Teams sharing stadium in modeling

[Juret]

In modeling, how would you deal with match-ups of two teams with the same home stadium? (home field advantage is an important factor in the league)

[FourLengthsClear]

Over the last 25 years, AC Milan are 0.6 goals per game better as the 'home' team against Inter Milan as they are as the 'away' team.
The ticket allocation over that period has varied from 70% to 87% in favour of the home team.

This is far too small a sample to be conclusive of course but it is not too far off the league average over the same period.

[Juret]

interesting! but if that is a far too small sample, it is way larger than I have for this match-up. I'll probably have to assume the game is played on a neutral field because I except neither team to have a spectator advantage, it's a small league.

[Wrecktangle]

4 legnths: how many occurances?

[FourLengthsClear]

4 legnths: how many occurances?

50 Serie A matches, 57 in total.

[Wrecktangle]

50 Serie A matches, 57 in total.

That's enough occurances to wonder if something is there. Have you run a z-score off the assumption that there is no HFA (0.0 points)?

[FourLengthsClear]

That's enough occurances to wonder if something is there. Have you run a z-score off the assumption that there is no HFA (0.0 points)?

Not specific to these two teams, no.

My 'study' was aimed at establishing just how big a component of HFA the attendance/crowd/atmosphere was on a league to league basis in soccer.