Tjanks

could the rock star who always posts "this day in 2020 versus today in 2024"

post some of that polling data?

I think it's pretty critical to see the polling with Biden versus the current polling from the same time against Harris.

Apologies if this has been discussed.

Could the EV R's enthusiasm be many (if not most) of the R's that typically vote on ED?
Seems a lot of the excitement is the increased R's EV turnout along with the "traditional" R's ED turnout would give Trump the wins.

Not a rock star but I think it's basically shown on the RCP averages toward the bottom; Trump is essentially up 5 vs. 2020 polls in the places that matter.

LMCane said:

---

could the rock star who always posts "this day in 2020 versus today in 2024"

post some of that polling data?

I think it's pretty critical to see the polling with Biden versus the current polling from the same time against Harris.

---

Edited out that data

That may not have worked. I would caution anyone putting too much weight on these comparisons. The industry has put a lot of effort into eliminating the bias they had and over counted, albeit without Trump, Republican support in 22. I think assuming they are undercounting right now is pretty loose.

Here's the page on RCP that shows their averages from 2016 and 2020 on the same day (Oct. 23).

https://www.realclearpolling.com/polls/president/general/2024/trump-vs-harris

October 23, 2020: Biden +8.0 | October 23, 2016: Clinton +5.9

agsalaska said:

---

LMCane said:

---

could the rock star who always posts "this day in 2020 versus today in 2024"

post some of that polling data?

I think it's pretty critical to see the polling with Biden versus the current polling from the same time against Harris.

---

Edited out that data

That may not have worked. I would caution anyone putting too much weight on these comparisons. The industry has put a lot of effort into eliminating the bias they had and over counted, albeit without Trump, Republican support in 22. I think assuming they are undercounting right now is pretty loose.

---

In 2016, Trump led/tied in 6 of 59 public polls in PA.

In 2020, he led/tied in 8 of 84 public polls in PA.

In 2024, he leads/ties in 39 of 54 public polls in PA.

Shy Trump voters are no more, polls have bad samples, or, something historic is happening...maybe all three? (1/4) pic.twitter.com/9sVHlrkZGx

— Athan Koutsiouroumbas (@Athan_K) October 23, 2024

#NEW @QuinnipiacPoll

Michigan :
🔵 Harris 49%
🔴 Trump 46%
Was 🔴 Trump +3, 2 weeks ago

Wisconsin:
🔵 Harris 48%
🔴 Trump 48%
Was 🔴 Trump +2, 2 weeks ago

— Political Polls (@PpollingNumbers) October 23, 2024

And ... the daily graph for Wed Oct 23, 2024 is out ... https://t.co/IbZyOXkf7h pic.twitter.com/SDHD5YGwyk

— Rasmussen Reports (@Rasmussen_Poll) October 23, 2024

Captn_Ag05 said:

---

#NEW @QuinnipiacPoll

Michigan :
🔵 Harris 49%
🔴 Trump 46%
Was 🔴 Trump +3, 2 weeks ago

Wisconsin:
🔵 Harris 48%
🔴 Trump 48%
Was 🔴 Trump +2, 2 weeks ago

— Political Polls (@PpollingNumbers) October 23, 2024

---

How reliable is Quinnipiac? This seems like a big shift back to Harris compared to other polls.

agsalaska said:

---

LMCane said:

---

could the rock star who always posts "this day in 2020 versus today in 2024"

post some of that polling data?

I think it's pretty critical to see the polling with Biden versus the current polling from the same time against Harris.

---

Edited out that data

That may not have worked. I would caution anyone putting too much weight on these comparisons. The industry has put a lot of effort into eliminating the bias they had and over counted, albeit without Trump, Republican support in 22. I think assuming they are undercounting right now is pretty loose.

---

I asked about this the other day, and never got a good answer about the "how". How do you adjust what is mostly a statistical model based on data you didn't/couldn't collect? Are they just using a fudge factor to adjust the final result based on past failures? Are they using different weightings? I haven't seen any pollster talking about what they are doing differently to adjust their results.

And if they are actually adjusting for it instead of just trying harder to reach shy trump voters, where is the proof that these adjustments have been made or that the adjustments are producing accurate results? If it is just in the weighting, we should be able to see changes in which categories are weighted how. If it is a fudge factor, that should be very easy to see. Nobody is talking about it or showing it, so why are you confident it has been done and is accurate?

Yes they use weightings applied to their poll data. Hypothetical example they may think Michigan usually has 54% voters are white females so even if the poll only had 40% white females they may normalize that.

Gallup suggests a +3% edge for identifying as Republican, yet this new Q poll sampled a materially higher perceentage identifying as Democrats.

4stringAg said:

---

Captn_Ag05 said:

---

#NEW @QuinnipiacPoll

Michigan :
🔵 Harris 49%
🔴 Trump 46%
Was 🔴 Trump +3, 2 weeks ago

Wisconsin:
🔵 Harris 48%
🔴 Trump 48%
Was 🔴 Trump +2, 2 weeks ago

— Political Polls (@PpollingNumbers) October 23, 2024

---

How reliable is Quinnipiac? This seems like a big shift back to Harris compared to other polls.

---

All the Harris surrogates and talking heads have been predicting that there would be a "final two week surge" in their favor that will pull them over the edge. Maybe that is what we are seeing here.

Hungry Ojos said:

---

4stringAg said:

---

Captn_Ag05 said:

---

#NEW @QuinnipiacPoll

Michigan :
🔵 Harris 49%
🔴 Trump 46%
Was 🔴 Trump +3, 2 weeks ago

Wisconsin:
🔵 Harris 48%
🔴 Trump 48%
Was 🔴 Trump +2, 2 weeks ago

— Political Polls (@PpollingNumbers) October 23, 2024

---

How reliable is Quinnipiac? This seems like a big shift back to Harris compared to other polls.

---

All the Harris surrogates and talking heads have been predicting that there would be a "final two week surge" in their favor that will pull them over the edge. Maybe that is what we are seeing here.

---

Ya'll are worried about Quinnipiac ? Dont they are unreliable

I wish I had a good answer for that. I think some of these guys consider it trade secrets but some are more open.

That would be a good question for Silver and I bet he addresses it post election.

4stringAg said:

---

Captn_Ag05 said:

---

#NEW @QuinnipiacPoll

Michigan :
🔵 Harris 49%
🔴 Trump 46%
Was 🔴 Trump +3, 2 weeks ago

Wisconsin:
🔵 Harris 48%
🔴 Trump 48%
Was 🔴 Trump +2, 2 weeks ago

— Political Polls (@PpollingNumbers) October 23, 2024

---

How reliable is Quinnipiac? This seems like a big shift back to Harris compared to other polls.

---

The sample weighting with the small/academic public polls is pretty wild.

QUINNIPIAC POLLS

Michigan:
🔵 Today: Harris+3 (Sample: D+4)
🔴 Last poll: Trump+3 (Sample: R+1)
🟠 Sample shift: D+5

Wisconsin:
🟡 Today: Tie (Sample: D+1)
🔴 Last poll: Trump+2 (Sample: R+3)
🟠 Sample shift: D+4

🙄🙄

— Eric Daugherty (@EricLDaugh) October 23, 2024

Bob Knights Paper Hands said:

---

Gallup suggests a +3% edge for identifying as Republican, yet this new Q poll sampled a materially higher perceentage identifying as Democrats.

---

I'm never quite sure how they do this, but from what I've read, it seems that the control question they use to "balance" the sample is primarily to ask respondents how they voted in the last presidential election. They know what that number "should" look like, so they adjust their sample until it is weighted in accordance with that prior election number. Now, when they show party ID in their sample of D+4, I assume that is a weighted response, and corresponds to the last general election. But, if Gallup is correct, then that is probably the wrong method to weight their sample.

As has been mentioned, some of these "polls" are simply the pollster taking data gathered by some other outfit and applying their own weighting, so there is obviously a lot of statistical manipulation involved in coming up with the top line numbers.

harge57 said:

---

Yes they use weightings applied to their poll data. Hypothetical example they may think Michigan usually has 54% voters are white females so even if the poll only had 40% white females they may normalize that.

---

I understand the weighting, but the 2020 and 2016 polls were weight adjusted too. What the poster I was responding to was saying was that they had changed their polls to fix the previous misses that were believed to be caused by shy trump voters. If they changed their weightings to account for it, how did they do it? Which demographics are they enhancing to overcome the shy voter undercounts? If the weightings still match the expected demographics of the population and they are not applying some other factor, then they haven't fixed the glitch unless they have found some magical way to account for shy trump voters who they can get to respond to surveys.

Silver is saying Trump…

https://www.foxnews.com/media/pollster-nate-silver-says-gut-tells-him-trump-win-election

Captn_Ag05 said:

---

LMCane said:

---

could the rock star who always posts "this day in 2020 versus today in 2024"

post some of that polling data?

I think it's pretty critical to see the polling with Biden versus the current polling from the same time against Harris.

---

Here you go

---

Every state tracks pretty well from 2016 to 2020 except for Georgia. That 2020 number looks off compared to the rest. Wonder why?

[Enough with the derails on this thread. Next one and you're taking some time off -- Staff]

35% of the Georgia vote is black and that was at the height of the BLM movement. They were energized and voted in higher numbers than usual.

🇺🇲 National poll by Wall Street Journal

2-WAY
🟥 Trump: 49%
🟦 Harris: 46%

Last poll (9/28) - 🔵 Harris +1
——
FULL FIELD
🟥 Trump: 47%
🟦 Harris: 45%

Last poll - 🔵 Harris +2
——
October 19-22 | N=1,500https://t.co/pCIMQoFVWf

— InteractivePolls (@IAPolls2022) October 24, 2024

Trump up 47-45 in latest WSJ poll. Article just dropped.

Quite the spread

#NEW NATIONAL poll - job approval rating

Trump:
🔴 Approve: 52% (+4)
🔵 Disapprove: 48%

Harris:
🔵 Approve: 42% (-12)
🔴 Disapprove: 54%

WSJ | 10/19-22 pic.twitter.com/xqqHVGSARd

— Eric Daugherty (@EricLDaugh) October 24, 2024

Harris down to .02 in the national RCP average

How accurate was WSJ the last two presidential elections with Trump?

the good thing is this has 1600 respondents versus the earlier main stream poll which showed Trump also up 3 but with only 450 respondents.

so between a two person and field lead of +2.5 Trump nationally

that would be a beat down if accurate.

keep needing these national polls to move Trumps way!

keep in mind in the election of 2020 this same poll at the same time had Biden +10 points. So a 12.5 point swing in the same poll.

#NEW NATIONAL poll

🔴 Trump: 49% (+3)
🔵 Harris: 46%

WSJ | 10/19-22 | N=1,500

— Eric Daugherty (@EricLDaugh) October 24, 2024

Bovada moved back down a couple notches back to -165 this afternoon

Probably a little reaction to the Hail Marys being thrown right now, which will prove to be nonsense quickly

📊 AZ, GA & NC POLL: Marist/NPR

NORTH CAROLINA
🟥 Trump: 50%
🟦 Harris: 48%
—
ARIZONA
🟥 Trump: 50%
🟦 Harris: 49%
—
GEORGIA
🟥 Trump: 49%
🟦 Harris: 49%
—
#6 (2.9/3.0) | 10-17-22 | Likely votershttps://t.co/w9CHQIiED2 pic.twitter.com/TFzcSFYKGi

— InteractivePolls (@IAPolls2022) October 24, 2024

FrioAg 00 said:

---

Bovada moved back down a couple notches back to -165 this afternoon

Probably a little reaction to the Hail Marys being thrown right now, which will prove to be nonsense quickly

---

also a few points move to Harris on polymarket

but looking at the last few weeks it does seem to be hopefully the result of betting like the stock market has ebbs and flows every day

hopefully we see it trend to an increasing gap again in next day.