Opinion of Nate Silver

[RFayette]

He's obviously smart and good at what he does, but his personal bias (and what he wants out of politics: a rural, dumb, poor GOP vs. a cosmopolitan, educated Democratic Party ... which we all know is painfully simplistic) drips from his writing.  Unfortunately, he's smart enough to "back it up" and not be obvious, but it's annoying.

I think he's gotten a lot less overtly partisan over the years.  If you think he's like that now, take a look at what he was like before he worked for either ESPN or NYTimes, way back when FiveThirtyEight was its own website.  

[Free Bird]

He's obviously smart and good at what he does, but his personal bias (and what he wants out of politics: a rural, dumb, poor GOP vs. a cosmopolitan, educated Democratic Party ... which we all know is painfully simplistic) drips from his writing.  Unfortunately, he's smart enough to "back it up" and not be obvious, but it's annoying.

I think he's gotten a lot less overtly partisan over the years.  If you think he's like that now, take a look at what he was like before he worked for either ESPN or NYTimes, way back when FiveThirtyEight was its own website.  

Link me to one. I'd like to see the comparison

[eric82oslo]

As far as I know, Nate Silver is a moderate independent, he's not a Democrat. He doesn't seem to have very strong political leanings anyways.

[An American Tail: Fubart Goes West]

538 has suffered a definite dive in quality since leaving the NYT for ESPN.

I've noticed it too. They don't put in as much effort in their election stuff as they did in 2012 and Silver did in 2008. I hope to see its quality increase over the next year and a half.

I'd still lean FF. He did really well in 2012.

[Nichlemn]

FF. The hate he gets from some here is pretty funny, it seems like some people have a grudge against him making a prediction against their favoured candidate once (most notably, recently claiming that Clinton isn't a clear favourite in the 2016 general election).

I don't know if his predictions are well-calibrated. But no-one can really know that, there hasn't yet been a large enough a sample size - but his results to date are pretty good.

It's his analysis and the underlying logic for his predictions are where he shines IMO. There's an awful lot of really terrible prediction out there (including in these forums) which Nate thoroughly debunks. Super cherry-picked, small sample sizes, or even completely baseless assertions. If only everyone in our prediction forums were required to read a selection of Silver posts and/or chapters from his book, it might help a lot.   

And most of his critics don't have any actual substantive criticisms about methodological flaws, they're  usually just thinly veiled whining like "Nate said X would happen with 67% probability, and it didn't! He obviously sucks!" 

[Penelope]

538 has suffered a definite dive in quality since leaving the NYT for ESPN.

Yeah. Nate Silver seems pretty intelligent and his analyses are usually pretty good, but the majority of FiveThirtyEight's content has become clickbait-with-numbers.