Just started reading this interesting thread from a few years back. And given that we now have 5 more years of data to work with, I was wondering if anybody out there is interested in taking up the conversation again?

http://www.footballoutsiders.com/dvoa-ratings/2005/evolving-predictive-p...

First off, let me state that I'm not a huge gambler, mainly office pool stuff. My first question is what kind of benchmarks should we be shooting for. I assume a 60% win rate for ats pix is very good, but what about for SU pix?

Also, in order to compare to pointspreads, I convert dvoa into points using a factor of 0.04 which seems to work pretty well. I have done an adjustment for home field advantage to mixed results but that is all I do in a very simple analysis.

Anybody else out there interested in sharing?

Posted by: krugerindustrialsmoothing on 10 Jan 2011

0 replies