Author: Enrique Irazoqui
Date: 11:30:25 09/19/99
Go up one level in this thread
On September 19, 1999 at 13:42:52, Ed Schröder wrote: >>>It's not so simple, consider a few points... >>> >>>#1. Instead of the move send weird stuff to the other PC as a result >>>the other PC will crash. Do that in case your score is below -1.xx. >>>Don't do it in every game. >> >>You and I checked several programs with CB and Donninger's auto232, and no >>program sent anything weird. >> >>>#2. Let your own program crash when you are down in score. Don't do >>>it every game. >> >>This can be "achieved" also manually, and it has happened twice in the few >>thousand games I autoplayed. >> >>>#3. Send the "move now" command to the other PC after say 10 seconds >>>in a 60/60 or 40/120 game. Hide it a little, nobody will notice. >> >>This can be "achieved" also manually, and it has happened twice in the few >>thousand games I autoplayed. > >These days crashes are rare I agree but I consider a time-out also as a >crash and I have seen too many of suspect time-out cases and they by >definition always favor the Rebel opponent. I checked the time-out parameter >and that can't be the reason. Similar games (behavior) seen? I noticed very, very few time-outs. Not enough in any case to build a pattern or to question autoplayers. >>>I have not the impression it currently happens but is all possible >>>if a programmer wants so. >>> >>>About books... >>> >>>You can easily recognize when the opponent is out of book simply by >>>checking the opponent response time. With this information you can >>>recognize the opponent. Think about this for a while. I have tried >>>it for my own curiosity and it simply works. Now you can do nice >>>things in case you know the opponent. Is it happening already? I >>>don't know but it can be done and quite easily. >> >>I wonder if this is so easy (the 2 versions I tried didn't make reliable >>identifications at all), and even if a program can identify the opponent >>one can easily argue that: > >Read again. I am not talking about first versions I once send you but >a system based on the moment when a program is out of book which >you can easily measure by the (long cq direct) response time of the >opponent. You can identify the opponent based on that information. >It's a piece of cake if you think about it for a few moments. > >>- it would be an intelligent development in the "artificial intelligence" >>field. >> >>- all programmers could do it, so what's the problem? > >The problem (book-learning alike) is that it hides the real strength of a >chess engine. These days it's not about the engine but more about >the extra elo you can gain by smart (and aggressive) book-learning >and opponent recognition will only make it worse. It operates hidden >hardly to see (notice) for the end-user unless you take a very deep >look in the system. It is like with learners. Once upon a time, it gave advantage to some programs. Not now. Programs either they have a learner or they escape with broad books. Opponent recognition is, as far as I can tell, something no program has. But if they all develop it, and this would be a genuine advance in intelligence, the engine will still have the last word. We are saying the same as 2 years ago about learners. >>>This whole auto232 thing is so fragile that I can imagine people >>>don't want to touch it any longer. >> >>It has always been fragile, but more reliable than the very few manual games >>that can be played. Proof: you and I play thousands of automatic games, and >>seldom any manual ones. Why is that? :) > >It depends on the intention you are playing these thousands of games. For >me that is to improve Rebel as it gives me a lot of useful data. It is useful as the only way to get enough comp-comp games to make accurate quantifiable comparisons, in spite of the few problems here and there. Although I learn much more from one game I play manually than from 10 autoplayed games, but this is another question. Enrique >Ed
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.