Wednesday, November 7, 2012

The Horse Race That Never Was

The Votamatic prediction has barely changed since June and was completely accurate on the night
If you've watched any American TV over the last few months you'll have been under the impression that this was going to be a close Presidential election, a "nail biter", a real horse race. Two weeks ago many of the networks were saying that Mitt Romney had the "momentum" and a very good chance of winning. The numbers have told a different story. The horse race was completely fake. Sam Wang at the Princeton Election Consortium, Nate Silver at the 538 Blog and Drew Linzer at Votamatic never once showed Mitt Romney winning the election. Drew Linzer's prediction has been consistent since June and was exactly right in terms of the final electoral vote for Barack Obama. Nate Silver also called the election correctly and Sam Wang's prediction the night before the vote also called the electoral college exactly (although his election day prediction gave Florida to Romney). During the Republican primary the TV media also pretended that there was a horse race going on when it was obvious that Mitt Romney was going to win that election too. Every single network was guilty of perpetuating the fraud of the fake horse race, the question is why. I suppose the cynical answer is that pretending that there's a horse race generates better ratings and that's why everyone bought into the "white lie". Advertisers don't care why people are watching a show, just that they are watching, so it doesn't matter if you're informing them or entertaining them as long as they're there. Or, it could be that the pundits that fill up most of the airtime on political programmes just don't understand mathematics and statistical analysis. I think this is pretty likely. Political science and communication majors don't seem to get maths and on TV at least are weirdly proud of this fact. People who understand very basic maths and science are called "nerds" or "brainiacs" or "geeks" when none of this is rocket science. It's the stuff I studied in high school for my Statistics O Level when I was 15. The talking heads on Fox News in particular don't seem well versed in any of the hard sciences (evincing bafflement and disbelief not only about Nate Silver's stats but also apparently confused by basic physics and evolutionary biology too) but they're not the only guilty ones: MSNBC, CNN, The New York Times, The Guardian, The Washington Post, The Daily Telegraph, even the BBC are just as guilty of pretending that was going to be a coin toss election when since June anyone who studied the numbers from the state polls could see that it wasn't going to be close at all. Three pundits in particular, Joe Scarborough of MSNBC, Dick Morris of Fox and George Will of ABC should be laughed off the national stage because of their anti science views, but they won't will they? And you'd think that the day after the election there would be a massive epistemic rupture on conservative websites like the Drudge Report, Fox News and National Review who actively lied to their desperate readers reassuring them throughout the campaign that Mitt Romney was ahead. Epistemic rupture my foot. Today these websites continue the same dissonance as if nothing has happened.