multiple regression for Ncaaf

[MILK1123]

How do I set a frequency/ count in columns, so that the number cells matches my dependent variable range? Example: Wisconsin had 13 seniors, 24 juniors, and 17 freshmen in 2010. I want to include these values in a regression analysis to see if the precent of each impacts the point margin. Point margin column has 122 cells. I set up columns for each class and filled them with the frequency, but with no viable results.

 

[shg]

Welcome to the board.

If you have the same team composition over all games (no sophomores?), there's nothing to regress.

 

[MILK1123]

There are different percentages of each class across 10 years of data. My point margin column is for the entire 10 years.

 

[shg]

OK; so you could use a column that had weighted average seniority of players (Senior=4, Freshman=1) and regress point spread against that. Maybe I'm missing your question?

How do I set a frequency/ count in columns, so that the number cells matches my dependent variable range?

I don't understand what that means or how your data is arranged.

 

[MILK1123]

I've thought about that, but it would only work for for year by year not game by game. When I try to do that for game by game, I get an error message in regression. I've even tried log and basicaly every thing I could find on the web. I just can not figure out how to set that info up to make it viable in regression. Thanks for replying.