There are no doubt going to be lots of conversations and think-pieces about the Hottest 100 in the coming days and weeks. Here are some statistics that you might find interesting and/or useful about the gender balance and country of songs in the countdown. I've also included a look at how accurate the predictions from @socialhottest were.
I've done something similar to this before, for the Hottest 100 of the Last 20 Years.
In all of these calculations, if a band appears more than once in the countdown, they also appear multiple times in these statistics.
Women in the Hottest 100
I've used two summary statistics to measure the gender balance in this years Hottest 100. The first is female bands, that is, bands in which women make a significant contribution to the music. So a band that is full of men but has a woman playing the triangle in the corner doesn't count as a female band. Obviously this is subjective, to give you an idea of the decisions I've made, I have not included groups full of men that had a female guest vocalist, or bands which have three men and a female drummer. Of the 100 bands in the countdown, 21 are female.
The second summary statistic is a more straightforward calculation of the number of female and male musicians in the countdown. I have included featured artists in this calculation. There are 273 musicians in this years countdown, of which 34 are women.
The first two figures below show these two statistics visually. The bottom figure shows the density of female bands throughout the countdown. This was to see if one gender or the other dominates any part of the countdown.
Women are pretty evenly spread through the countdown until you get to the top twenty, when you get a complete dearth of women. The only two women in the top twenty are Lorde at number 18, and Sia at number 9.
Countries represented in the Hottest 100
The figure below shows how strongly each country is represented in the countdown. All up 59% of the countdown is Australian.
How close was @socialhottest?
Finally, the Twitter account @socialhottest did a 'Warmest 100' style prediction of what the results would be, by manually harvesting and entering votes from social networks (primarily Twitter and Instagram).
So how close were the predictions? The figure below is a way of visualising this.
Along the horizontal axis you can see the actual Hottest 100, and the vertical axis shows the @socialhottest prediction. If @socialhottest predicted a song in the correct place, it appears on the diagonal dotted line. If they predicted would do better than it did, it appears below the line, and if they predicted it would do worse than it did it appears above the line. Obviously the closer the predictions are to the dotted line the better.
Any song that is on one of the lists but not the other I have recorded as coming equal 105th on the list that it is missing from.
So how did it do on the whole? A bit of a mixed bag. The predictions are not very good toward the bottom of the top 100, but they do improve markedly as you get closer to number 1. That's not really surprising, as there are more votes for those songs and hence more robust estimates for their final proportions. It did well on the top five but still couldn't pick the number 1.
This shows that triple j removing the ability to share your votes through social media has made predicting a much harder exercise than it has been before.
If you've got any ideas for other statistics to look at, or think I've made a mistake, please let me know!
UPDATE: People of Colour in the Hottest 100
Someone on Twitter asked me to look at the number of people in colour in the countdown. I've thrown this together but I can count 13 bands that have at least one person of colour in them. Again, I may have made an error, let me know if you spot one.
Updates: The Person of Colour section originally said 11 bands had a person of colour
11:33pm: updated the female musicians data after some feedback (in the comments) - left a few featured artists out of the data.