Thread by @MysteryPollster

Hard to believe it's been 14 years since I started the MysteryPollster blog and posted this widely shared exit poll post
mysterypollster.com/main/2004/11/e…
While much has changed, and mid-day exits no longer leak, the main thrust holds up well...
[THREAD]

To wit: If you think there's some secret stash of inside data showing who's *really* winning what, think again. Even at or just after poll closing, exit polls won't tell you more (or with any greater certainty) than pre-election polls are telling you right now

Also, the ABC-CBS-CNN-NBC "exit" polls now consist of a much larger portion of pre-election polling (to gauge early voting) than ever before…

My advice: The best way to stay on top of who's going to win is to watch the @UpshotNYT estimates as the vote is counted, needle and all.
nytimes.com/interactive/20…

And one of the networks paying for the exit poll is telling its reporters to "embrace not knowing". Good idea!
politico.com/story/2018/11/…

As for what exit polls can tell you about "why" people voted the way they did, here’s some worthy caution from @BrendanNyhan

https://twitter.com/BrendanNyhan/status/1059796617589592065

Finally, for the "who voted" part, be careful on that too. NPR's @asmamk has a really nice summary on that point, part of a longer piece worth reading on the exit poll:
npr.org/2018/11/05/662… And…

This piece from @Catalist_US's @yghitza makes the point even more powerfully, especially on the share of voters with college degrees. Even looking backward, estimates of who voted vary widely, and exit polls appear far off the mark.
medium.com/@yghitza_48326…
-30-