I have 'and the army, goes rolling, along…' running in my head at this point.

#NEW Nate Silver 2024 election forecast update

🔴 Trump: 50.2% [+3]
🔵 Harris: 49.5% [-3.1]

[+/- over last week] pic.twitter.com/YqUjMXHdrj

— Eric Daugherty (@EricLDaugh) October 17, 2024

There are very few uncommitted voters... without leaners, it was Trump 49, Harris 48 https://t.co/ScFxnza0cx

— Scott Rasmussen (@ScottWRasmussen) October 17, 2024

Michigan Presidential:
Trump 49% Harris 49%
Last month, Harris +3
Conducted by @ScottWRasmussen for @NapolitanNews
789 LV, Oct 10-16

— Scott Rasmussen (@ScottWRasmussen) October 17, 2024

I hope everybody realizes what this election is going to look like if Wisconsin polling isn't "fixed," which underestimated Trump by 5-6 points in both 2020 and 2016. https://t.co/BsQlts84JB

— Eric Daugherty (@EricLDaugh) October 17, 2024

Somewhere up above, I believe someone referenced that right track/wrong track has gotten POTUS election right every time since 1980. And keep in mind, some of these folks (not calling Scott out specifically) don't want to break too hard on Trump momentum now because…they want to be able to sell their services to folks on either side in the next cycles, and claiming insight into things that 'could' still happen late if people 'had listened' to them is helpful in this regard.

This is yet another indicator of sentiment in polling:

#NEW MICHIGAN poll - who are your friends/neighbors voting for?

🔴 Trump: 45% (+10)
🔵 Harris: 35%

MIRS | 10/14 | LVs

Poll was Trump+1-2.

This begs the question - poll-answerers may say they're voting for Harris. What about everyone else?

— Eric Daugherty (@EricLDaugh) October 17, 2024

SwigAg11 said:

---

1836er said:

---

LOL. It's October 17 and they're publishing their FINAL Michigan poll.

I wonder why many of these pollsters are deciding to make mid-October polls their Final polls, rather than continuing to poll for the next three weeks?

What could it be... hmm... what could it be...?

---

When have these polling firms typically published their final polls in previous elections?

---

RMG's (the poll mentioned above) last presidential poll in Michigan four years ago polled from October 26-29, which means it was probably published 2-3 days before the election. It was Biden +7.

https://projects.fivethirtyeight.com/polls/president-general/2020/michigan/

I am curious what the Polymarket was in 2020 a few days before the election.

Looks like the betting had Biden winning around 308 EVs. So it got it right.

Trump in 2016 and Trueman are the only 2 upsets known.

I don't know the reason, but a lot of pollsters stop polling in the last couple of weeks of an election.

will25u said:

---

I don't know the reason, but a lot of pollsters stop polling in the last couple of weeks of an election.

---

The time constraints make polling more expensive. Need more resources.

Which is why public polling is mostly crap and the internal polls are much more accurate because you get what you pay for.

texagbeliever said:

---

Looks like the betting had Biden winning around 308 EVs. So it got it right.

Trump in 2016 and Trueman are the only 2 upsets known.

---

Much appreciated. At work and not allowed to access the site.

I tend to believe developed betting markets over any political poll, for the obvious reason that people putting up large sums of money have thoroughly researched well beyond public data, not unlike equities markets.

📊 MICHIGAN GE: @Rasmussen_Poll

🟦 Harris: 48% [=]
🟥 Trump: 48% [=]

MI Senate
🟥 Rogers: 44% [+1]
🟦 Slotkin: 44% [-3]

[+/- change vs 9/19-22]
——@AmericanThinker | 10/9-14 | 1058 LVhttps://t.co/6ub4OUOIhG pic.twitter.com/owAzN4fTwz

— InteractivePolls (@IAPolls2022) October 17, 2024

aggiehawg said:

---

will25u said:

---

I don't know the reason, but a lot of pollsters stop polling in the last couple of weeks of an election.

---

The time constraints make polling more expensive. Need more resources.

---

added bonus for the pollsters: copout excuses for being wrong.

"situation was fluid, but we were right at the time"
"circumstances changed in those last 2.5 weeks before election day, but we were right at the time"
etc.
etc.
etc.

will25u said:

---

I don't know the reason, but a lot of pollsters stop polling in the last couple of weeks of an election.

---

I think most of them stop reporting results publicly once early voting is underway to avoid the appearance of trying to influence the vote. Same reason the exit polling info doesn't get released for each state until the polls close.

📊 WISCONSIN GE: @Rasmussen_Poll

🟥 Trump: 49% (+2)
🟦 Harris: 47%

Last poll (9/22) - 🟡 Tie
——
WI Senate
🟦 Baldwin (inc): 46% (+1)
🟥 Hovde: 45%
——@AmericanThinker | 10/9-14 | 1,004 LVhttps://t.co/6ub4OUOas8 pic.twitter.com/rcoGsBY6v9

— InteractivePolls (@IAPolls2022) October 17, 2024

With the addition of that poll, Trump has taken the lead in the Wisconsin average, giving him a lead in every battleground. Small leads, but significant that he leads them all.

shouldn't Wisconsin be an obvious lock for Trump at this point or is there real suspicion that pollsters' adjustments finally got it right this time?

Trump won WI in 2016 by less than 1%, despite being behind Clinton by over 5% in final polling.

Trump lost WI in 2020 by less than 1%, despite being behind Biden by over 7% in final polling, and despite all the shenanigans that went along with a mass-mailed scamdemic-induced crazy weird election.

in 2024, Trump polling in WI is looking like a virtual tie.

Yes, that would be my guess, but past isn't always prologue.

If 2016-2020 polling trends apply Trump should do better in WI than MI and PA if all three states are now polling very close... because the gaps between where he has polled versus election results in WI have been slightly larger than the gaps in MI and PA... plus he just tends to do slightly better in Wisconsin period.

One thing that might be an indicator of whether they have "fixed the glitch" is each "grouping" of polls using the same data source moving in the same direction by roughly the same amount relative to 2020 and 2016. But if each polling group has a different take on how to fix the glitch, you would expect to see them all over the map, with some moving too far towards R and some not far enough. My take on the most recent round of polls is that they all seem to have moved towards R by ~6-8 points relative to 2020, which at least to me is more likely to be a real move in the electorate and not an indication of them having added some adjustment to account for their previous misses. If they were trying to fix their previous misses, I would expect to see a much larger spread between the various polling groups than we are seeing right now.

But who knows...will be an interesting night in a few weeks.

My hunch is and my head says you are correct... which seems to mesh with Gallup's recent polling on party identification.

I won't feel good about it until (if) in turns out that way in reality though.

The latest New York Times polling finds President Trump has significantly cut into the black voting block for Democrats.

Only one percent unaffiliated or independent in the first two, but seven in the current? I wonder if there is a significant amount that still cannot call themselves republicans.

aggiehawg said:

---

The latest New York Times polling finds President Trump has significantly cut into the black voting block for Democrats.

---

after Trump got more of the black vote in 2020 than any republican in 40 years, he could double it in 2024.

that's one thing in the crosstabs that doesn't jive with the overall polling, it seems.

jeremiahjt said:

---

Only one percent unaffiliated or independent in the first two, but seven in the current? I wonder if there is a significant amount that still cannot call themselves republicans.

---

That could also be true.

It's actually very believable as I think polling has pretty consistently shown. Black men have moved for multiple reasons including the law fare and economy and immigration stuff.

It seems to me the external polling and whatever the candidates are using and seeing internally are starting to align. The external polls have all starting trending Trump in both battleground and national.

I know this is a polling thread and I'm trying to relate this to the polls: the candidates actions and tactics speak to what I think they are seeing in polls. Trump seems relaxed and confident and surging, Harris seems desperate as she's doing the interviews, the Fox interview, trying to appeal to black voters, and has moved her campaign from "Joy and Hope" to "Trump evil".

nortex97 said:

---

It's actually very believable as I think polling has pretty consistently shown. Black men have moved for multiple reasons including the law fare and economy and immigration stuff.

---

I wonder if the polling will end up understating black support for Trump with the backlash that seems to have happened against Obama's scolding black men.

📊 ARIZONA POLL: CBS/YouGov

🟥 Donald Trump: 51%
🟦 Kamala Harris: 48% pic.twitter.com/c38NDGSbHb

— InteractivePolls (@IAPolls2022) October 17, 2024

Strong Democratic Overnight Results: Trump 48%, Harris 48% https://t.co/23qhvfr6VZ pic.twitter.com/QM3gKA3yzp

— Rasmussen Reports (@Rasmussen_Poll) October 17, 2024

So at this point Harris needs a massive polling error in her favor, and that goes completely against the historical trend.

texagbeliever said:

---

Just your friendly reminder that the right direction / wrong track has correctly predicted the presidential winner since 1980.
Currently the democrats are -35 or so points.

---

Fox showed Wrong track at 79% last night

agreed that is usually a good predictor of the polls final outcome

Captn_Ag05 said:

---

Can we please stop giving Nate silver and his "model" legitimacy? He's a fraud and deeply partisan.

---

the thing about Silver is that he took a LOT of grief from the leftists about 3 months ago when he was showing Trump winning.

then he had the biggest lead for Harris. I do really want his tool to continue to show a growing lead for Trump- he does do 7000 simulations every few days so the better our numbers, the better the Silver Tool will look.

4stringAg said:

---

It seems to me the external polling and whatever the candidates are using and seeing internally are starting to align. The external polls have all starting trending Trump in both battleground and national.

I know this is a polling thread and I'm trying to relate this to the polls: the candidates actions and tactics speak to what I think they are seeing in polls. Trump seems relaxed and confident and surging, Harris seems desperate as she's doing the interviews, the Fox interview, trying to appeal to black voters, and has moved her campaign from "Joy and Hope" to "Trump evil".

---

I noticed the exact same thing.

if you didn't know the state of the race from these polls:

Trump campaigns at Coachella in California, Madison Square Garden in NYC, Colorado.

Harris after two months hiding goes on Fox News and gets demolished.

"Brat Summer" Harris turns into today yelling angry Harris.

she must be seeing some internal polling to have everything change in the last month.

oh no said:

---

shouldn't Wisconsin be an obvious lock for Trump at this point or is there real suspicion that pollsters' adjustments finally got it right this time?

Trump won WI in 2016 by less than 1%, despite being behind Clinton by over 5% in final polling.

Trump lost WI in 2020 by less than 1%, despite being behind Biden by over 7% in final polling, and despite all the shenanigans that went along with a mass-mailed scamdemic-induced crazy weird election.

in 2024, Trump polling in WI is looking like a virtual tie.

---

Real Clear Politics polling average just moved wisconsin into Trump lead

This is the polling agencies' third attempt at trying to figure out the Trump electorate. It's been proven particularly difficult in Wisconsin. They have made some adjustments to how they try to account for the shy Trump voter but I suppose we will just have to wait and see. Obviously none of them want to gain a reputation of getting things completely wrong. My guess is the polls will be closer to reality this time.

Is it unpossible the pollsters with bad records corrected or improved their methodology? I love the idea that poll XYZ is always off by 4pts, but wouldn't they be trying to better that between election cycles? I can't help but be skeptical about assuming they'll keep being so wrong.

rononeill said:

---

Is it unpossible the pollsters with bad records corrected or improved their methodology? I love the idea that poll XYZ is always off by 4pts, but wouldn't they be trying to better that between election cycles? I can't help but be skeptical about assuming they'll keep being so wrong.

---

If it was one election, maybe, but both 2016 and 2020 were both off quite a bit. The point has been, that even if they are off half of what they were, with current polls, Trump wins in a landslide. It's very hard to poll Trump supporters.

2022 polls however was grossly over polled for republican senate seats so one could argue that on the other side, but polling off year elections and without Trump is way different.

I might spend some time looking later at polling data from 2012, 2008, 2004, etc. The innacuracy of the last few election cycles had been very different, at least as far as I know, from the norm.