@Locrian and for anyone else who is interested…
There is Pennsylvania margin of victory market at PredictIt:
What will be the margin in the presidential election in Pennsylvania?
The farthest Trump leaning bracket is GOP 3% or more. You can, right now, buy 100 shares for $120 that will be worth $1000 in five days if Trump wins PA in a landslide.
Hmmm… well, I’m not paid until after the election…
For that kind of return, if you believe it’s a sure thing, you should borrow the money. One of those usurious payday loans would make financial sense if you believe what you’re saying.
For the record, I am not recommending that course of action and am not affiliated with PredictIt or any payday lender.
I went with my nephews wife to vote yesterday (I voted last week but I went to keep her company). It was quite the adventure.
She moved a couple of weeks ago., which came up in conversation while we were on our way there. I was like “just give them your old address” but she told them, they said “it doesn’t matter for early voting we can change the address now” BUT she had crossed county lines when she moved.
So we found a polling place in her new county and went there, about 30 minutes away, where they wanted proof of her new address, 2 pieces. She had her USPS address change not in the car but nothing else, but she was able to pull up her new utility account on her phone. So she got her ballot and voted. At no point did she consider giving up and waiting another day.
I felt really great about this. She and my nephew are both fairly apolitical and more conservative than I am, and they live in a rural area Neither is registered with a party affiliation. But she voted straight D this time and was talking about the importance of an overwhelming Biden victory so Trump can’t pull any shit.
I’m still a little worried about my nephew, because he works at one of those construction companies that paints all their vehicles with those over the top read white and blue flag/ American eagle designs and his coworkers try to fill his head with all that antifa/ socialist crap. Plus his mom is an air headed nut job that embodies every annoying liberal stereotype- frankly, if I were him I might be wrapping myself in MAGAgear just to piss her off. But I’m trusting his wife to get him to vote right, they’re newlyweds so she has some…levers, so to speak that she can pull.
This is anecdotal, but virtually all of the white people I’ve personally encountered in the trades are Trump people. Pretty much all the guys involved in construction, paving, masonry, excavation etc, if they’re white, they’re Trumpers. White truckers overwhelmingly seem to be Trump fans. Mechanics, welders, etc…Trump fans. I have to put the qualifier “white” in there because there are a lot of black and Hispanic people in all those jobs too and I have not observed them to wear their politics on their sleeve in any way; for all I know, a bunch of them support Trump too, but I can’t imagine it’s in the same numbers as the white guys.
I live in a blue city in a red Midwestern state.
[quote=“Mike_Mabes, post:220, topic:922525, full:true”]
is the first part possibly true? If so, do pollers adjust for it?
[/quote]Yes. traditionally, conservatives are less willing to answer polls than liberals. Anecdotally, I have heard several pollsters saying this trend is accelerating recently. Part of that likely is that historically, working-class voters are harder to poll, and the GOP is trending working class while the Dems trend professional class.
The two main ways pollsters account for this are either weighting (where you more or less simply put more weight on the answers you did get from whatever group you think you undercounted) or just calling more and more of that group until you get a sufficient number.
The problem with that, of course, is you have to start asking yourself whether the people of said group willing to answer your poll are truly representative of the larger group. And you really can’t know the answer for sure.
Of course, the same problems hold for ALL polling, especially nowadays when response rates are very low (<10%). Anyone willing to spend 30 minutes on a weeknight talking to a stranger is by definition a very atypical person. There are various strategies for mitigating this (e.g. polling online, etc), but at the end of the day people who don’t like being polled don’t get polled, and there’s no polls-based-solution to that.
There is of course, other, non-polling data.
Historically, incumbents that are unopposed in their party and nonetheless have good primary turnout win reelection. Trump had excellent primary turnout, and one model based on this has him at 91% likely to win. https://spectator.org/helmut-norpoth-trump-win-november/
In 2018, there was a lot of coverage of the Democrats’ success in registering new voters. That was smart, because historically, the party that’s been doing better in new voter registration wins elections. In 2020, the GOP has been registering more voters.
Voter enthusiasm can be measured, and every measurement goes in favor of Trump.
For that matter, poll questions like “who do you expect to win” or “who are your neighbors voting for” are historically better predictors of actual results than who respondants say they themselves are voting for. (Though it’s not a long history). Those poll questions favor Trump.
In fact pretty much every kind of data you can use to predict an election says Trump wins, maybe by a lot … and yet topline question on the polls say the exact opposite.
Other than the opinion of 538. com, what is your basis for thinking this pollster is reliable?
My prediction for the upcoming election is that the polls are mostly wrong. I think that Trump supporters will not say that they are voting for him. There has been enough back lash for supporting Trump that his voters are just keeping quite.
There have been consequences for openly declaring support, income and employability. A co-worker just lost his job for talking to students in favor of Trump. Of course they had other reasons, but if he had just kept his mouth shut until after the election he would still have a job. I believe that there are a non-insignificant number of voters who are wisely keeping quiet.
And the enthusiasm for the Biden-Harris ticket is missing. There is no message for why we should vote for them other than get rid of Trump, and I don’t think that is enough. Bad Trump is not a message.
I think that there is an unpleasant November surprise coming for most of the people on this board who never listen to the opposing point of view. My Democratic wife, long term Democrat, confided the other day that even though we live in a solidly blue state, she voted for Trump.
There is no message from the Biden campaign other than Bad Trump. I don’t see that as enough to win.
Did she explain why? I honestly, truly, seriously, genuinely want to know what the reasoning would be for a “long term Democrat” person to do this.
Well, for some values of “bad” is sure should be enough.The huge numbers of early voters this year is a sign, perhaps, of anti-Trump enthusiasm.
We’ll see.
I think that between GOP voter suppression and dismantling the mail system, Trump,will claim some kind of victory that his violent and ignorant base will support.
In truly dire situations, “grim determination” may well beat out “positive enthusiasm” as a motivation. That’s where many people are right now.
In terms of the polls, if you look at the last two presidential cycles, the poll misses could be ascribed to missing the “likely voter” assumptions. In 2012, it missed the bump in black turnout. In 2016, the whites-without-college-degrees turnout was missed. This year, we’re seeing record turnout on both sides (pointing to perhaps very little poll miss, like 2018 - everyone’s motivated).
And while I don’t put much stock in on the ground voting reports, if there is a surge, the reports I’ve seen are favoring it being in young voters (favoring Biden). I know “counting on the youth vote” is like “counting on Florida” for Democrats, but there are some strong signs in the early ballots at least.
This! My mom got a call from a neighbor in her retirement home this week. Mom said she mailed her ballot. Her neighbor said, “I hope you voted for Trump because he’ll keep the blacks out of here.” South Jersey, Burlington County. And no one around there seems to be voting for Biden/Harris because they think they’ll be great, just not as bad.
I just don’t understand why people trust polls. To me, they seem to be concerned about reputation and getting the credit for correct predictions, not necessarily by polling properly. Find me one reputable polling org that polls over 100 million voters, then that might be something I’d look at.
I just don’t understand why people trust polls.
Instead of anecdotes about someone’s spouse or someone else’s mom’s racist neighbor? Yeah, that’s crazy!
Uh, you know how pollsters earn their reputation and get credit? It ain’t by wearing snazzy pants or fudging their numbers.
Moderating
Note that Discourse accidentally parsed this into a link to another site (not 538). I broke the link.
Colibri
It was two things: Assholatry and the weird reason to trust polls. I should’ve used a line to separate them.
Not predicting the result, but if the vote goes against Individual-ONE, he will fight it all the way to January 20th, every way he can. I predict, provisionally, that he will observe that he still controls the government and will order Biden’s Secret Service detail back (not him personally, but the demand will percolate through Treasury).
If he does not get his way, there will be twelve weeks of complaining and unimaginable chaos.
Looking at the historical data on Gallup, Trump’s approval is probably about 3-5% below where it was for Obama and George W Bush, both of whom had tough re-election contests but managed to pull off a victory. But then again, Trump can say he win in ‘blue wall’ states, which Bush did not.
Trump needed to gain states; he hasn’t, and he is probably going to lose most, if not all, of his blue wall in the rust belt. He might hold FL, GA, NC, and OH. OTOH, he might lose AZ, and it’s still possible he could lose in TX.
If you scroll down 538’s Approval Rating page, select the “Net Approval” box (which is approval-rating minus disapproval-rating) and scroll just a little more, you will observe a difference of right around 9 percentage points with each of those two at this moment.
Polling 100M people is no longer a poll, that’s an election. There is no polling company or political organization in the country with that kind of money. Plus, it’s totally unnecessary, as that’s not how statistics work. Here is an SA article giving some of the math: