A thread for discussion of the count in the Alabama Senate race, where the face of modern Republicanism, Roy Moore, does battle with Democratic candidate Doug Jones for the seat vacated by Jeff Sessions’ appointment to the Trump administration as Attorney-General.
Author: William Bowe
William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.
View all posts by William Bowe
Thank you William. I hope Jones gets up to really take the sh*t up to Trump.
Jones powering ahead on NYT live
Watching Alabama senate election with keen interest. Never been a more exciting time. Thank you 2017.
Live Estimates of the Final Vote
Updated at 9:17:20 PM ET
Our estimates are based on the results reported so far, the results of previous elections and demographic data. Unlike reported results, they attempt to account for precincts that have not yet reported their votes. The shaded area in the gauge below shows the range of our forecasts.
Right now, our most likely estimates span Jones +15 to Moore +6. The more we know, the narrower our range will be.
Our best guess is that Mr. Jones has an advantage, but paths to victory remain for Mr. Moore.
https://www.nytimes.com/elections/results/alabama-senate-special-election-roy-moore-doug-jones?hp&action=click&pgtype=Homepage&clickSource=story-heading&module=b-lede-package-region®ion=top-news&WT.nav=top-news#eln-forecast-section
Nate_Cohn: And in terms of turnout, both white, rural, GOP counties fell *short* of our estimates, with Randolph–the least educated–falling well short. For now, our model assumes these turnout errors are uniform. But if they’re not, this could be an ominous sign for Moore.
This perhaps the reason NYT is giving Jones a chance…..”Unlike reported results, they attempt to account for precincts that have not yet reported their votes. The shaded area in the gauge below shows the range of our forecasts.”………..
NYT pretty accurate with the Virginia vote… Let’s hope for the same with Alabama today!
HARRY ENTEN 9:23 PM
Following up on that Nate Cohn tweet we posted below: I want to point out that Fayette County is all in. Moore won it by 50 points — well ahead of his benchmark there. Here’s the twist, though: There were less than 5,000 votes cast there. In 2014, there were nearly 7,000 votes cast there. In other words, turnout is way down from 2014 in a heavy pro-Moore area.
Huge surge to Moore on the Upshot’s model just now. Now at 86%.
There must be some BIG Dem areas to come in then.
So its Moore winning rural white counties by a larger margin than needed, but turnout is way down?
Ie, big numbers of voters in democrat heavy counties could get Jones across the line?
Jones crushing it now on NYT. Looks like hardcore Repugs have turned up and so-sos have taken a rain-check.
Surge to JONES, sorry!
Looking like a trend of Moore’s supporters choosing to stay home rather than vote.
NYT algorithm has heroic call of Jones 81% likely to win
Presumaby for this reason (Harry Enten at FiveThirtyEight):
Yes. Only 37% of the vote in Jefferson County (Birmingham) is counted.
JonahNRO: Now Hannity’s explaining that Trump didn’t **really** campaign for Moore.
From David Wasserman at the 538 blog:
At the moment I am seeing Jones behind by 35K -40K votes..He needs some big areas to cover that.
Upshot model now back to where it was before …
sonar @ #18 Wednesday, December 13th, 2017 – 12:33 pm
Yes, he should net at least +30K in Birmingham, and +10K in Madison. That will get him there, without even tapping Montgomery or Mobile.
krassenstein: Roy Moore is Up by 6% with 56% of Precincts reporting.
Good news is that there are only 140 Heavy Favored Roy Moore Precincts remaining to count, but 400 Heavy Favored Doug Jones Precincts to count.
Hope is alive!
#AlabamaSenateElection
Upshot’s model looking precarious, but they remain excited about black turnout at FiveThirtyEight.
Are the computers dumber in the big cities? Why do they take any longer?
Is this predictive thing on the New York Times the same thing that was so accurate at the election?
It Uses Green’s trick of comparing individual booths to previous results to get an idea of the swing.
Jones currently 66% with 62% counted. Better than the polls were predicting for him.
https://www.nytimes.com/elections/results/alabama-senate-special-election-roy-moore-doug-jones
EDIT: 66% “chance of winning”.
AlexConant: The polls have been closed for 90 minutes & GOP could actually lose Senate seat. In ALABAMA. No matter what happens next, hard to overstate what a huge political disaster this is for Moore-apologists like Trump, let alone Moore-champions like Bannon.
antonbruckner – depending on turnout, urban precincts can often have as many votes to count as entire rural counties – not just in Alabama but across the US.
Jimmy – but I thought computers count them. It’s almost automatic. Or do they have to be fed into a computer?
Da, Russkiye komp’yutery.
anton – I may be wrong, but I don’t think Alabama uses electronic voting.
Jones really does seem to be crushing it in Birmingham compared with last time. Hard to believe that’s not on the back of a big jump in turn-over.
Jimmy – they definitely do. There was a court case about what they do with the electronic data (they’re gonna dump it). But I suspect that someone has to scan the “votes” (maybe).
Love that NY Times thing.
Currently estimating Jones 50.9% Moore 47.6%. Expecting Jones to lead the remaining count by 14.5
Jones 70% chance of winning.
From 538
Guytaur – surely the estimated remaining vote is a guess based upon the last election. If turn-out is up, maybe Jones will romp home?
A11
Yes conservative estimates based on previous elections.
Looks like it’s from the NY Times. Are they sharing?
https://www.nytimes.com/elections/results/alabama-senate-special-election-roy-moore-doug-jones
Gap closing rapido
Question
Thte two most accutare models are agreeing on projections from raw data
ForecasterEnten: All the talk about black voters not turning out… If Jones wins it, it’ll be because black voters turned out.
Jones is drawing closer now…margin down from nearly 70k to less than 30k…nail-biting
Guytaur,
It looks like the NY Times graphic.
anton – thanks for the info re voting machines.
I wonder if Jeff Sessions thinks it was all worth it? Despised as Attorney General by Trump and causing his party to maybe lose the Alabama Senate seat by his going.
Jones locking onto target!
The only large numbers of votes remaining to be counted are urban votes for the Dems. This race is looking less and less about whether Jones will win (I think he already has), but rather how much Jones will win by.
Argh! Jones down to 64% chance. Excruciating! 🙂
anton – I may be wrong, but I don’t think Alabama uses electronic voting.
Oh noes! Not the ‘Hanging Chads’!
Question
Sorry you are right. 538 showing the Times result.