I still don't like the Top/Middle/Bottom 6 theory.
Its expedient because users are used to looking at premiership tables so to present it this way is giving them something they are used to.
The key to good Data Visualisation is to present it in the best manner that people can understand the meaning of the data.
If you have this data for instance. I'll keep it simple (I've made up the numbers)
Average Handballs per game
Adelaide 158.8
Brisbane 157.2
Carlton 156.5
Fremantle155.8
GWS 155.6
Hawthorn 155.4
North 155.3
Port 155.2
West Coast 155.1
Gold Coast 155.0
Sydney 154.8
Melbourne 154.7
Western Bulldogs 154.6
Collingwood 154.5
Essendon 142.3
St Kilda 142.1
Geelong 141.8
Richmond 140.7
If you split these into groups of 6 Western Bulldogs and Collingwood are deemed to be in the bottom 3rd even though they had only 2.5% less handballs than the leader.
Therefore it is better to group them for instance as teams high usage of handballs those within 5% of the leader and those with less usage those with greater than 5% less than the leader.
The second way gives you a better interpretation of the data.
