Nate Silver wrote a bizarre article at FiveThirtyEight last night, For Columnist, a Change of Tone. It is a parody of an article in The New York Times and it attacks Paul Krugman, implying that Krugman has been attacking the new website because it left The Times. Since it is done as satire, it is hard to know what to make of it.
As you probably know, Silver became something of a media star when he predicted the last few elections quite accurately. And liberals especially liked him in 2012 because his model told liberals what they wanted to hear: Obama was winning the election. At no time during the general election did his model have Romney even close to Obama. That included the period after Obama’s less than impressive first debate. Silver was the main counterweight to all the pundits who talked about ridiculous notions of “political momentum” and the distribution of yard signs.
If we take Silver’s article as fundamentally serious, it shows that Silver really doesn’t understand the nature of the many complaints about his new venture. All it shows is that there is a correlation: since leaving The New York Times, Krugman has been critical of Silver. But that’s not the only thing that happened when Silver left. He also greatly expanded his work and has put out articles about global warming and economics. These are fields that are distinctly different from poll aggregation. So the correlation could just as reasonably cause one to conclude, “Krugman doesn’t like Silver since he branched out into fields he clearly knows nothing about.” Or there could be other reasons. The “leaving The Times” narrative seems highly unlikely.
But is Silver serious? He ended his article with a paragraph that I might have used to lampoon him:
While it can be easy to extrapolate a spurious trend from a limited number of data points, the differences are highly statistically significant. At his current pace, Mr. Krugman will write 425 more blog posts about FiveThirtyEight between now and the 2016 presidential election.
I get the impression that Silver isn’t really sure what he wants to say. On the one hand, he seems to be bothered by all the attacks on his new site. On the other, he seems wryly aware that his statistical work has been thin.
The best take on the article is that Silver is saying that everyone is jumping on the site based upon little data. But the article makes it appear that Krugman was right in his initial assessment of FiveThirtyEight, “What would be really bad is if this turns into a Freakonomics-type exercise, all contrarianism without any appreciation for the importance of actual expertise.” I’m afraid so.
The biggest problem, is not the current criticism of Nate Silver. It is rather how much praise was heaped upon him before. I really liked his work during the 2012 election. But he was not the only one doing that kind of work and not even the best. To his credit, Silver always said that there was nothing special about what he was doing. And as is now becoming clear, there is nothing very deep about Silver’s thinking. For that reason, it is probably best that he stick to sports and other kind of “horse racing” analysis.