Author: Robert Hyatt
Date: 07:50:06 06/28/98
Go up one level in this thread
On June 28, 1998 at 05:34:15, Vincent Diepeveen wrote: > >On June 27, 1998 at 12:28:16, Bruce Moreland wrote: > >> >>On June 27, 1998 at 09:53:00, Vincent Diepeveen wrote: >> >>>No theoretical bullshit, it's clear that all evidence shows how little >>>knowledge DB has, now it's time to show the audience why it's >>>so hard for low rated people to program chessknowledge. >> >>On your web page you once said that you thought that the average chess >>programmer's chess knowledge was about 1400. >> >>I think you are off by about 600 points, I think I am typical at about 2000. > >Account of Bruce on the ICC server: >Blitz rating:1696 > needs x games to reopen account :[8] > won games 154 > lost games 110 > drew 19 > total games 283 > highest ever :1813 reached at (10-Aug-1995) > >This year you only played 2 games. > >That's about 3 years ago when you had 1813 and you are rated >now at1696 at blitz and it has been some time when you last played over >there, and this considering you're quite quick with mouse compared >to most, which matters with a standard deviation of >around 300 rating points at ICC server. > >>But I have also studied chess as it relates to computers. Whether or not this >>has helped my over the board play, I do not know, but I am quite happy >>discussing rams, duos, dispersion, distortion, and yes, majorities. > >Your computerchess insight is not doubted by anyone. > >>Hsu is weaker than 2000 as far as I know, but he didn't write DB's eval code, >>Murray Campbell did, in consultation with players stronger than Murray, and as >>far as I know, Murray is pretty strong. > >So what's his USCF/FIDE rating? > Murray is about 2300 USCF. I don't know his official rating at the present, but he's been a master since I first met him around 1978 or so. >Pretty strong is everyone to you Bruce. I greatly doubt that 2000. > >If i were your chesstrainer i'd advice to play some games now in order >to proof that. If you don't play you lose your ability to estimate stuff. > >I know plenty of people who play chess, and they always overexaggerate >their strength. Note that i still did not start with the USCF versus European >rating >discussion. > >>>I doubt that Hsu has ever heart of the word 'pawn majority'. >>>I dropped some day this word among some chessprogrammers, >>>and they all didn't even know what the word means. Because being >>>the programmer he's the one who needs to implement so he needs to >>>be the one that must exactly understand what it is and what it is not. >> >>The stronger ones probably do. > >>But even if they don't, it really doesn't matter *that* much, I think. I don't >>have pawn majority code in mine. It is pretty complicated to write, and even >>when I hash it it takes a lot of time to execute. > >>I get burned from not having it, I am sure. I often have problems with >>opponents having distant passers, but the cases where it is a majority that >>causes problems by itself aren't that common, it seems, or at least they get >>covered by other things. > >>bruce > >What i mean is: at a certain time it might get your weakest link. > >I played Ferret only a few times on the internet, and i didn't win a single >game. > >This was because of 3 reasons from which 2 are major > - a) my own laughable tactical ability at blitz when it's passed midnight > - b) the agressiveness of ferret > - c) the reasonably well book you're having considering the old > variations i play. > >The strong points of ferret > a) it's tactical ability. I don't know a program tactical as strong as >ferret. > but compared to my own play and GM play *any* program is tactical > strong enough. However > b) the depth Ferret gets at your alpha in blitz is really *amazing*. > c) this is really cool. Just like few of the strongest commercial programs > i'm the most amazed in how well Ferret sets up an openingsposition > when still getting few ply. > d) it's ending for a computer program is quite well. > >The weak points of Ferret > a) it knows quite little from advancing. After it has this lead because of > some good moves it has problems to advance, meaning it doesn't > know exactly what it must exchange. > b) it relies a *lot* on its tactics. Like most programs (mine included of >course) it > plays things i think are not needed, so simply risking things because >it > relies on tactics. > >b) is the main reason why programs that outsearch others mainly win from >their opponent. I've seen Lonnie play plenty of Nimzo98 against Ferret, >and those games are really weird. No doubt both programs are best >blitz players in the world, but there will be a time that other programs >get 11/12 ply too in rapid (wrongly called 'standard' at icc) chess. > >Now pawn majority is just an example of something which is needed to >advance, and i agree with you that it doesn't apply to many positions, >but it's something i think that a program needs. > >I just took this, because Bob said he has the code ready to use, and >would like to show how *hard* it is to program knowledge into a program, >and the huge number of NOR's and NAND's needed for it. I don't undestand "huge". This is not a hard thing to recognize. Certainly not as hard as computing sin(x) for any x, as the pentium FPU can do. And it leaves plenty of transistors to do hundreds of other equally interesting things (hint: count the instructions in the CPU, the FPU and then add in the MMX instructions, and then remember *all* this fits on one chip.) > >In my program pawn majority is 500 lines. Meaning that it's simple code in >this sense that a simple human definition caused me a lot of trouble >implementing. > >I *doubt* whether a program that doesn't even know the pattern >pawn g6,g7,f7 open h-file is not a bad doubled pawn, has something like this >code. > >I doubt that any program getting 80% lazy evaluation is among the positional >best programs. > >Greetings, >Vincent
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.