Author: Roger
Date: 19:43:20 12/03/99
Go up one level in this thread
On December 03, 1999 at 20:03:47, walter irvin wrote: >On December 03, 1999 at 19:19:42, Bertil Eklund wrote: > >>On December 03, 1999 at 13:37:58, José de Jesús García Ruvalcaba wrote: >> >>> Instead of giving code numbers to the secret entries, give them codenames! I >>>suggest using the names of famous masked wrestlers, like 'el Santo', 'Blue >>>Daemon', 'el Solitario', 'Lizmark', 'Canek', 'Atlantis', 'Mil Máscaras', >>>'Universo 2000', 'Pentagón', 'Pierrot', etc. >>> That way, you can publish the list with all the entries, without compromising >>>the secret nature of the programs, giving the readers a chance to compare their >>>progress from one list to the next. Also, the use of masked wrestlers names >>>seems fully appropiate, as their identity is also secret. >>>José. >> >>Hi! >> >>I have already suggested that. I´m tired of all this nonsense already and I >>suggest we should publish all results of all programs. If any company have any >>prevarications about our results let them prove why our results are misleading >>and let this forum judge who are right. >> >>We are still forbidden to publish results with Shredder4 but in my opinion it´s >>our duty to do so whatever mr Weiner says. An alternative should be that mr >>Faber(as suggested by himself(thanks!)) or someone else publish the results! >> >>He have tried all kinds of manmanouvers to make our list doubtful, but from now >>on I think he should deliver clear proofs for that. He or someone else also >>seems to have used torpedoes that tried to prove with statistics that our >>organisation is corrupt. In this forum it was a matematician who tried to prove >>that the SSDF was corrupted, I guess the same man tries to understand the world >>with statistics also. This man thought for instance that christmas was in >>october or november, but maybe he could prove it with statistics. >> >>Only my personal opinion Bertil > >i think as far as computer programs go ssdf is it .i think ssdf is 10000% right >on the money . the only problem is the list does not seem to translate very >well or close to a fide rating .some where the list is wrong either at the low >end or the high end .maybe the ratings would be better if a match with fritz >5.32 128mb k6-2 450 mhz and a fide rated 2650 + rated player at tour time >control at least 6 games .then on basis of the result of that match the ratings >could be adjusted up or down .at least then it could be closer to fide elo . Perhaps the list can be recalibrated as more games versus titled players become available. Recalibrating with one program alone for a small sample of games would not be advised. Roger
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.