Ill preface this by saying its 6am and Ive been at work for 12 hours. I may need to address or clarify my responses later.
This puzzled me as its not overly complex.
The metro average is clearly derived from the number of people watching broadcasts of AFL football. The number of people is divided by the number of broadcasts. This average is done by city, and is done by total metro area.
Frankly I dont believe it CAN be done this way and be accurate on a per broadcast basis. The number of cities is not a relevant factor in the average only the number of people in the ratings total and the number of broadcasts they watch.
Why thanks.
You'll find a hell of a lot of matches are single or double city games only. Over the course of a season very few games are actually national - about a quarter.
Since I calculate on ratings a per match basis and not a ratings per city one, I dont believe it is wrong. The number of games per city isnt really relevant.
Given i use a ratings per broadcast total, and not a ratings per timeslot or other methodology, the ratings of other cities are in fact going to have an effect on the overall average being lower than the Melbourne top figure.
We reconcile it by acknowledging that most cities rate lower than Melbourne and it brings the average broadcast audience down.
When we measure average audiences per city we get an average audience per match broadcast in that city.
What we get then when we add average city ratings together to get some sort of super average I have no idea, but its nothing to do with ratings per game at that point and seemingly everything to do with trying to beat the percieved advantage the NRL has by only having 3 matches per round all broadcast nationally.
Well, as a comparator to the nrl, such a "super average" is far better to a metro average that is blind to the fact that only a quarter of matches are actually telecast across all metro areas (and perhaps just as many a broadcast in one)
If I were you (ie I was providing an excellent free service to all the data nerd football fans and code warriors) I would add an extra row in to the summary and aggregate/average all the standard c7 games...
...another minor addition I would make is to add a "viewer hours" to reflect the longer afl games. Ironically, the only time I've seen viewer hours used is when someone was trying to devalue the afl tv rights value!