Author: Uri Blass
Date: 09:52:07 10/16/00
Go up one level in this thread
On October 16, 2000 at 11:07:10, Robert Hyatt wrote: >On October 16, 2000 at 01:35:52, Christophe Theron wrote: > >"> >> >>You have told me the same when Rebel-Tiger I has been released last year. :) >> >>Others have said the same. The theory was that Chess Tiger 12 had not been >>released, and that was an unfair advantage because the competition had no >>opportunity to "tune" against it. >> >>Well one year after I still don't see any program outplaying it badly... >> >> >> > >You missed the point. After one year I don't see _it_ outplaying the >other programs significantly. It started on top of the SSDF. It is no >longer there. That was the point. I personally believe that if _any_ >programmer finds something "new", it will be "understood" by the rest of >the programmers within a reasonable amount of time. I do not expect tiger or other programs to be number 1 when it is one year old but I think that the reason is simply the fact that everybody gets better. I do not think that tiger lost the first place in the ssdf list because other programmers learned it. Tiger was not leading by 100 elo but only by 30 elo or something similiar. Other programmers added knowledge that is not dependent on tiger. For example Fritz5.32(that was number 2) did not use the nalimov tablebases when Fritz6 is using them. I guess that this change without more changes is responsible for closing all the gap or most of the gap. Some other good changes helped Fritz6 to be clearly number 1. 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.