Washington DC – In what is seen as a huge victory for “Big Data”, a sophisticated bayesian model predicted the outcome of the American presidential election a whole seven minutes before the result was confirmed.
Spokesperson for the research unit that carried out the analysis – Nate Platinum – said “This is a huge step forward for big data analytics. Nobody knew the outcome, we were a full seven minutes away from knowing the outcome and our Bayesian model got it 100% right”.
“We ran two models – our model based on the polls we were monitoring on a minute-by-minute basis (we call this our “polls plus plus plus plus” model) and our model that used the polls and other information about the color of the morning sun, what Yoko Ono said recently, the frequency of entirely new cat gifs on the internet on any given day and whether or not anybody has seen fire-breathing unicorns lately. This is our “polls++++++” model. Its a completely different model.”
“Our polls plus plus plus plus model is 100% accurate, but our polls++++++ model is even more accurate.”
“Both models predicted the outcome of this election and got the result 100% correct, a full seven minutes before the result was known.”
“Did I say how this was a huge victory for statistics?”