Then the story.
Silver's political analysis is better than his sports analysis. The reason is pretty simple. Every sports contest can be subject to random chance or flukes. But every election has dry runs in the form of polling that show us the lay of the land.
Put another way, it is impossible to predict the outcome of a game, because there is no data that tells you what will happen in any particular game, whereas elections come with polling and polling CAN tell you what will happen if you read it right.
All Silver did was read it right.
But then again, so did a number of other people. Silver is not the Gay Wizard, he's just the highest profile practitioner of using polling averages. Blumenthal, Wang, Moulitas, Cohn and plenty others called the election the same way.
This wasn't Silver vs. the Old School, it was Numbers vs. the Old School. And Numbers won, with or without Silver.