I first heard that Nate Silver has added a climate change denialist to his blog on PZ’s blog, but it took me a bit of time to check it out. By the time I got to 538 and under the Science tab, sure enough, there it was. Roger Piekle, Jr., well-known shil for the carbon folks claiming that climate disaster are costing more simply because we’re richer (hey, go see where the damage was in Sandy, sure a few rich peoples’ beach houses, but plenty of crummy neighborhoods, plus the subways in NYC have been there for decades and never flooded before).
Now if there is anyone who should know how to deal with statistics it should be Nate, but he invites a well-known liar, who cherry-picks statistics to try to prove his preconceived (and bought-and-paid-for) notions. Come on, Nate, you know that bias in statistical analysis is bad, you say so in your book.
Even in his book Nate points out the problem. In reporting on how the Japanese chose to ignore power law in earthquakes and deliberately put a “kink” (instead of straight line, they chose an arbitrary inflection point is the scarce data portion of the curve). That meant they built the seawalls too low around their nuke plant, just to say investors a few bucks. After the single large earthquake datapoint was added to the dataset, there is was, the actual power law relationship validated. Nate uses this as an example of how to lie (or, at best, deceive yourself) with limited data.
So now he hires a guy to do exactly that on his blog. The general belief is that Nate wants to stir the pot a bit, get the justly deserved outrage, and thus drive hits to his blog and thus his pocketbook. In the absence of any hard data this is as good an excuse as any. Even his buddies at HuffPo are denouncing his climate change denialism.
So fame and the money it will bring can corrupt anyone. So sad, Nate. So as others are saying stick to sports forecasting and stay out of science you don’t understand.