I think another factor is that there was an early AZ call in 2020 by one news organization because they expected the mail-ins to lean D as they do nationally, but they actually leaned R due to new COVID dynamics specific to AZ (hence “STOP THE COUNT!” in GA, and “DON"T STOP COUNTING” in AZ by the Trumpians). AZ did go blue in the end, but the narrowing margin I’m sure made many squirm for a few days.
Yup - I think some learned to be cautious.
FWIW Predictit currently has:
AZ Senate 97 D/ 7R
AZ Gov 58 D/ 43 R
NV Senate 85 D/19 R
NV Gov 88 R/ 12 D
US House 93 R/ 9 D
Obviously there can be some wild swings on election day, but I’ve found it to be a pretty decent predictor, for lack of a better word
1 Like
Looks like one local news outlet is calling AZ for Kelly:
I can’t comment on the level of statistical rigor in that newsroom, however. I don’t think any larger outlets have made the call yet.
+1. I think 538 has beat the markets for the last few elections. But the markets are especially convenient in this waiting period, when nobody wants to declare the wrong winner.
1 Like
If I was in the election prediction business I would love to get assigned to Kentucky and Indiana where I have lived my whole life. In both states there isn’t even any need to know the names. Look at where they’re from, then pick blue or red based on results last 20-25 years.
1 Like
The hardest part of forecasting is getting the weights and sample size. I guess also there is a potential of people lying during the polling. Is that correct? I always feel like they don’t sample enough people for the different polls.
I’m sure there is some amount of lying. But there is also a meaningful amount of “I’m not saying” that may be skewed to one side. This was the case with Trump 2016.
Plus, if they’ve ever tried to poll me, I’ve not picked up the phone…
Polls (excluding the biased politcal funded ones like Trafalgar) were actually fairly accurate this cycle, underestimating Democrats on average. 2nd best performance by polls since 1998 if you go by average margin of error.
1 Like
You’re the rule:
We’ve been dialing for 3 hours, and poor Nick is still striking out.
FOUNTAIN: Yeah. At this point, I’d pretty much given up. I’d called 236 numbers and hadn’t gotten a single completed survey. The rejections on top of rejections - they started to get rather amusing in a dark-humored sort of way.
The last time I answered a political polling survey phone call is the last time I will ever answer a political polling survey.
It was [political party] masquerading as a polling company. When I told them I was planning on voting for [political party’s candidate] the next question was along the lines of “Would you like to volunteer for [political party]?”
No thanks, and now I feel like a sucker for getting my phone number on to your “He’s one of us” list.
You know what would help? An honest Caller ID.
Polls would be biased toward people WANTING to pick up the phone, but at least getting people to pick up the phone.
My wife likes tell me, “It’ sPam, your girlfriend again.”
Just tell me you’re Gallup or whatever.
2 Likes
In NV Masto has closed the gap with Laxalt to under 1000 votes. Looks like this one’s going to end up staying blue.
Futures markets are 95% certain that AZ & NV will stay D.
GA is about 75% for D.
If I’ve done my calculations correctly, that will leave the Senate 49+2 D+I & 49 R.
If all three go blue, that would give the Dems a lot of leeway to let one of Sinema/Manchin play bad cop/good cop as needed
1 Like
Wow, this 8pm Maricopa drop was supposed to be the better one for Republicans, but Hobbs widened her lead, and Kelly is now projected to win.
Trump sends out a fundraising blast email allegedly to help with the Senate races not yet called. However if you click through Trump gets 90% of the money, Herschel gets 10%, and the rest get 0.
https://twitter.com/samstein/status/1591192739797876737?t=By38lR46c2xnQFUXKKxDzA&s=19
This i’m sure warms the cockles of the Trump supporters here.
2 Likes
Correct. In that scenario the Dems pick up PA and no other Senate seats change party.
But Laxalt is still ahead in Nevada. The thought is the remaining votes will skew blue?
This is an “election data” guy from NV:
In a new batch of 5.1k ballots counted in rural, Republican-heavy Douglas County, here are the #nvsen results:
- Cortez Masto received 1.9k votes (37 percent)
- Laxalt received 3k votes (59 percent)
This increases Laxalt’s statewide lead from 800 to nearly 2k.
https://twitter.com/s_golonka/status/1591253248945815554?s=20&t=RvhQuafHeb2quHJCMhgmLA
The remaining uncounted ballots (possibly in the range of 40-50k ballots) are also predominantly from urban Clark and Washoe counties, where Cortez Masto leads Laxalt in total votes.
https://twitter.com/s_golonka/status/1591272734830780416?s=20&t=RvhQuafHeb2quHJCMhgmLA
1 Like