Author: Uri Blass
Date: 12:43:00 12/18/02
Go up one level in this thread
On December 18, 2002 at 15:19:54, Bruce Moreland wrote: >On December 18, 2002 at 11:22:48, Omid David Tabibi wrote: > >>>I have not tested Omid's variant. >> >>There are two types of people: those who when see an idea, implement it, and >>only then criticize it; and those who criticize first and only then try it. I >>previously thought that only Vincent is in the latter group... > >This is completely unfair, and you are turning this into something personal. > >I take issue with some of the elements of your paper, and perhaps with the >larger issue of how people do testing in computer chess academic papers. That >is all. > >If your intent was to show that VR=3 is better than R=2, in your own program, >you have shown that, but VR=3 is a variant of R=3, and you needed to investigate >the relationship between R=3 as well as between R=2 and VR=3. > >You considered it axiomatic that R=2 is better than R=3, but your own data >strongly implies that R=3 is better than R=2. > >None of this has anything to do with whether or not your idea is good. Your >contention is that I should have to prove that your algorithm is bad before I >criticize your article. That is not how things work. I could stipulate that >your algorithm is good, and this would not affect the substance of my criticism. > If you write a paper, the burden of proof is on *you* to show that your >algorithm is good. This is true whether or not your algorithm is actually good. > >Meaning, that just because your algorithm is good does not exempt you from >having to prove it, and just because someone else comes along later and proves >it doesn't mean that your original paper was done properly. > >bruce I agree I think that the important question is not if the paper was done properly(a lot of papers were not done properly). The important question is if the idea is a good idea. Only Vincent claims that it has to be a bad idea. other people say that it works for them or do not work for them or work for them with some modification but do not say that it is a general bad idea. I believe that it may be productive with some rules that use verification search only in part of the cases. 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.