Author: Heinz van Kempen
Date: 05:55:02 01/26/06
Go up one level in this thread
On January 26, 2006 at 08:24:05, Rolf Tueschen wrote: >Heinz, > >it's pretty difficult to criticise you with your remarkable openess and >sophistication. My point is another one. It is dangerous the more someone so >reflectedly like you is busy with such sort of ranking lists. The point is the >question what we could do if all different testings rely on the same false >assumptions? Ok, nothing wrong with it in the sense that some people want to >have fun and just playing around in a hobby sphere. Nothing wrong with such >passtimes. But again - you cant close your eyes and justify the overall >preference for a new entry like Rybka. What if we could find out that the >advantage of Rybka is based on two factors? > >1) the always existing advantage of a new entry which was freshly tuned on all >old ones > >2) something perhaps new in the code of Rybka that will become known > There is always progress and the two factors you are mentioning of course play an important role. We are well aware that authors learn from each other and apply concepts (I am not referring to cloning here, but to include ideas in an own way adapting it to an existing program, if it is somehow compatible). For Rybka I am sure that there is a lot of chess knowledge included and you cannot deny this when watching closely the games. >If then after a while the situation is equal on a slightly higher level, what is >the sense of it? Isnt it a problem for you if you see Hiarcs being tuned >successfully and whoopie, the big advantage is already minimalised? I'm not a >supporter of SSDF but one factor in their testing design is remarkable. They >dont take any new entry, no matter how complete, Beta or not, and produce data. >I wouldnt say that this is all well planned and justified. I remember the >general critic that always the strong new entries were tested with weaker >hardware... But with such a "slowness" you at least reduce the possibilities to >become part of the PR of something. Ok, fine, in the end that critic was still a >valid one somehow, but this is a different topic. > >Q: Dont you fear to produce just neccessary data for the interests of other >people? The question goes in all directions and testers of course. When you think that it is propaganda to publish such games and results this is your right of course. For me I can only tell that I do this because I am enthusiastic about what I am seeing. The level of play has reached a really exciting level. Same goes for Fruit having reached new heights with apparantly very clear and sophisticated structures of the program. I know that you do not like that especially those two are mentioned so much, but this is due to the enthusiasm their appearance caused after years of almost standstill and it is legitimate to utter this enthusiasm without being one interested in doing propaganda for certain programs. In the same way I can mention that I am also delighted about progresses seen in Fritz 9 and Hiarcs 10 and I do not really doubt that their authors also profited from ideas that were shared by some. I do not know how to explain all this better. But I think you will not really doubt that testers want to be independent and try not to favour certain programs too much. It happens anyway that new and sensational programs will be in the focus of interest of anybody and maybe overtested. That´s normal. Best Regards Heinz
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.