yeah, man. You copy/pasted that before. Is there a link to that study or the math behind it? Like I said in the other discussion, the math seems really screwy in that post. Why would they change their regression model, and say that based on the old data their regression model is better too?

Regression models are an old concept. You'd think that in something as important as this that different fits would have been tried, and if a different one was more appropriate based on their data, they would have changed their findings...