Author: Gian-Carlo Pascutto
Date: 15:19:20 03/30/02
Go up one level in this thread
On March 30, 2002 at 16:25:01, martin fierz wrote: >>My main argument against it would be that the SSDF ratings have >>(approx) 30 elo error margins to either side. If you compare the >>ratings of two engines, you have to cope with this twice. The >>speedup ELO numbers are on the order of 50-130 ELO. > >statistics, GCP! adding two random errors does not give twice the error, but (if >they are equal), sqrt(2)*single error. sqrt(a^2+b^2), which reduces to your formula for a = b I wasn't sure of it when I posted this, which is why it was intentionally vague (I never implied they had to be added). -- GCP
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.