Author: Uri Blass
Date: 18:03:02 12/22/03
Go up one level in this thread
On December 22, 2003 at 20:16:10, David Dory wrote: >On December 22, 2003 at 17:36:13, Uri Blass wrote: > >>On December 22, 2003 at 16:28:50, David Dory wrote: >> >>>On December 22, 2003 at 10:39:20, Uri Blass wrote: >>> >>>>On December 22, 2003 at 09:44:43, David Dory wrote: >>>> >>>>>On December 22, 2003 at 07:08:09, Duncan Roberts wrote: >>>>> >>>>>>Different software engines have different strengths and weaknesses in different >>>>>>types of positions and I once saw mentioned the idea that one could raise the >>>>>>elo level of chess software by 150 points by having some software which would >>>>>>interface with the top 5 programs and would have all of the strengths and none >>>>>>of the weaknesses of each individual program. This would be achieved as the >>>>>>interface program would ask the individual program to only play the type of >>>>>>position it played best at. >>>>>> >>>>>>kasparov once mentioned that in certain positions junior plays at 150 elo points >>>>>>higher than the competition, on the other hand he said fritz is more 'certain'. >>>>>> >>>>>>An interface program should be a far tougher challenge for kasparov to crack. It >>>>>>would truly reflect the best of computer science against the best chess player. >>>>>> >>>>>>I do not know much about computer chess, but I assume that to implement this in >>>>>>at least a basic way should not take a great deal of time. (a week ?) >>>>>> >>>>>>Is this right? and if so (although it is easy to ask) why is nobody doing it.? >>>>>> >>>>>>There must be many good programmers on this site whose chess programs while good >>>>>>cannot realistically hope to reach the 'top 10'. Surely (assuming the top 5 >>>>>>chess program authors co-operate with this) they would be making a much bigger >>>>>>contribution to computer chess by implementing an interface program. >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>>>duncan roberts >>>>> >>>>>For this to work, you'd need to have your top 5 programs all working on separate >>>>>computers, so they wouldn't interfere with each other's computing time. >>>>> >>>>>Then sure enough, you'd get Junior with one best move, and Fritz with another, >>>>>and Shredder with still another best move, while ChessMaster might agree's with >>>>>Fritz's best move. So what move would you, or your interface program, pick? >>>>> >>>>>Your interface program would have to be as smart as Kasparov to tell which move >>>>>really was best (and sometimes Kaspy'd be wrong!). It isn't enough to classify a >>>>>position as "open", "closed", "beginning", "middle", "end", etc. That's WAY too >>>>>broad, and the whole classification might change dramatically in 4 ply or more, >>>>>so now the position being searched would need to be classified again, wouldn't >>>>>it? >>>>> >>>>>Pretty soon your work to classify the position would be eating up the programs >>>>>time to search, if you weren't careful. >>>>> >>>>>All programs do better in open, tactical positions. >>>> >>>>better than who. >>> >>>I believe in the concept of playing the position on the board - not the >>>opponent, whenever possible. >>> >>>So better is not related to "who", but to "which", as in which move has been >>>chosen to play. >>> >>>And all computer programs do better in open or tactical positions where the >>>programs can find something "good" to play, within it's search horizon. >> >>No >>better is relative. >>If you say that they more often find the best move you may be right but it does >>not mean that it needs to prefer open position because the opponent program also >>may have more problems to find the best move. >> >>> >>> >>>>We compare program against programs and not against humans and I also do not >>>>care about comparison against humans because most humans have no chance against >>>>programs and the interesting question is comparison of programs with other >>>>programs. >>>> >>>> >>>> >>>> >>>> They all try to avoid closed >>>>>and blocked positions, simply because the best move is not frequently within >>>>>their search of the game tree. >>>> >>>>No >>>> >>>>Not all program try to avoid closed position. >>> >>>The stronger programs have included evalution terms to favor open, tactical >>>positions over closed one's. >> >>No >> >>Crafty did it but there are commercial programs with symmetric evaluation and I >>believe that Junior does not have evaluation to prefer open positions. >> >> >> Yours will do it also when you get seriously >>>interested in whupping two legged opponents. >> >>I find computer-computer games as more interesting because chess program already >>beat almost all the two legged opponents. >> >>> >>>> >>>>My program simply tries to find the best move based on its evaluation and does >>>>not care about close or open position. >>>>I believe that it is the same for most programs. >>>> >>>>Uri >>> >>>Yes, amateur and/or weaker engines need other improvements first. But given the >>>choice of two moves of equal value, the program should choose the one move that >>>leads to an open and more tactical position. >> >>No >> >>The program should choose the move that give it better chances and against >>computers I see no reason to assume that it is the move that leads to open >>positions. >> >>Uri > >Strong programs, sooner or later, will have to play against humans, and for >*best* play against a human opponent, the program will have to have code to >prefer more open and tactical positions. > >Whether this interests you or not at the moment is immaterial. It's been well >known for more than 30 years. > >Saying: > >>No, >No, >>>No > >Just won't change this fact. > >David It is the choice of the programmer. I see no reason that a program has to play public games against humans. The fact that Fide does not allow them to play in tournaments to get fide rating is also a good reason not to care about games against humans and only care about games against other programs. programs are simply not allowed to play in most of the tournaments that humans play and I see no reason to care about results in games that the program is not allowed to play. Games on ICC are mainly blitz and I see no reason to care about it when it is known that humans are weaker in blitz and that in longer time control humans may cheat by using a computer. Uri 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.