AL-Gravis/Big League: Moore +4

[The Other Castro]

Roy Moore - 49%
Doug Jones - 45%
Uncertain - 6%

Trump approval: 55-42 (+13)

You must be logged in to read this quote.

https://bigleaguepolitics.com/big-league-gravis-alabama-poll-moore-surges-ahead-jones-49-45-special-senate-election-race/

But:

Gravis Marketing‏ @GravisMarketing 5 minutes ago
Note if we used the previous turnout model jones would be ahead by 2%.  in light of the consistent movement in the raw data, the voting model was shifted from plus-15 Republican to plus-23 Republican.

https://twitter.com/GravisMarketing/status/939308265518911489

[junior chįmp]

Alabama is despicable

[Devout Centrist]

“In light of movement in the raw data”

Translated:

“We’re herding”

[GlobeSoc]

“In light of movement in the raw data”

Translated:

“We’re herding”

[Doimper]

Gravis Marketing‏ @GravisMarketing 5 minutes ago
Note if we used the previous turnout model jones would be ahead by 2%.  in light of the consistent movement in the raw data, the voting model was shifted from plus-15 Republican to plus-23 Republican.

https://twitter.com/GravisMarketing/status/939308265518911489

LOL

Has any other pollster outright admitted to herding before?

[Sorenroy]

Gravis Marketing‏ @GravisMarketing 5 minutes ago
Note if we used the previous turnout model jones would be ahead by 2%.  in light of the consistent movement in the raw data, the voting model was shifted from plus-15 Republican to plus-23 Republican.

https://twitter.com/GravisMarketing/status/939308265518911489

Even with the original model, there is a two point shift towards Moore. Not exactly what you want to see this close to the election.

[100% pro-life no matter what]

Do Alabama Hispanics and Asians always vote identically to Alabama whites or is this just a case of the white vote dropping to the Asian/Hispanic level (~62% Republican)?

It seems that Moore has also closed the age gap greatly (going from ~22% of the 18-29 vote in the last poll to nearly 40% this time).  Maybe some of the unenthusiastic young Republicans are coming home?

[Brittain33]

I'm so confused by what they said. "Our survey had Jones leading, but we decided Republican turnout will be much higher than we though last week so we gave some points to Moore." I don't have high hopes for a Democrat in Alabama but this just seems off, especially in light of SurveyMonkey.

[Former Kentuckian]

I can't freaking wait until this Hell-spawn of an election is over, God.

[Sestak]

-> "How do we prevent them from accusing us of herding?"

-> "By telling them that we're herding."

[Devout Centrist]

Ironically, both the unweighted Survey Monkey poll and the unweighted Gravis numbers show Jones leading by 2. I applaud Gravis for hedging their bets in such an obvious fashion.

BIG SHIFT!!!!!!!

[Tender Branson]

I still have a feeling that Doug Jones will narrowly win in an upset, despite these recent polls.

[MT Treasurer]

I still have a feeling that Doug Jones will narrowly win in an upset, despite these recent polls.

[Since I'm the mad scientist proclaimed by myself]

“In light of movement in the raw data”

Translated:

“We’re herding”

[AtorBoltox]

Moore will probably win but this poll is still utter junk

[Co-Chair Bagel23]

Always knew he would win. People mocked me when I said Handel would win by 53-47 or 54-46, and she won 52-48. Why, because I may be stupid, but not stupid enough to change my mind for little things and hop on bandwagons for small things. I stick with my predictions unless something drastic happens, and I sometimes end up right. Though I hope I am wrong on this one. Moore will win around 53-45, and do not be surprised if it is by even more.

[Eraserhead]

Moore leading in Gravis is great news for Jones!

[BundouYMB]

LOLOLOLOL if Gravis didn't herd they would have been the gold standard. Too bad the right-wing hacks who run Gravis cared more about pushing a narrative!

[Holmes]

Gravis should've stayed with their model!

[Landslide Lyndon]

Gravis should've stayed with their model!

Yeah, seriously WTF. Gravis would have nailed a second straight election if they just trusted themselves.

[Pericles]

Gravis should've stayed with their model!

Yeah, seriously WTF. Gravis would have nailed a second straight election if they just trusted themselves.

[Devout Centrist]

That’s what you get for herding, Gravis

[Sestak]

That’s what you get for herding, Gravis

[Pollster]

An incredibly important lesson that all pollsters must learn: don't change your methodology until you have an independent, credible reason to do so.

[McGovernForPrez]

I love how people were predicting their downfall before it even happened. They're probably gonna go back to their old methodology after being screwed like that so I don't think it matters that their official result was wrong. They don't get to take the credit, but it still means they've got a lot of promise as a polling firm.