Author: Uri Blass
Date: 08:12:38 08/21/02
Go up one level in this thread
On August 21, 2002 at 10:08:51, Manfred Meiler wrote: >On August 21, 2002 at 05:58:25, Uri Blass wrote: > >>On August 21, 2002 at 01:42:07, Mike S. wrote: >> >>>On August 20, 2002 at 17:25:43, Uri Blass wrote: >>> >>>>(...) >>>>No reason to be surprised. >>>>This test was designed to put Fritz on top because the analysis was done by >>>>Fritz to decide about the positions. > >>> >>>What exactly do you mean? >>> >>>Do you really mean, "designed to put Fritz on top"? If not - and I assume (and >>>hope) you didn't mean that - then I suggest to be a bit more careful. Now it >>>sounds like an accusation. > >> >>I mean that the testers used Fritz to analyze >>before deciding about the positions. >> >>Vincent said that when he asked about a position they gave him Fritz's analysis >>as a proof. > > >Mr. Blass, > >how do you know so exactly in which way this test suite was designed ? I admit that I do not know. The only guess that I made is that they did their computer analysis with Fritz before deciding about the positions. I did not mean to say that they planned to put Fritz as number 1. >Where's the proof for your statement above ? >Only the fact that Mr. Diepeveen has got Fritz's analysis as a proof from the >test authors Michael Gurevich and Heinz-Josef Schumacher ? That would be >ridiculous... > >Do you really believe that the authors Mr. Gurevich and Mr. Schumacher worked >almost two years (altogether) in designing this large test suite only to put >Fritz on top ? I did not say that this was their target. >Do you really believe that me myself would invest so much (computer) time - 112 >(versions of) engines x 100 test positions x 20 minutes - in testing and >documentation of my results - only to put Fritz on top ? I said nothing bad about you. > >Why should we do that ? For money or other profits ? >Maybe it's impossible in your eyes but - these test efforts are only a private >hobby of "ordinary" friends of (computer) chess with interest to chess engines >and test suites (like the earlier BS-2830, GS-2930 or BS-2001) - without >professional claim, without any commercial interests. > >I have to state: It obvious makes little (or no) sense to give my EXCEL results >of WM-Test to the public. It's not very funny to see such >insinuations/accusations in consequence. I apologize if you felt bad about it. I will try to develop my test suites when the target is going to be simply to avoid mistakes in games of movei or to find good surprising moves in the same games. It is going to take time because in a lot of cases it is not clear what is the losing mistake but I expect to have convincing analysis for every position in the test suite(I expect most of the positions to be easier than the positions in the WM-test suite but there are going to be some positions that are not easy). I do not like most test suites and one of the reason is that they often do not include convincing analysis to prove that the test is correct. 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.