Captn_Ag05 said:

---

Well, well, well

#New General election poll

🔴 Trump 47%
🔵 Harris 47%

Last poll - 🔵 Harris +4

YouGov (Yahoo) #B - LV - 10/4

— Political Polls (@PpollingNumbers) October 7, 2024

---

That's a big jump…

Have several pollsters stopped with their polling? Was it just the news outlet pollsters stopping?

It's a little suspicious nothing really in 3 days from the 'big names' but it's also been across the weekend.

Wow!

"If" the polls are off.

Basically, "if" polls like TIPP are wrong when they show Harris +16 with Independents. https://t.co/fgDN2deZyP

— Mark Mitchell, Rasmussen Reports (@Mark_R_Mitchell) October 7, 2024

May have put this next tweet up already, sorry if so;

Good Morning !

We have Trump up by 2% nationally and @atlas_intel has him up 3%. Published reports from battleground states have Harris weakening. How does the polling industry respond? By releasing ZERO new national results for 3 full days. Our response below ... https://t.co/E80bHWlt4f pic.twitter.com/egNnonAyVb

— Rasmussen Reports (@Rasmussen_Poll) October 7, 2024

RE: weighting…

Reader Reminder: We are weighting nightly election results to the 2020 Recall Vote (Who did you vote for in 2020?). If we didn't do this - which we believe helps our accuracy - our Trump v Harris results would be Trump +5 and our new Generic Ballot today would be Republicans +5. https://t.co/fFyZxAAAZH

— Rasmussen Reports (@Rasmussen_Poll) October 7, 2024

I apologize for not understanding what you answered with in regards to my question on Rasmussen focusing on recall polling.

Is it that they believe that to be a more accurate metric? Has this shown to be true in the past?

Quote:

---

How does the polling industry respond? By releasing ZERO new national results for 3 full days

---

Is that accurate? They just stopped reporting poll numbers?

#NEW PENNSYLVANIA poll

🔴 Trump: 50% (+4)
🔵 Harris: 46%
🟢 Other: 3%
🟡 Undecided: 1%

Democracy Institute | Oct. 2-3 | N=400LV

— Eric Daugherty (@EricLDaugh) October 7, 2024

#NEW WISCONSIN poll

🔴 Trump: 48% (+2)
🔵 Harris: 46%
🟢 Other: 3%
🟡 Undecided: 3%

Democracy Institute | Oct. 2-3 | N=375LV

— Eric Daugherty (@EricLDaugh) October 7, 2024

#Updated Electoral Map Based on @RedfieldWilton polling

🟦 Harris 276 🏆
🟥 Trump 246
——
Michigan - 🔵 Harris +2
Wisconsin - 🔵 Harris +1
Pennsylvania - 🔵 Harris +1
Nevada - 🔵 Harris +1
Georgia - 🟡 TIE
Arizona - 🔴 Trump +1
North Carolina - 🔴 Trump +2
——@Polymarket U.S.… https://t.co/qnyDOpaT6f pic.twitter.com/k5CUjKO4cN

— InteractivePolls (@IAPolls2022) October 7, 2024

I'm probably not the best 'explainer' here on this, sorry, but here is a lengthier unbiased piece on it. The short 'nortex version' is they weight the adjusted count, if using recall voting as a parameter, such that by way of example, if the electorate was 52 percent women last time (in 2020) then this time that is the value they adjust for as to the female voter rate, and similarly for other factors such as party affiliation, instead of just picking a weight as to what they think the electorate will look like by gender (or party, etc).

As the linked author discusses, it's actually a lot more complicated/nuanced than what the Rasmussen guys have to say on it, imho. Still, at least some weighting or presentation as to that option would make sense as it's certainly an indicator as to a valid option/pattern, imho, and not that tough for the pollsters to put out.

🇺🇲 National poll by Yahoo/YouGov among likely voters (full field)

🔴 Trump: 47% [+2]
🔵 Harris: 47% [-2]

[+/- change vs 9/13]

#4 (2.8/3.0) | 10/2-4 | LV | ±3%https://t.co/4leY88fOSz

— InteractivePolls (@IAPolls2022) October 7, 2024

Yougov typically has a fairly large Dem bias, imho.

The most likely map according to Nate Silver, occurring 22% of the time pic.twitter.com/PP2qzGfmRb

— Basil🧡 (@LinkofSunshine) October 7, 2024

So Trump doesn't win even 1 swing state…Silver seems to have gone straight campaign mode from where he was a month ago.

The MOST likely map has a 22% probability?

I'm reading that as 22% of the scenarios have that map as the result.

TRM said:

---

I'm reading that as 22% of the scenarios have that map as the result.

---

Yep. Theoretically 78% of the scenario maps could have trump winning, just with different states.

Silver would be a great example for an 8th grade science lecture on precision vs. accuracy. Very precise but not very accurate.

Something tells me Trump is not going to over perform polls as much as he did in 2016 and 2020.

People were reticent to admit they were voting for Trump then, but I don't think they are as shy about it this go around.

sam callahan said:

---

Something tells me Trump is not going to over perform polls as much as he did in 2016 and 2020.

People were reticent to admit they were voting for Trump then, but I don't think they are as shy about it this go around.

---

Given the level of Dem rhetoric, I'd say people that switched to or back to Trump are going to be even more shy about saying so. LITERALLY HITLERS! DANGEROUS TO OUR DEMOCRACY!!

Silvertaps said:

---

So Trump doesn't win even 1 swing state…Silver seems to have gone straight campaign mode from where he was a month ago.

---

Silver is full of Sh**. This is going to be a Trump blowout.

I think you need to post the next post in that thread. Silver's second most likely scenario has Trump winning 19.9% of the time. So his model has it really close.

The next 4 most likely, happening 19.9%, 4.5%, 4%, and 3.9% of the time

Together they make up 54% of outcomes!

The 270 Kamala map happens 2.15% of the time pic.twitter.com/IyhJOxd3he

— Basil🧡 (@LinkofSunshine) October 7, 2024

Why are we so confident Trump will over perform the polls? I would like to think (hope), but this isn't a given.

Definitely not a given. GOP underperformed the 2022 polls.

Polling is not static in terms of strategy, mathematical correction, etc. The pollsters change their techniques to adjust to prior error. Most of these organizations have spent quite a lot of time over the last few years working on getting their accuracy improved and making mathematical adjustments to their models.

DonHenley said:

---

Why are we so confident Trump will over perform the polls? I would like to think (hope), but this isn't a given.

---

It's like the "Bradley effect". Some people won't admit to pollsters that they are voting for Trump.

Also this:

Quote:

---

A related concept is social desirability bias, which describes the tendency of individuals to "report inaccurately on sensitive topics in order to present themselves in the best possible light."

---

DonHenley said:

---

Why are we so confident Trump will over perform the polls? I would like to think (hope), but this isn't a given.

---

Saw a tweet this morning that quoted an individual in the Trump campaign. Said their internal numbers showed multiple swings states outside the MOE in their favor.

I'll see if I can find it. I think reports like that add to the confidence levels.

DonHenley said:

---

Why are we so confident Trump will over perform the polls? I would like to think (hope), but this isn't a given.

---

This is my concern.

Also, the polls can only tell so much. They can't even remotely predict the ballot harvesting efforts in GA, PA, AZ. Heck, PA stated about 2 months ago that there's no chance they'll be able to confirm who won their EV's, at least a few weeks out.

Is Nate Silver a died in the wool Dem shill, paid to put out misinformation in his "data"?
Is Rasmussen overconfident in it's algorithm?

The next 4 weeks are going to be insufferable if you're hanging on every single poll.

AtticusMatlock said:

---

Definitely not a given. GOP underperformed the 2022 polls.

---

Not really true. Midterm polling is a challenge, because pollsters are not typically polling at the congressional district level with regularity. They largely do the head-to-head overall congressional ballot choice. It was almost spot on in the average of the polling vs. the actual vote outcome. Republicans won the congressional vote nationwide by 2.8 points and the average polling difference was 2.5.

📊 VIRGINIA Poll by @WasonCenter

🟦 Harris: 52%
🟥 Trump: 41%
🟪 Other: 3%
—
VA Senate
🟦 Tim Kaine (inc): 55%
🟥 Hung Cao: 35%
——
Fav/unfav
• Walz: 47-35 (net: +12)
• Harris: 51-42 (+9)
• Vance: 34-50 (-16)
• Trump: 36-58 (-22)
——
#20 (2.8/3.0) | 800 LV | 9/28-10/4…

— InteractivePolls (@IAPolls2022) October 7, 2024

#NEW NATIONAL poll

🔴 Trump: 50% (+3)
🔵 Harris: 47%
🟢 Other: 2%
🟡 Undecided: 1%

Democracy Institute | Oct. 2-3 | N=1,200LV

— Eric Daugherty (@EricLDaugh) October 7, 2024

DonHenley said:

---

Why are we so confident Trump will over perform the polls? I would like to think (hope), but this isn't a given.

---

It's not so much confidence in over performing the actual polls it's the assumptions and weighting that is applied to the polling results by the pollsters. They apply demographic weighting to the poll results based off of past election trends and Trump has shown to present significant shifts in the traditional demographic splits. I.e. young people, Hispanics, unions, black men, general enthusiasm, etc.

Maybe they now have enough data from two Trump elections to be more accurate, but they weren't accurate in 2020.

Dialing up the sample size. We'll hit 500+ tomorrow, folks.

Weighted D+2, and by gender, age, race, region, and recalled vote. pic.twitter.com/zgcx1aKjgS

— Rasmussen Reports (@Rasmussen_Poll) October 7, 2024

I'll post two things that I don't think add up:

Part 1: Harris+2 in New Hampshire (Biden won it by 7.5).

#NEW NEW HAMPSHIRE poll

🔵 Harris: 47% (+2)
🔴 Trump: 45%
🟢 Other: 6%
🟡 Undecided: 2%

Democracy Institute | Oct. 2-3 | N=325LV

— Eric Daugherty (@EricLDaugh) October 7, 2024

Fine, small sample size, but NH is basically one tiny, college educated white state (about 61 percent white, 12 percent black, relatively older, few kids/young adults, ballpark figures).

Part 2 Chaser: Kamala is supposedly up 18 with college educated whites:

Kamala Harris seems to be on track to put up the best numbers ever for a Democratic presidential nominee among white voters with a college degree & among all voters with a college degree.

Those with a college degree are likely to make up a record high share of the electorate. pic.twitter.com/BIjMbOX3fC

— (((Harry Enten))) (@ForecasterEnten) October 7, 2024

These things can't both be true, imho.

If I had to guess, I don't think Trump is really losing college whites by that much, but I dunno.

Good news:

If this poll is right, the rightward shift from 2020 will be 7.5 ppt nationally, at the upper limit of the 3 to 8 point shift we see in various states. If Trump wins pop vote by 3, it'll be an electoral college blow out at 350+ EV. https://t.co/qvFnUXYqYH

— au ng (@athein1) October 7, 2024

Quote:

---

The poll found 26 percent believed inflation was the most pressing matter, while 23 percent thought immigration was the most crucial issue. The economy and crime tied in third place with 13 percent of votes, followed by abortion, health care and the environment.

---

Great post, but isn't it also true that Trump turns out low propensity voters that are not accounted for i.e. not polled??

nortex97 said:

---

If I had to guess, I don't think Trump is really losing college whites by that much, but I dunno.

Good news:

If this poll is right, the rightward shift from 2020 will be 7.5 ppt nationally, at the upper limit of the 3 to 8 point shift we see in various states. If Trump wins pop vote by 3, it'll be an electoral college blow out at 350+ EV. https://t.co/qvFnUXYqYH

— au ng (@athein1) October 7, 2024

Quote:

---

The poll found 26 percent believed inflation was the most pressing matter, while 23 percent thought immigration was the most crucial issue. The economy and crime tied in third place with 13 percent of votes, followed by abortion, health care and the environment.

---

---

The problem with polls like that is teasing out the "why" for people's responses. 26% may think inflation is the most pressing matter, but 10% of that may be from people who think it is being caused by price gouging by big conglomerates, while 16% think it is out of control government spending causing it. Those two groups will vote differently and not as one block.

It is kind of like the polls always showing huge disapproval for congress. Those polls mix in hate from both sides. There is always something to hate about congress, regardless of who you vote for. Just because everybody hates them doesn't mean there is enough momentum on either side for a meaningful change to occur in the elections.

Polymarket betting odds shot up big time the last few hours for Trump

54-46 trump

Was tied a few days ago

Anyone know what caused the odds to go virtual earlier today ?