Author: Sandro Necchi
Date: 00:48:29 11/01/03
Go up one level in this thread
On October 31, 2003 at 20:46:18, Mike S. wrote: >On October 31, 2003 at 19:53:05, Harald Faber wrote: > >>On October 31, 2003 at 00:03:05, Mike S. wrote: >>>(...) >>>Would you call it FAIR when a major competitor of yours wouldn't be tested at >>>all?? Hi, I think you are making confusion. Other people can do it, but not SSDF. Even if they will look the outcome: 1. the program get 2800 rating and get no.2 Some people could say, well with his book it would have been stronger, so it is no.1 Some other could say, well the learning by CB GUY has give at least 20 points and the good book another 30 points at least... Other may say something intermediate and so on. So, what is the rating of CM9K as really is...I mean the one you get when you buy it? It will be still unknown and different opinions will be just opinions... The next question arise... Why we do not test other chess programs with different books than their book? I would add than I am thinking to design a F1 car...if I have enough funs I can go to Ferrari and tell them that they have to give me their engine and Schumi as well because my funs want to see how the car I designed will perform. Does this make sense to you? Sandro >> >>When the competitor does not the least to be able to be tested, well, for me >>this looks as if the competitor does not want to be tested at all. > >It doesn't matter if the competitor (it's programmer and/or publisher) want's to >be tested or not. I'd even say it must not matter. Test are not only for them, >but for the fans too, even mainly for them who are or may be customers. You can >compare this with other tests of all kinds of products, cars, cameras, washing >machines. An independent testing organisation shouldn't be dependant :-)) from >decisions of people related to the product. I think this is obvious. So they >don't have to wait for any permissions or efforts to have a program tested. > >(I've already written once or twice, that if think the common SSDF practise to >have the progs sent to them - for free obviously - is somewhat doubtful, >considering their independance.) > >Also, the qualitiy of the SSDF ratings would be in question when important >engines are missing. Some years ago there was a similar situation, and it was >just courtesy of the SSDF not to test some important programs because the >publisher said he didn't want it, but the users certainly didn't appreciate the >decision. > >When publishers and progammers want a testing lab which they can *control* >completely, what and how to test etc., than they would have to create *their >own* testing organisation. I guess we share the opinion that this isn't >realistic, because some companies would have invest money to get results, they >are not happy with (and most probably can estimate before anyway). > >As for the usage of general.ctg, I would rather have used a simple >database-generated tree as a replacement. But this would probably have it's >downsides too. >http://www.talkchess.com/forums/1/message.html?325009 >I guess the King fans would call that unfair and say, why don't you use >general.ctg like already done in other similar cases... > >It's a compromise, and therefore, not perfect. > >mfg. >Michael Scheidl
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.