Author: Mark Young
Date: 14:20:09 05/30/99
Go up one level in this thread
On May 30, 1999 at 16:55:42, Melvin S. Schwartz wrote: > >On May 30, 1999 at 16:27:57, Mark Young wrote: > >>On May 30, 1999 at 16:04:43, Melvin S. Schwartz wrote: >> >>> >>>On May 30, 1999 at 14:36:35, Mark Young wrote: >>> >>>>On May 30, 1999 at 14:23:19, José de Jesús García Ruvalcaba wrote: >>>> >>>>>Hi Mel, >>>>> here it goes a fragment of a post by Dr. Hyatt you are constantly referring in >>>>>the thread about SSDF testing: >>>>> >>>>>---------------------------------------------- >>>>>On May 27, 1999 at 15:51:59, Prakash Das wrote: >>>>> >>>>>>Well, I still don't understand the point of this "world championship" then. Sure >>>>>>, you can call it "hey, if nothing it will be fun", a "test of systems", etc. >>>>>>But what are exactly is this exercise trying to prove? If program A on hardware >>>>>>B, beats program D on hardware E - does that say much about A compared to B? >>>>>>This belies the principles of science - you have to have a uniform platform >>>>>>for all participants to make any kind of judgement. >>>>> >>>>> >>>>>I don't understand what you don't understand. This is an "open" competition. >>>>>Anything is allowed. Any sort of hardware and software combination that can >>>>>play chess. It has _always_ been that way. It will always be that way. The >>>>>question being asked is "what is the strongest electronic chessplayer on the >>>>>planet?" Not "what is the strongest program?" Or "Who is best on equal >>>>>hardware?" or anything else... >>>>> >>>>>---------------------------------------------- >>>>> >>>>> As you can see, you were attributing to Dr. Hyatt words by Prakash Das. >>>>>José. >>>> >>>>As Homer Simpson would say in Mels place....DOH!! >>> >>>My dear fellow, >>> >>>You seem to enjoy starting an argument and pursuing it to no end. You also have >>>some character flaw that makes you want to be insulting. >> >>Because you seemed to be so confused about SSDF, Testing, and the Rating system. >>It is no wonder you are also confused about this. If you check the thread. You >>are the one that posted to me. I addressed my post to Hans Christian Lykke. >> >>> >>>Now, I will address your rather crude comment - no doubt from a rather rude >>>individual. >> >>Sorry, but you are the one floating a flase Quote from Dr. Hyatt and giving this >>as proof that we are wrong and Dr. Hyatt agrees with you. Sorry that is funny. >>And as I said before I don't think Dr. Hyatt would agree with you... >> >>> >>>The fact I used the wrong author of my quote does not change my position. If you >>>or Dr. Hyatt feel it is not important to test programs on equal software - have >>>a nice day! >> >>I always have a nice day... > >Take a look at the post by Timothy J. Frolick on 5/29. -------------- You are correct. It is not always possible to have two identical computers and the same time. To be "perfectly scientific" we should decrease the number of variables. Since we can't always do that we resort to statistics. You will note that Hiarcs will win a certain percentage of the time. Now if Hiarcs were winning half of the games it would be at least 80 points stronger than Nimzo because of the 2.3 fold increase in processor speed. So the point is not the winning of the games but rather the percentage won. I assure you that the person doing the testing does not have a desire to skew the results. Don't have hard feelings over these posts. Knowledge is a cumulative endeavor and tomorrows computers and programs will make mincemeat of all of todays machines. We have to keep testing in order to improve the things. Tim Frohlick -------------------- Look, you still don't get the point here. I agree with the above post, it isn't "fair" to pit one program against another with one of them on a lesser machine and then when the one on the better machine wins to claim it is a better program. However, that has nothing to do with the SSDF and their methods. This is not match play, it is a method for determining a rating. These two things are not the same. The SSDF takes a program and puts it on multiple types of computers and establishes a rating for the program on that paticular hardware. For instance, program "A" will be put on a P90 and on a P200. Both of these will be considered a seperate entity. Both will establish its own seperate rating. Therefore over time the SSDF will establish a stable rating for program "A" on a P90, and a stable rating for program "A" on a P200. This then gives the SSDF a good benchmark by which to establish the rating of program "B" which may also be tested on a P90 and a P200, or as is done now only on a P200. In the end the SSDF winds up with a lot of programs with stable ratings that can be used to benchmark even more programs and so on and so on. There is nothing here that is at all unfair. Each program establishes its own rating for a particular hardware. Now I feel like I'm teaching kindergarten here, so I hope you finally get it. Mark
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.