Wednesday, October 08, 2008

A Statistician's View of Electoral Politics

I like to think I'm an informed voter. I'm sure most people think they are too. But it's harder and harder to avoid the blatant partisan blather that passes for commentary on the cable news networks. No matter what issue is raised or what question is asked, the panelists spin into an attack on the opposing candidate. Enough already.

There's a great new web site that caught my attention earlier in the year called, and I've been recommending it to many of my friends. And here's the twist; after reading it every day for weeks I discovered that it was the brainchild of Nate Silver, a baseball analyst best known for his predictive models to project player performance. Nate has applied his advanced statistical analysis to study political polls, compiling polling data from hundreds of sources and measuring the effectiveness of different polling techniques.

Wikipedia offers this overview:

[The site] compiles polling data through a unique methodology derived from Silver's experience in baseball sabermetrics to "balance out the polls with comparative demographic data" and "weighting each poll based on the pollster's historical track record, sample size, and recentness of the poll." also uses computer models to simulate the election 10,000 times per day in order to provide a continually up-to-date assessment of probability for electoral outcomes.

Notably,'s new polling methodology gained national attention for beating out most pollsters' projections in North Carolina and Indiana in the heavily contested political primary race between Senators Barack Obama and Hillary Clinton in 2008.

Nate launched the site in March, and by the end of September it was attracting 2.5 million unique visitors a week.

A lot of folks probably aren't interested in this sort number crunching, but I find it fascinating. It's an injection of science into what has, at least for the general public, been a nebulous process. Maybe Nate will do for politics what Michael Lewis and his book Moneyball did for the game of baseball.