help with running a regression on nfl stats

[rbskee]

Hey guys, just wondering if I could get some help with the best way to use my data.


After the challenge system was introduced into the nfl in 1999 the % of home wins went from 58.54% (from 1970-1998) down to 57.52% (from 1999-2009). I have data for all the games. I've got the variable ptsdif with home team score minus the away team score therefore a + = win and a - = a loss.

I firstly created a dummy variable for post99 which i made 1 and 0 would be for pre99 games. I ran the regression with ptsdiff as the dependent and post99 as the independent and got a p-value of 0.3379 for my dummy variable which i guess doesn't help me much.

I was just wondering if there is some other way I can use the data. I just want to be able to say there was or was not a significant change in home wins after 1999 which can possibly be explained by the coach challange rule.

Thank you.

 

[koupparis]

My stats are a little rusty, but I think you might want to run a t-test. Tests for differences in means under various assumptions of population size and variance.

 

[rbskee]

Ahh okay. So split the data into 2 separate pieces, and then run differences tests?