The story is pretty simple.

On election night, Fox News called Arizona very early. This surprised pretty much everyone. The Fox News team had improved their election model in 2020, and was confident their call was correct, even if unconventional.
Controversially, however, they called it for Joe Biden. Arizona has historically been a red state, and the polls (and final result) were close there. Nobody had a really strong expectation it was gonna go for Biden until Fox News just.... called it.
This hurled Trump surrogates into a desperate place. If they lost *Arizona*, that was a bad sign for the night. More importantly, it happened before midnight: it humiliated the President while millions were still watching.
The Trump campaign immediately turned on a dime. They pivoted from "Look, we're crushing it with Hispanics! Massive turnout! Incredible performance!" to "The election was stolen."

Who stole the election changed over time. Fox News. Some rando in Atlanta. Poll workers. China.
The conspiratorial delusion of the stolen election grew, but it's rageful trigger came in that innocuous decision to call Arizona early. For Trump supporters, that call "changed the night." It was the turning point. It is somehow deeply significant.
So this week, Fox News fired all the people involved in that decision.
Now, let's look at this story from another angle.

There was less true uncertainty about who would win at 11:20 PM on election night than in any election ever, because sooooo many votes had been cast early.
While counts of these votes were not announced, early indicators based on turnout by voter registration and polls of who showed up, as well as differences in pre-election rhetoric, strongly suggested early voters would swing for Biden.
Meanwhile, for all that people like to crap on polling, the reality is that the predictive models for election night as returns come in really are getting quite good. They're imperfect of coruse, but they're improving all the time.
As a result, Fox News' team was very well positioned to have a good idea of what a "Trump win" would need to look like, in terms of how day-of results were going to need to look for Trump to offset the likely split in early and absentee votes.
So from this perspective, Fox News' call was a *triumph* of good analysis: they correctly called a close race even though the number of votes still not counted massively exceeded the expected margin of victory.

Analysts in finance get paid lots of money for that kind of call!
Having a rational, explainable, reproducible model which produces correct out-of-the-money calls is an extremely valuable thing. As a professional forecaster in a different business, this is a very exciting and valuable thing to see in a model.
Unfortunately, the model's call was politically inconvenient. Trump's team was counting on night-of calls swinging their way in order to make similar arguments as were made in 2000: inject enough uncertainty into the public discourse and they hoped to influence the outcome.
So the Trump team, and Trump supporters around the nation, threw a tantrum.
Ultimately, what happened was:
1) Fox News recruited high-quality analytic talent into the conservative movement
2) That talent delivered, providing accurate results faster than competitors
3) Those results didn't fit "The Narrative" that the Trumpist echo chamber wanted
4) Fired
Conservatives JUSTIFIABLY complain about mainstream media biases. Because they're real and influential.

But our media critics really need to pay attention to stuff like this. It's open hostility on even *irrelevant* truth claims.
Because guess what?

It didn't matter what Fox called! The results weren't influenced by that!

But in the media-and-commentary-addicted mind of politics-overdosed Trump supporters.... it really mattered what The Media said.
It was *deeply significant* to these folks that their false view of the world was not publicly ratified and legitimized, even though that view was *false*.
It was more important to these folks that the call made fit their narrative than that it in fact be true or honest.
Now, look. A few years back I was looking for funding for a project. I approached a funder for it. It was a data-intensive project, but with outcomes likely to be friendly to conservatives. The funder was conservative.
The funder said, wow, this proposal looks really interesting...

.... but we don't really fund "data stuff." We're more interested in commentary. Do you want to repackage this as some op-eds?
(same project was pitched to a second conservative funder who much more reasonably denied the funding pitch on the grounds their funding program had a specific mission related to talent development; that was fine)
Generally, I find that conservative publications are often confused when you ask, "So, how should I format my graphs for your publication?" because they just are totally unused to quantitative arguments being made in articles.
A serious problem in the conservative media system is that the only expertise that gets respected is legalese: rhetorical argument, appeal to precedent, citation of others, moral justification..... *evidence* is treated with hostility.
So this makes it even worse FWIW. They just decided to shoot the messengers apparently. https://twitter.com/EzraG/status/1352615463604379650
You can follow @lymanstoneky.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled:

By continuing to use the site, you are consenting to the use of cookies as explained in our Cookie Policy to improve your experience.