I have been reading Larry's posts with great interest. So if his tests are for real (I believe they are), then maybe there should be no more Rybks release until 2010 so that others may catch up. :)
I personally think it'll take the others until 2011 to catch R3
well what you say here has me thinking , if rybka truely is that much better then maybe a match rybka 3 vs hiarcs 12 at pawn odds . at least you dont have to pay hiarcs to play .rybka on superfast hardware 8 core vs hiarcs 12 8 core pawn odds ? who is your money on ? if pawn odds are too much then draw odds is ok .
at least you dont have to pay hiarcs to play .rybka on superfast hardware 8 core vs hiarcs 12 8 core pawn odds ? who is your money on ?My money would be on Hiarcs. I tried this with against Chessmaster 11 and it was very difficult for Rybka 232a SP.
Uri, I also doubt that anyone in this forum has an IQ of 180+, but you never know. There are a few people in this world with super IQs.
I also agree that people can increase their iq with training, but more than 20 points? - highly unlikely. The tests that are used to pin down IQs in the 180 range are so difficult that no training can help at all.
I also agree that people can increase their iq with training, but more than 20 points? - highly unlikely. The tests that are used to pin down IQs in the 180 range are so difficult that no training can help at all.
My IQ is obviously not high enough to understand what this post is doing in this thread. :)
I guess the this is due to the server change (damn, I'm so clever!) :)
Oh, this could be pretty bad if some of the threads from the Warrior Room appear for the unwashed, knuckle-dragging masses of hamsters to view?
i am as thick as pigshit mine was only 132 many years ago and is probably about 50 now after following this forum for last 12 months.
Maybe 160 to 180 is hard but my opinion is that for the average person or above average but not something special(IQ 100-130) it is
not impossible to improve by more than 20 elo.
Uri
not impossible to improve by more than 20 elo.
Uri
I may try something like this, Rybka 3 vs. other top programs on my octal at pawn odds. I think though that the traditional "pawn and move" (f7) handicap is a bit much for the best competing programs, and there would be the problem of variety. So maybe the best way for such a computer match is 16 games, one with each of the sixteen pawns (covering both colors). Since colors would be alternated, this would be "pawn odds" in the truest sense. I'll give it a try, initially at blitz, later perhaps at a more serious time control.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill
