spacer image
spacer image

Welcome! You're looking at an archived Snarkmarket entry. We've got a fresh look—and more new ideas every day—on the front page.

December 3, 2008

<< I Am A Robot. Can I Help You? | Me and My Seven Genius Friends >>

How Nate Silver Brought Sanity To Polling

Matt Yglesias says, eh, Nate Silver’s not all that great; Ta-Nehisi Coates says Silver should call Matt a washed-up punk on YouTube (since Silver is clearly Souljah Boy to Yggy’s Ice-T).

I say 538 wasn’t great in this election season (just) because Silver’s formula worked; it was great because it so consistently tempered the insanity of polling fluctuations (including at by identifying erratic data, bad sampling, house effects, and other quantitative noise. In other words, Silver’s formula (and his explanatory rationale for it), instead of just being an aggregate output, actually helped its readers to make sense of the broader universe of polling, from process to results.

As a result, the blog wound up being one of the best political reporting sites on the web. It helped take political junkies from obsessing about “the polls” as an undifferentiated black box out of which numbers spewed into something they could understand and criticize. I also can’t say enough about its calming effects — every time a friend would call me freaking out about some new polling “shift” (usually as a result of one poll’s numbers following another’s, or Drudge beating a cherry-picked drum), I was able to talk them down, using Silver and 538 as my authority.

When virtually every political blog is devoted to channeling outrage, it’s salutary to have one that, even when challenging the CW, reassures.

Posted December 3, 2008 at 8:37 | Comments (4) | Permasnark
File under:


Every time a friend would call me freaking out about some new polling "shift" (usually as a result of one poll's numbers following another's, or Drudge beating a cherry-picked drum), I was able to talk them down, using Silver and 538 as my authority.

This right here is insight. Man, how many times did I do the same thing this year?

I'll give you a great example, in the form of a missed opportunity. Look at Silver's November 11th post on Proposition 8. This is right around the start of the silliness about black and Hispanic voters being the "difference-makers" on Prop 8.

* Silver debunks the notion that the surge of new Obama voters contributed to Prop 8's success. New voters voted against Prop 8 62-38.
* He points out that the demographic group most responsible for the success isn't any racial minority, but older voters.
* He doesn't point this out, but many people have noted and I, with my Silver-attuned polling senses, have noticed, that most of the "Blacks Pass Prop 8!" stories were based on a single poll with a very limited sample size of African-American voters (and no cross-tabs or balancing for age). In other words, a poll of somewhat dubious generalization value.

But here the data didn't calm anyone down. The "Blacks Pass Prop 8!" story was too good to check.

I can hear synapses fusing shut every time the "Blacks Pass Prop 8!" story returns.

As a journalist, it often troubles me when we become the noise instead of amplifying the signal. Nate and Co at 538 showed the value added of adding context.

It wasn't just analysis. The micro-reporting they did following the story comparing the Obama volunteer operation versus the McCain operation was stunning. Yes, Dems usually rely more on volunteers than Republicans, but when you had Republicans on 538 and Politico begging for a call just for a sense of affirmation, you could see the trouble McCain was in. The post with the photos from the volunteer shops right before the election said it all.

I think 538 also highlighted the danger of the CW anecdote becoming the story when it might actually be representative of nothing. An anecdote in the hands of good a writer is powerful but also can be dangerous. It can misrepresent the broader reality.

Thanks for the post.

spacer image
spacer image