You' ve probably heard about Nate Silver by now - a young statistician that has been incredibly accurate at predicting baseball outcomes AND the 2008 and 2012 US presidential election results (More about this at Simply Statistics Blog).
He uses a bayesian approach to combine data from several polls and historical data to predict election outcomes. Nate Silver gave 'an interview' last month for Tim Harford where he explains shortly about bayesian statistics and his predictions for the US elections.
How good were his predictions?
[Image Source: Simply Statistics]
