Time series forecasting problems

I am trying to predict weekly power ranking of NBA data with variables offensive rating, defensive rating, strength of schedule, home or away games, net rating, and pace.
So the problem arises how to multivariate time series forecasting with these variables or I need to add more variables and should I go with some other methods of regression

It seems like you are regressing Y (power ranking), which I presume takes on integer values, with several X's, some of which may be continuous and some of which may be discrete. Perhaps a Poisson regression would be best in this scenario.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.