By now, if you haven’t mailed off your same-sex marriage survey, then your vote won’t count – which is very good news only if you planned to vote ‘No’.

Some terrible, terrifying news came from The Conversation, however, who have used advanced data analytics to predict the likely outcome of the vote – and it isn’t looking good.

The system they used was developed at Griffith University’s Big Data and Smart Analytics Lab, and predicted Trump’s victory last year. As they explain, these analytics have “proven uncannily accurate at predicting the outcomes of hard-to-call polls.”

So, they looked at 458,565 ‘anonymised’ Australian tweets referencing same-sex marriage in October, weighed the sentiment of these tweets, then analysed the data.

72% of these Tweets were in favour, which seems like good news, until you realise that many users tweeted out numerous messages in support of same-sex marriage – over a thousand in some cases.

458,565 tweets comes down to just 207,287, with 57% in favour.

This is worse, but still acceptable.

Naturally, though, the demographics of those who Tweet certainly don’t match the voting population. Less than 15% of tweets were made by over-55s, who make up 36% of voters.

Adjusting the tweets to represent the age and gender make up of our voters, and the YES vote drops to 49%.

Sorry.

Of course, polls such as these aren’t super reliable. Those who tweet about social issues tend to be heavily on either side, meaning those with more moderate views, or who simply don’t need to broadcast their every opinion, aren’t collected in such a survey.

I suppose we will have to wait and see what the results say. Then, if the answer isn’t what we want: Riot. Revolution. Something.

Check out The Conversation for a more detailed data breakdown.

Get unlimited access to the coverage that shapes our culture.
to Rolling Stone magazine
to Rolling Stone magazine