I posted this a few days ago across the street, and have yet to get any feedback.....so I'll try here
I've been dabling with some correlations using the Four Factors. My data is from Pomeroys site, but I imagine one could retreive it from anywhere stats are compiled.
I'm unclear on how any or all of them can be used for predictive purposes. Has anyone encountered a definative formula that produces positive results using the Four Factors in any combination?
I only have a "working Knowledge" of Excel, so although I maintain a database and have a successful model, I have trouble analyzing data as it is displayed in my current format.
On a VERY small sampling (Jan.21-27), I experimented with correlations using my models predictive evaluation coupled with eFG and REB's. I used Statfox data and arbitrarly chose a 50 pt rating difference between matchups....same with Reb's. In other words, if one team was rated 50 pts higher than its opponent in both eFG rating AND Reb rating AND fit into my model as a play....then I counted it. Everything else was disregarded.
I noticed that particular application exposed 24 "plays" that fit the criteria. The "plays" went 17-7. I know that is small enough to be inconclusive, however it's peaked my interest. Now here is where the trouble lies. I don't have the experience with EXCEL to properly create the formulas, so it is a tedious and time comsuming process for me.
In summation, I'm not afraid to pursue the analysis, and if anyone else has seen, read or accomplished a similar endeavor and found it to be useful, I'll continue. However, I'd hate to wade into this only to find it's been attempted previously and was a waste of time.
If anyone has an opinion on the usefulness, and more importantly, the predictive value of the Four Factors as a handicapping tool I'd like to hear it...
Thanks... and Good Luck on your action...