The Left Invests Totally in Nate Silver-Type Prognostication
Sep 6, 2017
RUSH: One of my favorite reads during the football season is Tuesday Morning Quarterback by Gregg Easterbrook. It originally appeared on what was called Page 2 on ESPN. And Easterbrook got fired for some innocuous, stupid reason like I had to resign for at ESPN. I forget what it was, but it was ridiculous. He moved on, and his piece ran then at the New York Times for a year or two.
Now it’s at the Weekly Standard. It’s every Tuesday and it’s about the NFL, and a little about college football. But he interposes every week, like he’s a big outer space buff, big astronomy buff, so he puts in the latest news on those things. And he’s a huge data analyst. So he throws in political observations and cultural observations in the middle of a football column. I can’t imagine how long it takes him to write this thing. I mean, it’s omnivorous, it’s voluminous, it’s very lengthy. And in the premier issue or column yesterday for the opening of the NFL season, what he always focuses on is predictions.
You know, every sports personality, every website, every network has their people that come on and they predict who’s gonna win opening weekend, who’s gonna be in the Super Bowl, and he thinks it’s all bunk. Easterbrook thinks that it’s nothing but wild guesses. He offers as evidence all of the Big Data analysis that was used to assure everybody that Hillary Clinton not only was gonna win, but that Donald Trump never even had a chance.
I read through this last night and I’ve got some excerpts I want to share with you. And it added another layer of understanding to me of why these people are so deranged. And then when I watched the opening episode of American Horror Story last night after having read Tuesday Morning Quarterback, I mean, it was like he predicted the opening scene of this premier episode. ‘Cause it was — I can’t describe it. You have to watch it. It’s a bunch of watching election returns on Fox, and they are hysterical.
They’re crying, they’re hugging each other, they think they’re gonna die. A gay couple think Trump is gonna come look for ’em and kill ’em. They’re cursing Nate Silver, and they’re cursing CNN, and they’re cursing the Huffington Post for lying to ’em. And let me share with you just a little of what Easterbrook wrote. “Here’s a fast forward version of FiveThirtyEight –” remember, his purpose here is to debunk all of these Super Bowl predictions as nobody knows anything. It’s impossible to do, it’s a waste of time. It’s fun, but nobody can in any way possibly know. And he’s exactly right about it.
“Here’s a fast forward version of FiveThirtyEight –” that’s Nate Silver’s website. Remember, he was at the New York Times, and somebody offered him more money. ESPN, incidentally. And FiveThirtyEight’s where he went and that’s where he does his work now.
“Here’s a fast-forward version of Five Thirty Eight predictions about the 2016 presidential vote. In June 2016, Silver said Hillary had an ‘80.2 percent’ likelihood of winning — not 80 percent, 80.2 percent.” You see, you put 80.2 and it conveys sophistication. It conveys super knowledge. It conveys authenticity. And then he says ESPN’s computer predicts the Steelers are gonna win 10.1 games. He says, what’s the nine-tenths of a game they’re gonna lose? What do you mean, 10.1 games? There’s no such thing.
“By July 24,” just a month later. “Silver had Clinton down to a ‘58.3 percent’ chance. Two days later on July 26, she was down to ‘53.3 percent.’ A week later Clinton had zoomed to ‘79.9 percent’ likelihood — quite a rapid rise, perhaps owing to one of Five Thirty Eight’s infamous ‘adjustments.’ By late August, Silver had Clinton at ‘87.7 percent’ likely to win, by mid-September at ‘66.8 percent,’ and by mid-October at ‘82.2 percent.’ The day before the vote, the website said Clinton had a ‘68.5 percent’ win probability; the morning of the election, it was ‘71.4 percent.'”
And there were more than that. That’s just a fast forward version of the all-over-the-place FiveThirtyEight predictions. He says, “This is a gloss, but you get the idea. Five Thirty Eight had Clinton’s odds rising and falling on a nearly daily basis, as if the election were the NASDAQ — though it is extremely difficult to believe there were bona-fide alterations in voter sentiment.”
Exactly. People don’t change their minds that often. What was this? I’ve even made that point here on presidential polling. People just don’t change their minds this fast. I’ve made the point that a presidential poll in July of the summer before the election is worthless! It’s meaningless. But you know the people in the media live and die by them because they are the source of news. Polling data and now big analytic data analysis has become a source of news! And so the more of it, the better. Why is Hillary dropping from 68 percent likelihood to 49%? She didn’t. ‘Cause she was never at 68 because there’s nobody who could know.
But Silver wasn’t alone. “The New York Times, cable news channels, the Huffington Post, and other organizations offered an almost day-by-day series of presidential election speculations dressed up in graphics that made them seem to be the scientific discoveries of great researchers, rather than what they were: guesses, however educated.
“In mid-July, the Times had Clinton with a 76 percent chance of victory. In six days she fell to 69 percent. A month later she was at 87 percent in the morning, rising to 88 percent that afternoon. (Millions of voters must have said at lunch). By mid-September, the Times had Clinton down to 79 percent, then down to 74 percent in mid-October. For the final week before the vote, the newspaper put Clinton at 84 percent.
“The point is not that all these forecasts were wrong. Practically everyone was incorrect about the November 8th election,” except Rush Limbaugh. “The point is that the prediction numbers on FiveThirtyEight and similar places appeared to be the result of deep dives into data,” no they weren’t “yet were darts-on-a-board exercises gussied up with charts, graphs, and squiggly lines to make them professorial.”
Even on election night, the New York Times had their little percentage dial there, and at six o’clock Hillary Clinton had like an 80% chance of winning. At 6:30, she had a 78% chance of winning. At seven o’clock she’s at a 75% chance. What is this? Votes had already been cast in most places. How can this be? What is this? And it was gobbledygook. I think Nate Silver had her at a 90% chance of winning at some point during election night.
“Did the public impression of a scientifically-proven, certain Clinton walkover convince many people to not bother voting? We’ll never know for sure, but this seems plausible among the data-conscious young, whose miserable turnout — 46 percent of those to age 18 to 29 voted in the 2016 presidential race, versus 71 percent of seniors.”
Easterbrook theorizes here that the low turnout, 18 to 29, was because these kids, these Millennials love Big Data, they just are into it, and the Big Data was telling them that Hillary was gonna win this thing in a walkover. It was gonna be a smear. It was gonna be Trump barely gonna get 20%, Hillary was just gonna cream everybody. So why vote? I mean, it’s already in the bag.
“The Princeton Election Consortium, which is only loosely associated with Princeton … said just before the vote that Clinton was ‘more than 95 percent’ assured of victory, with an ‘83.5 percent’ chance Clinton would carry Wisconsin, which of course she lost. These and other faux-precise predictions could have sounded like assurance there was no need to get out and vote.”
That’s his theory. And it may be in the case of some people. “The situation became so ridiculous that the Huffington Post, which shortly before the vote called Clinton 98 percent likely to win, angrily denounced FiveThirtyEight merely for Silver’s caution that a Trump win was not out of the question.”
Silver, in every projection that showed Hillary even at 90%, had this little proviso that there’s still a pathway for Trump to win, and that so ticked ’em off at the Huffing and Puffington Post that they posted an angry denouncement of Silver for even alluding to the possibility. Now we know what happened.
And in fact, there was a story in the New York Times on November 10th shortly after the election: “How Data Failed Us in Calling an Election — It was a rough night for number crunchers. And for the faith that people in every field — business, politics, sports and academia — have increasingly placed in the power of data. Donald J. Trump’s victory ran counter to almost every major forecast,” except for Rush Limbaugh’s “undercutting the belief that analyzing reams of data can accurately predict events.”
Dr. Pradeep Mutalik, a research scientist at the Yale Center for Medical Informatics. You know what I’m gonna do? The next time I need to go to the hospital, “Take me to the informatics division.” The what? “Yeah, I want to go where they do all the Big Data crunching. That’s where the answers are.” The Yale Center for Medical Informatics. Not information, but informatics. This guy’s name, Dr. Pradeep Mutalik, said, “It’s the overselling of precision.”
“Virtually all the major vote forecasters, including Nate Silver’s FiveThirtyEight site, The New York Times Upshot and the Princeton Election Consortium, put Mrs. Clinton’s chances of winning in the 70 to 99 percent range.”
I have made the point countless times that one of the biggest problems in politics is that people believe the media. Republicans too. They believe all of this preelection stuff. They believe these polls. But the people on the left, they live and die by Nate Silver, folks. They live and die by him. And they bought it hook, line, and sinker.
They went into election night thinking it didn’t matter, no matter what else, the size of Trump’s crowds, none of it mattered because they believed all the polls. They believed all this data. They believed all this expertise, which Easterbrook is right about. It’s nothing but really dressed-up wild guesses. And that’s one of the reasons that I believe now that they are still not normal. I think they still are insane and crazy. They trust the media implicitly.
That’s why when I come along and question anything about it, I am public enemy number one, hated and despised. They trust it. It’s gospel. New York Times is gospel. Nate Silver is gospel. And they got it not just wrong, they were promising the left that Hillary wasn’t just gonna win, that Trump was gonna be so embarrassed that he wouldn’t even show up and concede in person. That’s how humiliating the defeat was gonna be. Well, we know what happened.
Now they think the polling was still right but the Russians stole the election from them. This is how this stuff has a life. They will believe absolute weirdest conspiracy theories because they believe that what are wild guesses are actually advanced scientific truth, so they invest in it totally. They invest emotionally — this is key. They emotionally invest everything they’ve got in this stuff. And they can’t handle the consequence when they are betrayed, and they’re still trying to come to grips with it. I think it’s a big factor. Whether it suppressed voter turnout or not, who knows.