VA-GOV 2017: Utter Panic and Doom (General election: Nov 7th)

[Attorney General & PPT Dwarven Dragon]

Updated the Sabato map to include races only considered competitive by DDHQ/Cook. This is what I'll be using for reference on election night.

[Attorney General & PPT Dwarven Dragon]

Things are certainly looking up for Gillespie. Northam is at his smallest RCP average lead of the entire campaign - 2.8%. A simple average of the last 5 polls is either Northam +3.4 or Northam +3.6 depending on whether you skip over Hampton and Quinnipiac. All of these numbers are not good for Northam because they are not big enough to accommodate a polling error of ~4% too D, which is what we saw in 2013. The race stays in my Lean D column for now as I try not to issue a Toss-Up rating this close to the election. But if one of these averages slips below Northam +2 (meaning it cannot even withstand half of the 2013 error), it may merit a move to Lean R.

[Attorney General & PPT Dwarven Dragon]

I don't see the rationale behind moving a race to lean R if the Democrat has an average of +2.

1. I expect that the polling averages will be off in favor of Gillespie, just as they were off in favor of Cuccinelli in 2013, though not sure exactly how much error will exist.

2. Late deciders broke R in 2013, 2014, and 2016.

3. The trend in the last few polls of any race tends to be exacerbated on election day. So far, the late trend is against Northam, although I acknowledge that could change.

[Attorney General & PPT Dwarven Dragon]

Everyone's acting like Gillespie's gonna overperform the polls again massively or something and treating it as if it's destined to be within 3 points

The polling average doesn't need to be off by anywhere near as it was in 2014 for Gillespie to win though. The 2014 averages were roughly Warner +9, this year the average is somewhere between Northam +2.75 and Northam +4 (several "fair" calculations are possible). As I said upthread, under most calculations, Northam can no longer withstand the entire 2013 polling error of 4%. That being said, I understand that betting on polling errors is a poor strategy at times, and will keep the race in the Lean D category as long as the average stays above half the 2013 polling error (2%).

[Attorney General & PPT Dwarven Dragon]

Everyone's acting like Gillespie's gonna overperform the polls again massively or something and treating it as if it's destined to be within 3 points

The polling average doesn't need to be off by anywhere near as it was in 2014 for Gillespie to win though. The 2014 averages were roughly Warner +9, this year the average is somewhere between Northam +2.75 and Northam +4 (several "fair" calculations are possible). As I said upthread, under most calculations, Northam can no longer withstand the entire 2013 polling error of 4%. That being said, I understand that betting on polling errors is a poor strategy at times, and will keep the race in the Lean D category as long as the average stays above half the 2013 polling error (2%).

if you take out the two wack polls with Gillespie up, Northam's lead is over 7 points. Those are the ones screwing the average. Hence, expecting Gillespie to outperform an average like that is ridiculous. I can't wait til this race is over so we can stop this stupid notion that only Republicans overperform averages in off years.

You can take out Hampton, but if you do that, it's only fair that you take out Quinnipiac as well. The polls that show Gillespie +1-+2 are not really outlandish and should be part of any fair calculation, though of course it's fair to use "the polling company's" three-way numbers (Tie) instead of their two-way numbers (Gillespie +2), but even doing that keeps it within that Northam +2.75 to Northam +4 range in a five poll or six poll average.

About the Hampton poll: Just wait like 5 days before they come out with their final poll with Northam +5 to cover their ass. They did this in 2016 and they'll do it again.

If that happens, I will of course take that into account.

[Attorney General & PPT Dwarven Dragon]

A new Quinnipiac Utter Junk poll coming out today. Sort of bummed they didn't go out into the field this weekend.

FTFY

[Attorney General & PPT Dwarven Dragon]

New Poll from the world of non-junk: Northam +6: https://blumenthalresearchdaily.weebly.com/

[Attorney General & PPT Dwarven Dragon]

Yeah, there's nothing wrong with the Northam ad, at least not these days. Gillespie can run the MS-13 ads, Northam can run his "Latino Fear" ads, and voters can decide whose scare-mongering is more accurate on election day. The era of conciliatory, positive campaigning is over.

[Attorney General & PPT Dwarven Dragon]

Ads like this are a fact of life now. In most cases, the American people don't care about a candidate's morals anymore, as long as they preach the "right stuff" on policy. While the dems are quicker to point out moral travesties in their party then republicans are, in the average case, neither major party cares about moral travesties on their side, whether it comes in the form of nasty campaign ads, affairs, or more recently, outright physical assaults of reporters. 90% of Republicans voted for Trump despite Access Hollywood, and 90% of Democrats voted for Clinton despite her whole list of scandals. That shows that the "holier than thou" attitude isn't worth it. It may be nice to think about, but it's not suited for today's political world.

As I said above, let Northam and his SuperPacs run their smears, and let Gillespie and his SuperPacs run theirs, and voters can decide who should be feared more on election day. This "outrage" over negative ads existing is nothing more than partisan hackery - dems would never criticize a democratic ad, and republicans would never criticize a republican ad. Let's just resolve to quit complaining about the tone and treat it as a fact of life.

[Attorney General & PPT Dwarven Dragon]

No one seems to remember that a white supremacist ran over someone with a car and killed them earlier this year. Gillespie running ads supporting the confederate monuments does nothing to discourage that people who created violence in Charlottesville and he has clearly tried to get those sort of people on his side.

Northam is up anywhere from 6% to 17% (no, I'm not calling Quinnipiac a fake poll like Trump would), so of course Gillespie would be desperate enough to try and get just anyone on his side.

Again your the biggest left-wing hack on this site,NO CANDIDATE IS WINNING BY MORE 3 POINTS.

Sure thing Mr.Unskewer...

[Attorney General & PPT Dwarven Dragon]

Good news for Northam: Today's wash. post poll pushes Monmouth out of the 5 poll average. The last five polls, skipping Hampton and Quinnipiac, are Northam +5,  Northam +6, Gillespie +2, Northam +7,  and Northam +7, which averages out to a fairly comfortable Northam +4.6.

[Attorney General & PPT Dwarven Dragon]

Fox News (i.e. take it with a grain of salt) is reporting that the Latino Voices ad is being pulled.

[Attorney General & PPT Dwarven Dragon]

Regional crosstabs suggest a large advantage for Northam, unfortunately.

[Attorney General & PPT Dwarven Dragon]

Long Sabato article for those interested: http://www.centerforpolitics.org/crystalball/articles/signs-and-portents/

[Attorney General & PPT Dwarven Dragon]

Well, uh, all hail Gov. Northam, I guess.

[Attorney General & PPT Dwarven Dragon]

Last 5 Polls (non-junk):

Suffolk: N+4
WaPo: N+5
BRD: N+6
Polling Company: G+2
Christopher Newport Univ. : N+7

Average: N +4.0 (Down from N+ 4.6 before Suffolk Poll)

[Attorney General & PPT Dwarven Dragon]

Last 5 Polls (non-junk):

Suffolk: N+4
WaPo: N+5
BRD: N+6
Polling Company: G+2
Christopher Newport Univ. : N+7

Average: N +4.0 (Down from N+ 4.6 before Suffolk Poll)

Polling averages don’t work unless you include the good-faith outliers.

If anything, Kellyanne Conway's poll should not be in any average, especially not in place of a non-partisan pollster.

Skipping over the Kellyanne Conway poll puts the Fox News poll in the average, so either way you get somebody that is not non-partisan.

[Attorney General & PPT Dwarven Dragon]

Larry Sabato got a good op-ed out on the on VA and NJ. He seems to think that Chesterfield is the county to watch. If Gillespie is in the high single digits he'll be close or might win. Cuccinelli and Gillespie in 2014 got 9% in Chesterfield and still lost.

I read it. I think better indicators might be Loudon and Virginia Beach. Northam claims a strong regional support in the tidewater, and historically had a margin n Virginia Beach similar to his statewide margin. Gillespie has a personal draw in NOVA, and Loudon broke for him in 2014 despite Gillespie losing the state - a rarity in the modern era. If one candidate is winning them both, then they should win in the end. In they are split, I suspect the margins in Chesterfield and early margins in Fairfax/Prince William will be the next best indicators of victory.

Of course all of this is moot if Northam is ahead with 70-80% of precincts in - when most of the state has reported but not much of Fairfax.

The early margin in Prince William is actually an extraordinarily bad indicator. Romney was leading 49-49 in Prince William with about a third of the vote in there, only to lose it 41-57 by the end of the night. Trump was within single digits in the early Prince William numbers, but proceeded to lose the county by 21 points.  Gillespie looked like he was going to win the county for most of election night 2014, only to lose it 48-50 as the final ballots came in.

[Attorney General & PPT Dwarven Dragon]

Settle down everyone. Northam is still going to win by 2-5%.

[Attorney General & PPT Dwarven Dragon]

5 poll average (As is usually done with averages, I am only counting one poll per company - so the older polling company poll has been eliminated, and BRD remains in the average even though it is technically poll #6. In response to some comments, I will use the three-way numbers for all companies.):

Rasmussen - TIE
Polling Company - G+2
Suffolk - N+4
Wash Post - N+5
BRD - N+6

Average - N +2.6

Northam still has the edge, but I'd put Gillespie's chances at like 40%. Still barely Lean D.

[Attorney General & PPT Dwarven Dragon]

5 poll average (As is usually done with averages, I am only counting one poll per company - so the older polling company poll has been eliminated, and BRD remains in the average even though it is technically poll #6. In response to some comments, I will use the three-way numbers for all companies.):

Rasmussen - TIE
Polling Company - G+2
Suffolk - N+4
Wash Post - N+5
BRD - N+6

Average - N +2.6

Northam still has the edge, but I'd put Gillespie's chances at like 40%. Still barely Lean D.

Three way for TPC (Toilet Paper Company) has a tie

That's the old poll. See their new poll: https://www.realclearpolitics.com/docs/VA_TPC_Topline_Oct_30-Nov_2.pdf

[Attorney General & PPT Dwarven Dragon]

http://firehousestrategies.com/vagov/

It appears Optimus Prime had a child named 0ptimus polling, they've got it at:

Gillespie:  40.4
Northam:  37.4

No undecideds or Hyra.

I am not posting this abomination in the polling board.

Øptimus is the same polling firm that had Roy Moore up 55%-45% three days before Moore won 54.6%-45.4% so...

Gillespie by 2-3 pp.

Optimus also lists an alternative metric:

You must be logged in to read this quote.

I will use that to avoid a calculation too favorable to Gillespie.

New Average:

Optimus - N + 1.5
Rasmussen - Tie
Polling Company - G+2
Suffolk - N+4
Wash Post - N+5

Average - N +1.7

This is less than half of the 2013 polling error. Northam should be scared. It is just barely below the 2% danger line, so I'll wait to see if we get another poll by Monday before considering a rating adjustment.

[Attorney General & PPT Dwarven Dragon]

With Roanoke:

Roanoke - Tie
Optimus - N + 1.5
Rasmussen - Tie
Polling Company - G+2
Suffolk - N+4

Average - Northam +0.7

Still time for the average to turn around, but Gillespie may have this. The late trend is certainly in his favor.

[Attorney General & PPT Dwarven Dragon]

With Roanoke:

Roanoke - Tie
Optimus - N + 1.5
Rasmussen - Tie
Polling Company - G+2
Suffolk - N+4

Average - Northam +0.7

Still time for the average to turn around, but Gillespie may have this. The late trend is certainly in his favor.

Im gonna slam my head into a desk.

Don't. Virginia will be fine regardless of who wins.

[Attorney General & PPT Dwarven Dragon]

With Roanoke:

Roanoke - Tie
Optimus - N + 1.5
Rasmussen - Tie
Polling Company - G+2
Suffolk - N+4

Average - Northam +0.7

Still time for the average to turn around, but Gillespie may have this. The late trend is certainly in his favor.

Im gonna slam my head into a desk.

Don't. Virginia will be fine regardless of who wins.

No, I mean I'm all for averaging out polls but we have had 4 questionable polls in a row. Your average is one decent pollster and four iffy pollsters.

2016 showed us that calling polls we don't like junk doesn't work. If something shows a ridiculously impractical result like Hampton did that's one thing, but a Tie or G+2 is not an unrealistic result.