Estimating Home Ground Advantage by Venue

In the previous blog I fitted models to the game margins of each team separately, seeking to explain the margin in any game in terms of the Venue at which the game was played and three "Excess" variables summarising from the designated home team's perspective its relative Venue Experience, MARS Rating and recent form.

There we found that, in explaining the eventual game margin: 

  • for all teams, Excess Ratings were important, contributing between about 50 to 80% of the explained variability 
  • for most Victorian teams, Venue was only moderately important, contributing between about 10 to 20% of the explained variability
  • for most non-Victorian teams, Venue was more important, contributing between about 30 to 40% of the explained variability 

Today I want to complete the analysis by providing the results of fitting the same model, not to the results for a single team, but to the results of all teams.

For this purpose I've used the MCG as the reference home ground, so all Venue coefficients should be interpreted as the Home Ground Advantage (HGA) for that particular Venue relative to the advantage at the MCG.

So, for example, teams playing as the designated home team at Aurora Stadium, of which there have been 48, have enjoyed an HGA of an additional 7.3 points per game relative to the HGA of teams playing at the MCG.

Note that the coefficients are positive for all but two venues, Cazaly's and Gold Coast Stadium, indicating that HGAs are generally higher at grounds other than the MCG. Also note that the only venues for which the difference in HGA is statistically significant are the SCG, and Subiaco, both non-Victorian venues, and Kardinia Park, which we saw in the previous blog behaves as if it were a non-Victorian venue.

The positive relationship that we saw for many teams between game margins and Excess Venue Experience, Excess Ratings and Excess Form is evident here too, with the coefficients on all three variables positive and highly statistically significant.

Substantially diminished, however, is the importance of game Venue. When we shift from considering games on a team-by-team basis to a more macro view, Excess Ratings emerges as the pre-eminent basis on which to predict game margins, with every 1 Rating Point increment associated with an 0.6 point increase in the final game margin for the designated home team.