Tuesday, November 20, 2012

DeLong on Nearly Everybody Else v. Nate Silver

DeLong assembles an amusing catalog of criticisms leveled against Nate Silver and Silver-style polling.  I assume a fair number of these are just routine hackery but what strikes me as remarkable is how many of the critics really seem to believe what they are saying.   DeLong's commentators do offer a few helpful nuances but for the most part, the critics still look pretty ridiculous.  Grant that Silver is going to blow one bigtime someday--that's  in the very nature of probabilities.  But for the moment, I'd say that the evidence is all on his side and it really is impossible hard to conceive why anyone would think they could credibly challenge him.   Reading Silver's book (The Signal and the Noise), however, I can see one good reason why pundits harbor such animosity against him: in his own direct critique, he really does make them look like idiots.

1 comment:

Jimbo said...

Yes, a lot of commentary on this. Bottom line: right-wing pollers and Rmoney internal pollers were deliberately biased to bolster their "reality". Nate Silver's algorithm included a large number of polls and, as all scientists know, the larger the sample the lower the confidence interval.

So, this argument is finished and so let's move on.