Ann Coulter's Bad Scholarship

Ann Coulter tweeted:

Professor whose statistical model predicted every election since 1912: Odds Of President Trump Range Btwn 97% & 99%-http://bit.ly/1p63RMW

After my previous positive reviews of her book scholarship, I wanted to highlight how atrocious this is. Let's look over the article:

Political Science Professor: Odds Of President Trump Range BETWEEN 97% AND 99%

The model has been correct for every election since 1912 except for the 1960 election

Ann said "every election". Did she even read the article? What a travesty.

Specifically, Norpoth predicts that Trump has a 97 percent chance of beating Hillary Clinton and a 99 percent chance of beating Bernie Sanders.

The predictions assume Trump will actually become the 2016 presidential nominee of the Republican Party.

So it doesn't predict either primary. It only predicts Trump is 97-99% to become president if you throw in the big assumption that he's literally 100% likely to win the Republican primary.

So that's two major factual errors in Ann's tweet.

Besides getting the basic facts wrong, twice, there's also the issue that the article and prediction model are utter crap.

“When I started out with this kind of display a few months ago, I thought it was sort of a joke,” the professor told the alumni audience

You know what would have been impressive? If the prediction model was published in 1911.

Instead it was worked out a few months ago and has never actually predicted anything? It's really easy to "predict" past data. It's called back-fitting and it's well known. Making a formula to fit past data is completely different than making successful predictions about the future.

(That it was back-fitting, not prediction, was predictable to me before I even clicked the article. Ann should have known better even if she literally didn't read a single word of the article.)

Norpoth, a 1974 University of Michigan Ph.D. recipient who specializes in electoral behavior alignment, said his crystal ball also shows a 61-percent chance that the Republican nominee — Trump or not — will win the 2016 presidential election.

Wait what? This is pretty incoherent. These numbers do not make sense. For this math to add up – around 98% chance for Trump to win if he's the nominee, and 61% chance for any Republican to win – requires Trump to have only around a 60% chance to be the nominee (if the other Republican candidates are somehow all around 0% likely to win the general election) or less.

I also checked out the Daily Caller's source:

Political science professor forecasts Trump as general election winner

“You think ‘This is crazy. How can anything come up with something like that?’ ” Norpoth said “But that’s exactly the kind of equation I used to predict Bill Clinton winning in ‘96, that I used to predict that George Bush would win in 2004, and, as you remember four years ago, that Obama would win in 2012.”

Note the wording, "the kind of equation". So he made up a new equation just now. He's made up other equations in the past. He keeps changing them each time, rather than re-using an equation that's ever predicted anything.

In contrast, Norpoth forecasted that a hypothetical presidential race with Ted Cruz or Marco Rubio on the Republican ticket would be a much closer race. The results showed Clinton with a 55 percent chance of winning the race against Cruz or Rubio with a 0.3 percent lead in the popular vote.

So Trump needs to have a very low chance to win the GOP primary for the math to work out. Meanwhile the prediction model saying he'll win the general election is based on him doing so well in the primaries! This is all a bunch of contradictory nonsense.

---

And Ann Coulter is promoting this utter nonsense on Twitter while making factual errors. This fits her recent pattern of saying anything – even stupid and dishonest things – that are on Trump's side. :(