Author: Sune Fischer
Date: 05:22:10 11/02/04
Go up one level in this thread
On November 02, 2004 at 07:42:24, James T. Walker wrote: >Just plot a graph with 999 data points all with a value of zero. Then plot >another graph of 1000 data points all with a value of zero. Then compare the >difference. >Jim Ratings can become negative, so zero is not the lowest rating possible you can get. Zero is actually an arbitrary number you just picked out of nowhere, you might as well have said -1000. The proof of this is easy, even without knowing the formula. Just consider the 1400 guy and someone who is 400 ELO weaker, then consider one that is 400 ELO weaker than him and so on, you quickly get negative ratings. I don't know what the lowest rating is, even if you assume perfectly worst play your opponent might also be playing perfectly bad or close enough to that so you don't lose every time. You can thus never get an expected score of 0 and an equivalent rating of -infinity. -S.
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.