Uncertainty About Uncertainty

The lesson I took away from Nate Silver's excellent book The Signal and the Noise is one that Donald Rumsfeld pointed out during the war in Iraq: "there are also unknown unknowns -- there are things we do not know we don't know," and Silver -- who believes this -- interviewed Rumsfeld for the book . . . though that chapter is rather anti-climactic, the rest of the book is comprehensive, entertaining, logical, and enlightening; Silver believes that the science and math behind forecasting is improving, and that our predictions are improving as well -- but the way we frame and use these predictions is growing more political, polarized, and manipulative . . . and so we need to realize with all statistics and predictions: political polls, numbers about the economy, the weather, sports, etcetera, that these numbers are simply a stab -- not a stab in the dark -- but a stab with a particular likelihood of hitting the target and eviscerating the truth from it and a particular likelihood of missing the target completely . . . and if you can think that way, you should become a scientist, and if you can't, then you should become a politician.

No comments:

A New Sentence Every Day, Hand Crafted from the Finest Corinthian Leather.