All errors should be reported to

Sunday, March 12, 2017

Nate Silver, still clueless after all these years

Nate Silver, the media's favorite guru who went 0-for-2016, ventured into Don Surber territory this week by discovering, at last, that the political pundits got Trump all wrong because they talked among themselves, rather than with voters.
Pauline Kael.

Readers of "Trump the Press" knew this last July.

But Silver has a higher profile and his admission of the obvious delights the media insiders who made this failed economist THE final word on political predictions.

Just one problem; he was wrong -- again and again. He did a mea culpa after Trump clinched the nomination -- and promptly made the same mistake in the general election.

Now he makes the same mea culpa.

From Silver's "There Really Was A Liberal Media Bubble":
It’s hard to reread this coverage without recalling Sean Trende’s essay on “unthinkability bias,” which he wrote in the wake of the Brexit vote. Just as was the case in the U.S. presidential election, voting on the referendum had split strongly along class, education and regional lines, with voters outside of London and without advanced degrees being much more likely to vote to leave the EU. The reporters covering the Brexit campaign, on the other hand, were disproportionately well-educated and principally based in London. They tended to read ambiguous signs — anything from polls to the musings of taxi drivers — as portending a Remain win, and many of them never really processed the idea that Britain could vote to leave the EU until it actually happened.
So did journalists in Washington and London make the apocryphal Pauline Kael mistake, refusing to believe that Trump or Brexit could win because nobody they knew was voting for them? That’s not quite what Trende was arguing. Instead, it’s that political experts aren’t a very diverse group and tend to place a lot of faith in the opinions of other experts and other members of the political establishment. Once a consensus view is established, it tends to reinforce itself until and unless there’s very compelling evidence for the contrary position. Social media, especially Twitter, can amplify the groupthink further. It can be an echo chamber.
Wait a second.

Silver and company are not worthy to hold Pauline Kael's hat. Having died 15 years ago, Kael is not here to defend herself for something she said in 1972, six years before Silver was born.

But I am.

Speaking to the Modern Language Association on December 28, 1972. Kael admitted that she lives among the elitists writing to the elitists in an elitist magazine -- New Yorker -- about an elitist interest in the cinema:
“I live in a rather special world. I only know one person who voted for Nixon. Where they are I don’t know. They’re outside my ken. But sometimes when I’m in a theater, I can feel them.”
She was being honest.

And come to think of it, she did not need to be inside the ken of the middle class or the silent majority. Kael was a specialty writer, and as I recall a damned fine one with wit and a fine eye for the movies.

Pauline Kael was competent.

Nate Silver is not.

I explained Kael's quote in the first chapter of "Trump the Press," which later devoted an entire chapter to Silver's botching up the nomination process.

The sequel "Trump the Establishment" details how the media botched the general election. I enjoyed chronicling Silver's election night embarrassment.

Yes, that's a plug for the books, but you can continue to follow Silver for free, or pay a few bucks and enlighten yourself. I'm inside your ken.

"Trump the Establishment" is now on Kindle.

"Trump the Establishment" is also available in paperback.

This is the sequel to "Trump the Press," which covered the nomination. The original -- "Trump the Press" -- is available on Kindle, or in paperback on Amazon.

Autographed copies are available by writing me at

Please follow me on Twitter.

Friend me on Facebook.


  1. All these guys like Rush, Silver , etc are now pretending they were in the war room with you a year ago Don. Fake battle ribbons and all. Kind of sad.
    As for PK, she believed she lived in heaven but could feel the souls who had lived in places like Poca screaming in torment elsewhere. She wrote well but she would not have respected you for any reason we're she alive today, believe me. But perhaps The best thing that happened to her was she died before the dreck deluge that movies have become flowed over the golden avenue on which she lived

  2. It will be interesting to see if Silver and co. Are on the upcoming ESPN chopping block. He didn't earn his first free agent contract.

  3. When do these guys who got it so wrong suffer the consequences of being on the wrong side of history? Who will hold them accountable and when will that be? If they worked for the Trumpmeister, they'd all be fired by now.

  4. From PJ O'Rourke in 'Give War A Chance' describing how he and others got the Sandinista election defeat so wrong:

    "That is - I blush to admit it - I accepted the results of an opinion poll taken in a country where it was illegal to hold certain opinions. You can imagine the poll-taking process: "Hello, Mr. Peasant, I'm an inquisitive and frightening stranger. God knows who I work for. Would you care to ostensibly support the dictatorship that controls every facet of your existence, or shall we put you down in favor of the UNO opposition and just tear up your ration card right here and now?""

    Now change it to "Do you want to support the Democrat, or shall we just put you down as being racist, sexist, homophobic, Islamophobic, and nearly every -ic we can imagine and use that to ruin your social life, have children taken away, and you fired for supporting that awful Donald Trump?"

    -Mikey NTH

  5. Many of these pollsters thought the Dimocrats would win the Senate and the House.

  6. One word, was, says it all.

    He thinks all that Lefty bias isn't still there.

  7. He knows what he knows, and doesn't want to know what he doesn't know. Makes for a big, even yuuuuuuuuuge, credibility gap.

  8. Oh my God. You folks are idiots. Take a statistics course...A d truly read the blogs.

    Statistics of any kind does not and never has never signified a prediction. It is a statement of derived probability.

    Nate silver does and his colleagues do an excellent job of explaining this.
    Statistics = Nuance.

    Statistics != black and whitE predictions

    Total false equivalency described here.

    Nate and his team did the absolute best job of predicting outcomes in the last election. No one else had Trump as even close to winning.

    So, before you right a blog post degrading statistical nuance, please, take the time to understand it first.


    1. Yeah. Except for Don Surber and Bill Quick. They actually got it right. Phu koff troll.

    2. LOL, Wayne. As I read that comment from Smdh, I could hear Feynman laugh.

    3. Having the best wrong predictions is not something to be proud of. Silver got the primaries consistently wrong too. Maybe he got lucky in 2012.

    4. Perhaps getting the primaries wrong should have set off a warning signal that something is not right with the models and the polling. And perhaps someone should have recalled what O'Rourke wrote about polling, the Sandinista election, and how candid the polled populace would actually be.* There is a reason why the Australian ballot is the standard.

      *School elections - "Will you vote for Popular McAsshole?"
      "Oh sure." Because self-defense; once given the chance to vote a secret - Australian - ballot Popular McAsshole learns that his votes equal him, his toadies, and a few others.

      -Mikey NTH

    5. Actually, the model always gave Trump a better probability than Nate did during the primaries. And Nate freely admitted that he was letting his disbelief that someone as... outrageous, shall we say, as Trump could win the nomination color his analysis. Which is why he had Trump as 10-times more likely to win the general as anyone else. Because he learned his lesson and simply reiterated what the model said instead of offering his opinion.

      Trump lost the popular vote by ~2 million and only won the Electoral College by ~80 thousand votes spread across 3 swing states. If that isn't the definition of an "unlikely event", then what is? And that's what a 20% probability is... "Unlikely", not "Impossible".

  9. "Statistics != black and whitE predictions
    Total false equivalency described here. "

    Hmmm, an idiot speaks and accuses others of being idiots. OK, enough with the gratuitous insults.

    1. Have you taken any statistics courses? If so, what textbooks did you use? Who was your instructor and at which college? I don't need to be any more intrusive than that. The reason I ask is because I'm actually pretty familiar with the field, so if you blow smoke, I'll know about it pretty fast.

    2. So, you claim Silver only did statistics but didn't make predictions? Here's just one quote by Silver which Don cites in Chap 40, p. 212, of "Trump the Press". Tell me why this is NOT a prediction. "If you want absurd specificity, I recently estimated Trump's chances of becoming the GOP nominee at 2 percent." Sounds like a prediction to me. How about you? Do I need to cite any more examples of Silver's predictions, or is that one good enough to shoot holes in your argument?

    1. Saying something has a 2% chance of happening means it's "very unlikely to happen", not that it "won't happen". If I roll a 100-sided die there is a 2% chance I will roll a 1 or a 2. If I roll a 1 or 2 immediately, that doesn't mean that statical probability is "wrong", it simply means that the very unlikely thing happened in a small (single) sample.... You cannot judge a statistical prediction off of a sample of 1, and if you actually know even a tenth of what you pretend to about the subject then you know that and need to stop pretending otherwise.

  10. I will go out on a limb and say that statistics calculates there is a 85% probability that the commenter "Teach" is a professor of some sort.

  11. Umm. Maybe I'm missing something, but Silver did not get it "wrong." His models showed there was something like a 20% chance of Trump to winning the last I checked (nearly 1 ine 5). And I remember something like a 50/50 chance at some point.

    Either way, he seems like a statician gathering his data from the souces available to him. He does not seem to be a political pundit to me.

    IMO, Silver is really great at what he does.

    It would've been greater if he had the resources to gather his own polls throughout the country, but he doesn't have that luxury.

    I've seen a slight hate toward him on the interwebs that I don't quite understand. Again, he's simply been using data gathered by 3rd parties. And it's not like he said there was no chance in hell, Trump would win. In fact, everytime O checked silver's sitee, there was a very significant chance.

    1. LOL. Not your claim that Der Silver "did not get it wrong" but your silly claim that "It would've been greater if he had the resources to gather his own polls throughout the country, but he doesn't have that luxury."

      He has the resources. He works for Disney. If Disney is too cheap to do it right, then to hell with him and Disney.

      I give every team in the NCAA tourney a 20% chance to win any of the games, so my bracket is perfect.

      Where do I collect my money?

    2. If people who actually make their money from polling don't find it profitable to take regular (or sometimes any) polls in most states (i.e.,any non-swing state), then what makes you think 538, ESPN, or Disney is going to decide to lose money on such a venture? Nobody "has the resources" to take the number or quality of polls Nate would prefer, at least not anyone who wants to stay in business. Sadly, (reasonably) perfect information is one of those things that everybody says they want but nobody wants to be the one to foot the bill for. So we have to make due with whatever the best the market will supply is, even if it's of middling quality at best.

  12. I believe if you go back in time a bit further, Nate the (not) Great had Trump ~1%. Please pardon me for my inability to appreciate the 'nuance' in that outstanding prediction. Outstandingly clueless, misinformed, and hillary-worshiping.

  13. British Prime Minister Benjamin Disraeli: "There are three kinds of lies: lies, damned lies, and statistics."
    - Elric

  14. So... Why is he clueless? Because he mentioned some things that you mentioned? You knew these things before it was cool, and he's poser? I make fun, but I'm seriously asking what is the point of this post other than an opportunity to plug your book? And to be quite honest, it's an effective plug because it ended up in my feed.