Author: Lance Perkins
Date: 17:09:21 07/13/04
Go up one level in this thread
Oh it its all about "decisions". And these decisions are limited by your time, skill and resources. Folks decide not to go in debt just to purchase a quad-proc machine for this event. Folks decide not to spend the time to add parallel-search in their code. They would rather spend time and money on other things. I can decide to go with an inferior book, because I can't afford to pay a GM to design one for me. You know how the Deep-Blue beta lost to a notebook due to a short book? If I can afford to fabricate my own chess ASIC and finish it before the next WCCC, then I could decide to do it. Can everyone afford it and make same decision? I guess you're missing the whole point behind the WCCC. So, you have not answered this question: what to you is the "world 'computer' chess 'champion'. What you want is the: world AMD 2ghz, Linux, Anand-book, 1.2MB binary, chess champion world 3.5K nps world chess champion world 13-ply-limit chess champion etc, etc. --- On July 13, 2004 at 19:23:20, Omid David Tabibi wrote: >On July 13, 2004 at 19:16:26, Lance Perkins wrote: > >>But would you, in your right mind, field in a donkey after hiring an expensive >>and talented jockey? >> >>If folks decide to use an inferior hardware in WCCC, then its their decision. > >It is not about "decision". You assume that Jonny, ParSOS, Falcon, IsiChess, >Deep Sjeng, Woodpusher, Movei, The Crazy Bishop, and FIBChess, all had quad >Opterons and they "decided to use an inferior hardware"? > >The problem is that only a few have access to such fast systems. If you provide >every participant with a quad Opteron, then of course there won't be any >problem, and then they can really "decide" what to use. > > > > >>There's nothing unfair about that. >> >>Consider the history of WCCC. Its about 'computer' chess. Back then, there was a >>variety of hardware, and that was really exciting. These days, it just so >>happened that PC's are cheap and easy to write software for. So, bring in the >>best hardware you can, even if you just bought that instead of putting it >>together yourself. >> >>You don't need WCCC if you just want to see how programs will do on the same >>hardware. You can do that yourself, in addition to the countless others who run >>their own tournaments. >> >>On July 13, 2004 at 18:53:56, Omid David Tabibi wrote: >> >>>On July 13, 2004 at 18:46:36, Lance Perkins wrote: >>> >>>>Nope. A horse race is a competion between different horses "and" different >>>>jockeys. >>> >>>What if some jokies are riding donkies instead of horses? >>> >>> >>> >>>>You see, they don't swap jockeys later and do another round. Jockeys >>>>play a great part - they too are sportsmen (like football players). >>>> >>>>I'm not really sure how you come up with your analogies. Even if you give humans >>>>the same time to think in chess, some humans think faster or better than others, >>>>and make use of time better than others. >>>> >>>>Are you proposing that in WCCC, all chess engines should evaluate the same >>>>number of nodes per move? You see, its not just about the time spent, since some >>>>engines are compiled with better compilers and will therefore process more nodes >>>>per given time. >>>> >>>>On July 13, 2004 at 18:29:53, Omid David Tabibi wrote: >>>> >>>>>On July 13, 2004 at 18:25:34, Lance Perkins wrote: >>>>> >>>>>>Indeed. And that is what you see in horse races. Horses have different skills >>>>>>and run differently depending on the kind of race track. And the jockeys have >>>>>>different skills. >>>>> >>>>>Right, and that competition is by definition a competition between different >>>>>horses. You won't see some jockies riding horses and others riding donkies, as >>>>>is the case in WCCC. >>>>> >>>>> >>>>>> >>>>>>The winner? The best jockey + horse combination. >>>>>> >>>>>>Have you seen anyone go to the Kentucky Derby with a Donkey and complain that >>>>>>its unfair? >>>>>> >>>>>>Can I go the tournament with minimal book and complain later that my oponents >>>>>>have books prepared by GM's. That can't be "computer" chess anymore. That's >>>>>>human GM chess. >>>>> >>>>>In a human tournament, would it be fair to give one participant 2 hours for 60 >>>>>moves, and the other participant 8 hours for the same number of moves? >>>>> >>>>> >>>>>> >>>>>>On July 13, 2004 at 18:08:23, Omid David Tabibi wrote: >>>>>> >>>>>>>On July 13, 2004 at 17:53:20, Fernando Villegas wrote: >>>>>>> >>>>>>>>Tha's fallacy, my friend. Programs that use different hardware ARE NOT >>>>>>>>indifferent programs than are just by chance riding faster horses; they are >>>>>>>>made to work with this or that hardware, SO to use it is part of the >>>>>>>>programming. >>>>>>> >>>>>>>My program is also made to work with parallel hardware. But as I replied to >>>>>>>Dann, running a parallel engine on a single processor machine is not much fun... >>>>>>> >>>>>>>You can be a great horse rider, but when you are given a donkey, you will most >>>>>>>probably lose to a much less skilled competitor riding a horse. >>>>>>> >>>>>>> >>>>>>>>My bst >>>>>>>>fernando
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.