I think we both have analysis showing it to be very un-anomalous and expected. This guy...

🫡 That was fun! Lmk next time you have a hundred csv files and an interesting question.

Thanks! 🫡

I don't think there's any precedent for that big of a reverse swing.

We're going to have dig up some prior recount data though, since that seems where it's headed.

"Done" might be an overstatement. The last few percent can be a little unusual and start to swing the other way, but as long her share of the remainder is above 50% she wins, which I think is very likely.

Thanks! I had fun doing analysis, I'm glad some people found it interesting.

☝️ Right, it's not really reversion to the mean so much as small highly variable drops that happen at the very end that don't have a huge impact on the overall share of that last 17%.

Scraped all the daily reports from king county elections (which unfortunately only goes back to 2017). Then grabbed mayor, council, and city attorney races where there was a pretty clear progressive vs centrist matchup, which was most of them.

Yeah, seems like that can happen. Maybe they start counting some very late mail-in ballots here. But the averaging over increases and decreases it still seems like overall flat or better in the last 17%.

I pulled a bunch of other daily drop data from other races and put it on a scale of percent counted here: bsky.app/profile/dcjo...

Still, every election is different, and this is incredibly close, so do check that your ballot was actually counted.
bsky.app/profile/wils...

Or to look at in another way, the share of ballots the progressive takes on the last 17% counted has always met or exceeded their share on the second to last 17% in the 21 races I looked at.