Northeastern University

And the winner is…

Print Friendly
May 24, 2012

UPDATE: It looks like our guys were right! Last night, Phillip Phillips was crowned the 2012 American Idol winner. According to the US twitter data, that’s exactly what Sternberg Distinguished Professor Alessandro Vespignani and his team over at Northeastern’s MoBS-lab predicted.

This means a couple of things:

1) Apparently international voting isn’t as much of a problem as I was starting to think. If the international data had been counted, Jessica would have won.

2) Perhaps more importantly, the straightforward simple method used by the MoBS lab to predict voting outcomes was robust enough to make an accurate prediction about a straightforward election. We can all breathe a sigh of relief now that we know our network scientists aren’t out of their day jobs. The method they use for American Idol provides a framework on which they can improve to make more important predictions in the future.

***

May 23, 2012

Well, kids. It’s here – the moment you’ve all been waiting for. Vespignani and team have spent their morning analyzing all the #idol twitter activity between 10pm last night and 5am this morning and have come up with a prediction…and a caveat.

Looking exclusively at US-based tweets, Phillip is our man, but only by a small margin:

But as the team explain in their article, there’s quite a bit of international Twitter activity during the voting period, despite the fact that votes are only supposed to come from US residents. There are plenty of loopholes, however, and in past simulations, the team has had more accurate predictions when they include the international data. Which last night looked like this:

So, what does this mean for the actual winner? I guess you’ll still have to tune in tonight to find out. But if Jessica wins, it might be a sign that international voting is a bigger issue than people may have known.

Did you vote and/or tweet last night? Who do you think will win?

If you’re in California, Michigan or Maryland, chances are you’d answer that question with “Jessica.” Most other parts of the country? Phillip:



One Comment

  1. I totally agree with your studies. This is not the real judgement.

Follow

Get every new post on this blog delivered to your Inbox.

Join other followers: