Since @FiveThirtyEight released its model today, I can finally tweet: I think it’s weird that so many people seem to blame 538, specifically, for failing to predict Trump’s win. They started seriously warning of a popular/electoral split in September 2016, before other observers.
-
-
Replying to @yayitsrob @FiveThirtyEight
On the one hand, others were way worse. On the other hand, the 538 presentation was terrible and misleading. Visually, it suggested a landslide (because people are used to polls, not probability models) and many crucial details were in footnotes (I read; how many else did?).
2 replies 2 retweets 23 likes -
I'm not a 538 defender but it is not their fault if people refuse to listen to nuanced analysis.
1 reply 0 retweets 2 likes -
That is not a defense if you are that high-profile in the business of communicating to the public and where the consequences are so important. Others are worse (very true!) is also not a defense. So so so many people made consequential decisions based on refreshing that page.
1 reply 0 retweets 0 likes -
I dunno. I am very stupid and I remember 538 being the most alarmist about Trump's chances, especially in the two weeks before the election.
1 reply 0 retweets 1 like
You cannot override the visual with tweets or footnotes. That's why I rail against the endless beach pictures even in good COVID stories. In stats, where people don't have familiarity with probability models, the 2016 visuals were catastrophic. I have endless threads about it.
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.