| |
BGonline.org Forums
Correlation between ERs and MWC
Posted By: Maik Stiebler In Response To: Correlation between ERs and MWC (Marv Porten)
Date: Wednesday, 22 April 2009, at 4:09 p.m.
For gnubg-style error rates one could use one of the formulas described here in the article by Kees van Doel to estimate the elo difference and then use the FIBS formula for MWC.
For Snowie-style error rates, one could probably somehow adapt the R(DEPD)-formula from the link.
I think the Doel formulas lead to smaller ELO difference estimates than Zare's. I rather believe in the former. Of course, all this is more guesswork than science.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.