Author: Uri Blass
Date: 22:12:59 06/22/04
Go up one level in this thread
On June 22, 2004 at 18:01:37, Rolf Tueschen wrote: >On June 22, 2004 at 17:53:10, Franz Hagra wrote: > >>>For using this test suite “WM-Test” and to avoid >>>misunderstandings/misinterpretations of my test results I recommend to look at >>>the attached “readme.txt”. >> >>Only missing point is a little about the significancy of the figures! >> >>Hagra > > >I think that they think that the support from CSS is significance enough. This >is the same with the false public relations by ChessBase when they quote SSDF >results and rankings as if that had any specific value. This is here in the same >tradition. Hoax, nothing but hoax. > >If now someone feels unhappy let me explain that test results without >significance, without validity, without reliability - - that then they are >worthless. That's so simple and easy to understand. > >Finally as a footnote, it is extremely false to appear here in this forum, >dropping such nonsense results without any value and then quickly chicken out. The results may be not important for you or for me but they are not nonsense results without any value. They have a value for the people who are interested in them. They may be interesting for people who analyze specific type of positions when sacrifices work(engines sometimes play wrong sacrifice and if the target was to get some estimate for the analytical playing strength of the engines then they could give also positions when the target is to avoid some wrong sacrifices that part of the engines play). Of course it is only an estimate and it is possible that you get misleading results in specific positions when the program lose the right solution and later find it again but it takes more than 20 minutes to lose the right solution if your hardware is not fast enough. We can hope that it does not happen in a lot of cases and every engine has only small number of positions when it happens. 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.