Author: Ricardo Gibert
Date: 18:40:53 12/21/03
Go up one level in this thread
On December 21, 2003 at 21:35:51, Ricardo Gibert wrote: >On December 21, 2003 at 11:21:41, José Carlos wrote: > >>On December 21, 2003 at 09:48:12, Uri Blass wrote: >> >>>On December 21, 2003 at 09:36:27, José Carlos wrote: >>> >>>>On December 21, 2003 at 08:33:22, Duncan Roberts wrote: >>>> >>>>>up to how many moves from the start of the game do we know DEFINITELY, there is >>>>>no material win for white. >>>>> >>>>> >>>>>I once saw that a program called bringer or something like that had checked the >>>>>first 15 moves or so. >>>>> >>>>>does anybody have more information? >>>>> >>>>> >>>>>Duncan roberts >>>> >>>> It was SOS, IIRC. It tried a material only eval from starting position and saw >>>>that 1.h3 (I'm not sure about the move) does not lose material for white in >>>>about 30 plies. >>>> For more details search the CCC archives. >>>> >>>> José C. >>> >>>How do you define material? >> >> I don't define anything, but if I remember correctly, SOS experiment was about >>removing all positional stuff from eval, and using only a constant value for >>each piece. Then use the rest of the program normally, so a mate is still a >>mate. > > >So trading a Queen for a pawn is an even trade? Oh! I see now! The different pieces do not use the same constant value. I misinterpreted what you wrote at first. I thought you meant that a units value did not vary with its type so that Q=R=B=N=P. Of course that would be silly. Never mind. > > >> I think (I'm not sure if my memory is correct) that Rudolf just set the >>initial position and let SOS only material search for some hours. At iteration >>~30 the program was saying 1.h3 score: 0.00. >> That's all I recall. >> >> José C. >> >> >>>Does 1.f3 e5 2.g4 Qh4# lose material for white. >>>If you only count material the answer is negative and if you consider the fact >>>that capturing the king is not allowed in chess then white never has less >>>material in that line. >>> >>>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.