One of the fascinating sideshows during the elections was the attack on Nate Silver and his predictions, and the subsequent validation of his model (and that of others who predicted with the same accuracy as well). Needless to say, his book, is now on the top of most bestseller lists as people try to figure out whether Nate was really a witch or not.
Hidden beneath the excitement of validation is also an underlying dichotomy in statistics – that between frequentists and Bayesians. While the details of this might be a bit too technical for a blog post (the earlier reference and this one can do a much better job than I ever can, if you are interested), it essentially boils down to whether you estimate the probability of a parameter given data, or whether you estimate the probability of the data given a parameter. Also if I remember my statistics correctly, a frequentist interpretation of the world requires an assumption of ergodicity, which is hard to arrive at without carefully choosing your sample populations for surveys. This article might explain the difference in easier terms.
Interestingly enough, recent research with toddlers shows that our early years comprise of Bayesian techniques of learning about the world. No wonder academics that have followed this fork in the road are gloating over the Silver lining to their cloud.
Where else is this a big deal? Take a look at the discussions about whether the God particle was indeed detected at the Large Hadron Collider or not as the two techniques collide.