Would like to analyse 3 completed seasons for correlation and regression. Watch
Which should I use for the number of teams?
1. n=20 (PL teams only)
2. n=72 (EFL teams only)
3. n=92 (PL and EFL only)
4. n=98 (all big 5 leagues only)
And for the seasons, should I just use the 3 most recent seasons (2016/17, 2017/18 and 2018/19)?
I have completed my analysis of the 2018/19 Premier League on this. Would you all like to see it?
I don’t think three seasons is enough to acquire meaningful results but then I’m not 100% sure what you’re doing or trying to find out . PMCC?
Also do you take into account the fact that luck plays a big part I’m many results especially between mid table teams so that the difference between a team finishing 8th and finishing 17th is usually down to pure chance .
I can highly recommend the articles of Kevin Pullein regarding this kind of thing.