Nate Silver's replies are a good place to see a demonstration of Twitter users' absolute inability to read what they're replying to.
Like, he's literally just outlining scenarios based on the data he has access to, and every reply is someone (on either side!) furiously accusing him of claiming something he isn't claiming.
"Trump needs to win AZ, NV, NC and PA to have any path to victory..."

"What??? You really think he'll win AZ????"
He's just interpreting the data with no particular slant, and it makes people so mad.
Everyone gives him shit because he said Clinton had a better chance than Trump of winning in 2016, but also that a narrow Trump win was well within the scope of the polling errors, and they read that as "Clinton will definitely win".
People don't understand probability.
538 literally said that anything from a fairly comfortable win for Trump to a Biden landslide fell within the realms of possibility, but the most likely result was a narrow Biden win. That's a lot of uncertainty, but it's an unusual election.
Silver: "Biden will probably win, but it could be tight in some races."
*Biden probably wins, with some tight races*
Twitter: "Nate blows it again with his wild predictions!"
It's not Nate Silver's fault that your ape brain isn't evolved to intuitively understand large-scale probability.
This morning was so frustrating, because everyone on the left was in *despair* about the election results so far. And they were meaningless because in some places they hadn't even started counting the postal votes.
Is it any wonder Trump and his ilk are baffled by their change in fortunes? We can't extrapolate properly. We assume that almost any amount of data *must* indicate a broader trend.
If 50% of the votes are in and Trump is leading, that doesn't mean that he'll be winning by the same proportion once 100% are in because those two datasets have no probabilistic relationship to one another.
You can follow @smolrobots.
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.