perfect moves by humans or perfect moves by engines? >And what is chess? A game which might be as silly as tic-tac-toe. Because it's >much more complicated, we play some moves and say "i have a good plan" >"this move is wrong"... but i don't think this is really chess. >Chess for me is the algorithm that will prove that chess is a "first/second >player wins/or draw game". This algorithm already exists! It is called "brute force" and with a few shortcuts and extra logic this basic approach delivers very good results even without solving the chess game. It is a matter of time, perhaps lots of time, before the algorithm will yield the final answer, so meanwhile all those engines are taking shortcuts and whenever possible play perfect chess. If you are looking for an algorithm that solves chess in some other way, non brute force inspired method, then perhaps you should consider the issue of how do you verify this "It is win/draw." outcome of "your" algorithm without using brute force or some other shortcut. :) As it stands, even the bellow-mid-range 1.5 GHz computers can easily defeat %99.9 of human population and %90 of all chess players. Most programers, being of the pragmatic type, will concentrate on what works and perhaps will make it a bit better over time. > >You don't understand that chess engines have created to play as perfect chess >they can. And as tablebases make their play better, they had to be used. >And if one day, someone find a way to have 32 piece tablebases or >one algorithm to play perfect chess, then we would have only one line which >engines would play all the time. Very true. There is no ambiguity or uncertainty about the fact that one day we will know the outcome of chess. Consider this: You need compression when storing the table-bases ... The engines of today are evolving in the direction of becoming part of the compression mechanism for the time when the engines will have to provide/generate 10-20 moves on the fly for the large 32 pieces table-base. Regards, Hristo >> >>> >>>If humans are allowed to use book during games then I do not find it as stupid >>>if they do not learn the relevant books before the games. >>>programs are allowed to use tablebases in all the serious tournaments so >>>I do not find it as stupid if they do not have the relevant knowledge. >>> >> >>This comparison is not stupid, because until the game continue man can't use >>book to see best move , but chess programes until the game continue see to Tbs >>(book) >>and we have two Tbs (book) how play game not.I don't want that chess ! >> >>>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.