Hi all,
I am looking at the moment at sports data where I have two observations per Game, one for the Team and one for the Opponent. Therefore I have a dataset twice as large as the matches played. First I tried to xtset teamid but this yields not consistent estimates, I think because I need to include the opponentID aswell. Because I have two observations per unit cluster(MatchID) does not work aswell. Has anybody a hint / solution on how to deal with this problem with fixed and random effects?
Thank you so much in advance!
I am looking at the moment at sports data where I have two observations per Game, one for the Team and one for the Opponent. Therefore I have a dataset twice as large as the matches played. First I tried to xtset teamid but this yields not consistent estimates, I think because I need to include the opponentID aswell. Because I have two observations per unit cluster(MatchID) does not work aswell. Has anybody a hint / solution on how to deal with this problem with fixed and random effects?
Thank you so much in advance!
