Author: Vincent Diepeveen
Date: 06:06:07 11/07/01
Go up one level in this thread
On November 07, 2001 at 01:27:01, Uri Blass wrote: Some have commercial interests. If important tournaments get played on the internet then you get advanced players where Super-GMs always win. Comment from a guy called jfernandez when playing in the world championship qualification for the worldchampionship chess in moscow: "comon i'm better than a bunch of old kibitzing GMs who lose loads of time because of slow and unclear communication". Who is better than a bunch of old kibitzing GMs? Right Svidler at a laptop. The first 8 already got $6000 or so entry fee when just *arriving* in moscow. So the first 8 who qualified were all Super-GM helped. I'm not sure whether Kramnik or Kasparov helped anyone, but for sure the rest of the world top was helping some team. I counted them all number 3..15 from the world top perhaps with 1 or 2 exceptions, helped some kind of team. Now winning a world championship computerchess isn't worth $6000, it's way more worth. Remember what happened with the auto232 protocol when chessbase modified it. Remember the hardware SSDF friendly requires you to ship to them, otherwise they can't garantuee you get on the list. Remember the Braingames computer-human way of decision taking. Chessbase or Chessbase. Now this is not anti-chessbase meant. That $1 million price fund, it's peanuts compared to the sales chessbase will get when playing kramnik, even having a good excuse why they lost: "kramnik could prepare". This though hell sure he doesn't even have a dual computer, only khalifman has an old dual computer. Also fritz7mmx is non-preprocessor, deepfritz still looks preprocessor to me, nevertheless they play nearly everywhere the same move. What i mean to say is that commercial interests in computerchess are limited to only a few persons, and their interests are big, because people either kick on world titles or they kick on massive PR. >I think that the best solution is to have no operators when the game can be >played on ICC. So let's get realistic please, chessmaster sold 5 million copies or god knows how much. Even if all the others sold so far only a total of 1 million copies, then we still talk about massive money here. >On November 07, 2001 at 00:11:17, Robert Hyatt wrote: >>On November 06, 2001 at 12:10:10, Thomas Mayer wrote: >>>Hi Vincent, >>>oh, you change the facts always the way you want to have them... >>>>First of all. I did not disallow to adjust clock. I did ask him >>>>whether he was insane when he said he wanted to restart his program. >>>You disallow it, because he can only do it with restart. The only way he can do >>>it - simple fact. >>> >>>>Also i have played for years in textmode, and now this guy doesn't. >>> >>>is there a rule where it is stated HOW you have to adjust time ? >> >>Actually there is. The rule used to say "the operator can adjust the time >>by entering the current chess clock time left, _if_ the program requests this >>information." > >I think that the best solution is to have no operators when the game can be >played on ICC. >If people want to see the game on a real board then it is possible to replay the >moves also in the real board. Parallel programs are non-deterministic, so no way to get the same moves. If i analyze with tiger i can for example *never* get the same moves. Preparing against tiger must be hell, sometimes it plans h6?, then you restart engine and then it doesn't even consider h6? Then it considers a6? Then i restart put it at tournament level and it plays a better move Nc6. Idem with deepfritz. It doubts like hell everywhere. DIEP also doubts a lot, but usually i can reproduce moves. Nevertheless there is no way i can *garantuee* this. I have some crap extensions (which i consider to throw outside it again) which are cool to solve some tricks in testsets, but they are of course alfabeta dependant, something which means that you're completely non-deterministic in principle! >If programmers want to meet and discuss about the games there is no problem >about it and they can replay the move on a real board but I do not see a reason >to force them also to operate the program during the game. > >I want to see a tournament of programs and not a tournament that is also of >operators. And i want a tournametn with 40 in 2 level, not some stupid blitz level where my 800Mhz machine gets 8 ply. >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.