In this age of social media it’s hard to keep a lid on things. As such, a rather clever individual has scraped the net for people’s Hottest 100 votes, and come up with a prediction for the countdown, which can be viewed at warmest100.com.au.
He reckons to have collected about 3% of the vote, which may not sound like much, but should actually be enough to draw general trends from. I’m very happy to see that almost all the songs I voted for have ended up on his list – but I’m less happy to see he’s tipping the awful Thrift Shop for the number one spot (I’m even less happy that he’s probably right…).
In any case it’ll be very interesting to see how his list compares to the real one come Saturday.
Predicted Positions for Songs I voted for…
97 – Spiritus – Lisa Mitchell
67 – All the Rowboats – Regina Spektor
46 – Run Alone – 360
26 – Bangarang – Skrillex
20 – Same Love – Macklemore and Ryan Lewis
16 – I Love It – Icona Pop
10 – My Gun – The Rubens
2 – Little Talks – Of Monsters and Men
D’OH !…
BETTING has been halted for Saturday’s Hottest 100 countdown after a website published its list of the likely outcome based on an analysis of the public voting system.
Listening to the ABC’s Triple J, cracking a beer and counting down last year’s most popular songs is an Australia Day tradition – with plenty of people taking a punt on what will be No.1.
But some bookmakers have now suspended betting, after Brisbane-based marketing worker Nick Drewe and his mate Tom Knox published their predictions on their Warmest 100 website.
The pair analysed 35,081 Hottest 100 public votes, submitted to Triple J via Twitter and Facebook, to come up with what they think will be the most popular tunes.
University of Queensland Chair of Applied Statistics Professor You-Gan Wang believes the Warmest 100’s predictions will be mostly correct.
“I think they’ll probably get most of them right,” he told Fairfax Media.
..As of Tuesday, a number of bookmakers stopped taking bets on the Hottest 100 – though it’s unclear how many did so as a direct result of the Warmest 100 site.
Tom Waterhouse said the betting had been suspended.
“Yes that market has been taken down,” the company said via email.
What lies behind your assertion that a sampling of 3% of the votes is enough to draw general trends from. I am not disputing it at all as it seems quite reasonable. I am simply interested to know the reasoning and your knowledge that allows you to arrive at that.
Don’t feel obliged to give a written reply. We can converse about these and many other delightfully genteel topics the old fashioned way on ‘Imperial Invasion Day’, a.k.a. ‘Australia Day’.
Statistics good sir! It seems counter-intuitive, but both history and mathematics show us that once you’ve got a few percent of a vote, you’ve got enough data to come up with a broadly accurate estimate of the rest of it.
The guy behind this analysis has randomly collected 3,602 sets of votes. What is more likely – that those 3,602 votes are pretty much average, or that he managed to locate 3,602 people who’s musical taste is so weird that their votes are completely at odds with the rest of the voting population? The second scenario is not impossible, but is so unlikely that the first assumption can be assumed to be correct.
Obviously, the more data you have the more accurate your result will be, but 3% is enough to get things broadly right. I would be very surprised if any more than a few of his predictions are more than 5 places off in the final poll.
Oh, and haven’t you heard? The evul guv’mint lead by evul Ju-LIAR is trying to take away our precious bodily fluids by renaming Australia Day to “Citizenship Day”. It’s all over Facebook, so it must be true!!