Author: Rolf Tueschen
Date: 10:43:36 05/25/02
Go up one level in this thread
On May 24, 2002 at 09:05:59, José Carlos wrote: > Ok, two things to discuss then: Let me paste the answer on the Dutch National, there it was DEEP FRITZ who played in the official championships. Now, some players gave up after a couple of moves, some let the game end in a defensless 17 moves mate attack and one wasn't playing at all, he asked draw after move 1, what was rejected, then he gave up. Reason for all this was the rule that you could not forfeit in advance because then you had lost the right to participate. Then there was a sensation in the game between Tiviakov and the machine where Tiviakov thought that a draw offer by Morsch disturbed him in his concentration. He thought Morsch should have resigned. All that you could read in the archive files here. Just enter Morsch, Tiviakov and rules. > > 1. Why using an opening book is not important. > > To discuss this, we need to get into computer science. A basic program is a >function. A function receives data, processes it and provide data. In our case, >a chess program can be viewed as a function that receives some data (the board >position, time left, game history...) and gives a result (a move in a certain >time). What the function does inside concerns only to the programmer. That is your view. But it's not true. As it was shown certain Crafty clones had to be analysed and "opened", if you like. So, I see no problem if you must prove that you do not use GM books... >When one >of my clients needs a program to print invoices, the only thing that matters >(for him) is that he provides some input to the program and the program prints >an invoice. He doesn't care about how the program is done. He just care about >the program does what it is meant to do. That's all. There might be some >hardware restrictions. Ok, if my client can only use a PII-300, my program must >be small and fast, and I can't do fancy graphical effects. Just an example. > From the point of view of the programmer, the most important thing is that the >program does what the user expects. If you are writing a chess program for very >weak players who mostly care about the GUI, then you focus on a pretty GUI, and >don't care so much about strength. If you want your program to top SSDF and win >tournaments against humans, you focus on strength. But all of that is only the >programmer's bussines. From outside, you don't see data structures nor search >algorithmics nor heuristics nor preprocessed tables... You see a program that >plays damn strong. But it should have neither Crafty nor GM books inside. :) > Finally, don't forget that every single thing inside every single program on >earth is human knowledge. There's no difference. If a table lookup is cheating, >a sum is cheating, a loop is cheating, etc. It's all the same thing. It is like >a novel, with a programmer instead of a writer. The programmer writes line after >line, everything coming from his head. The machine only does, damn fast of >course, what the programmer wrote there. We have now a moment of highest interest. What you are saying with not the least self-doubts is for me the highest and most trivial form of fallacy. BTW decades after the debates in Atom physics. The simple rejection to your simple statement is this, no, it is not all allowed what a programmer is doing. And you know yourself where the clue is in your argumentation. You said, if SSDF, if tournaments, then strength. Yes, but fairness? That is the most important factor in sports. If you like it or not. It's the British who brought it into the World. :) So, could you please modernize your argumentation included these points? > > 2. How to convince FIDE to allow programs. > > I'm not an expert in this regard. FIDE was created for humans. Human >tournamens have a set of rules that are valid only for humans. If we start with >the fact that the computer can't move the pieces itself on the board and push >the clock button, everything else is not necessary. > FIDE has a different set of rules for blind players. They can touch the >pieces, and use a different board. I've played against blind players some times, >and I got nervous seeing him moving his hands on the board an touching the >pieces when I was trying to think. But I accept that. Why? Well, I respect blind >people, and understand they need different rules. > If we try to extrapolate this to computer players, it is much more difficult >to say "I respect the computer and understand it needs different rules". I do >respect the programmer, but then let the programmer play, not the computer. Yes, I agree. You found two rather uninteresting comparisons. The most important is however, what a computer player should be. What parts etc. > Bob points some interesting reasons why computers are not allowed in FIDE >tournaments. I believe it is difficult to avoid it. I, having written a chess >program, wouldn't like to play Fritz in an official tournament. > So for me, it is ok that programs don't play in official tournaments. We have >ICC and and GM challenges. That's more than enough for me. Moreover, in the next >years the computers will be too strong to make playing against humans not >interesting. > Just my opinion. > Yes, it is absolutely fine by me. But I see many fallacies in the typical GM shows against computers. Just like DB2 in 1997. The questions of fairness most of the time are not even discussed. But these days are gone, where it was simply the moral number one, anything goes. Because at the time it was so funny to watch the machines play chess at all. But if the machine and their lovers start to claim mastership and titles, we should be allowed to take a closer look at the principal parts of the machine plus program. FIDE rules come into play. Not the technological side of how moving the pieces, but the question of cheating during the game. Look please, if I were Bob, I would also try to neglect the whole cheating debate, because we were talking about my whole past life in computerchess. It's not so easy to accept that most of the past practice was in truth not in accordance to the FIDE rules. However the guys had other things to do than discuss justice in times when the original machine stood thousands of miles from the tournament chamber... It would have disturbed their concentration. The end of this naivety was IMO the failure of Feng Hsu to come into talkings with Kasparov after 1997 event. And Kasparov let his representative tell him that at first DB should qualify itself in the usual events of tournament chess. In that moment the classical computerchess had come to its end. Simply because little show events are diferent to tournament chess under FIDE rules. For many Kasparov's lesson is still too big stuff to swallow even today. (Even in show events human masters are no longer willing to knowingly accept to be outplayed, see the Kramnik vs. Fritz event.) Rolf Tueschen > José C.
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.