
BGonline.org Forums
NFL Sunday critical field goals  OT (math)
Posted By: Casper van der Tak
Date: Wednesday, 25 January 2012, at 6:34 p.m.
In Response To: NFL Sunday critical field goals  OT (math) (Chuck Bower)
I think it is nice to try the log form of the equation, the coefficient before the log of the explanatory variable then gives you the required power (instead of assuming it is 3). That will potentially improve the model a little  or confirm that the quadratic form is fine.
For the estimation technique  if a lot of observations are at close to 0% or 100% success rate, you need to be careful, if not, weighted least squares is fine. My guess would be that you do not need to worry.
I would not worry about error in variables, that is very complicated to address and does not gain you a whole lot. Academic Journal referees would love it, but as you mentioned...
Good luck, and I still hope that some day someone in the NFL will be smart enough to use your work!

BGonline.org Forums is maintained by Stick with WebBBS 5.12.