Nate Silver is the man.
A 31-year old 1996 East Lansing high school graduate was recently named one of Time’s 100 most influential people. Pretty good stuff. How did he pull it off? Pretty simply, he believes in data.
To be honest, lots of people believe in data. Silver’s angle is that he uses data endlessly to build statistical projections that are about as good as any you can find anywhere. So, logically, what does he use these projections to forecast? Baseball. Well, he started off in baseball, but now he is branching out. Based on his early success creating PECOTA, a statistical system for evaluating baseball players based on comparable players, he moved on to other contexts using similar methodologies. His most famous achievement is correctly predicting electoral voting outcomes in 49 / 50 states in the most recent presidential elections (he missed on Indiana) and every single senate race. Not bad! Considering how many people do professional political polling, his accuracy is unbelievable. At the heart of his approach is just pounding the data, using as much data as he can possibly devour, and weighting the data by its own past accuracy.
Embracing the fame that has come with his incredible penchant for accuracy, he spent last year’s election cycle on every national news show in the world, from the Colbert Report to CNN and the like. So, what is next? He recently tried his hand at predicting another hotly debated contest —- the Oscars. How did he do? Not so good. Taking stock of his performance, he has decided to hold off on projecting new contexts until he has his models better tuned. Turns out even Nate needs a little time to make the data talk, and rushing too quickly into the Oscars didn’t work.
I wonder what he has planned next? He already uses this stuff to play online poker. Maybe next he will go for predicting hurricanes and stock market shifts. One thing I’m sure of…. I’ll bet he gets rich!