Author: KarinsDad
Date: 10:57:07 01/29/99
Go up one level in this thread
On January 29, 1999 at 13:27:07, Rajen Gupta wrote: >On January 29, 1999 at 02:18:03, Jouni Uski wrote: > >>As predicted we have new number one and even number two. >>In list H7 has 2576 and F5.32 2575. But Hiarcs 6 and Fritz5 had over 2580, when >>they got first rating so where is the progress?? Nowhere I fear. > >The progress may not be dramatic every year and only may be incremental, but it >is definitely present lets say if you compare 3-4 year periods rather than just >year on year.It is like when a new intel chip comes out and the comp magaxines >do their obligatory comparisons and round ups- ''eg the 450 Mhz is not much >stronger than the 400 which itself is not much stronger than the 350, and so on >but in reality if you compare lets say the 450 to a chip several steps down >their will be a significant difference. > >I suggest some one does a test consisiting of the best programmes of jan 1996 vs >the top programmes of the current list on equivalent (preferably the best >possible) hardware. That will give us a true idea of how much software progress >has taken place. >Rajen gupta Sounds like a good idea. Everything should be identical including hash table size, program settings, hardware, OSs, everything. And unlike the SSDF, you would not have to limit earlier versions to what settings and hash table sizes that were used in previous tests (to ensure that you are playing against the "x level of ELO" for that program). The test is not to find out what the ELO would be, but rather to find out how much better the programs play. KarinsDad
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.