Sign up for The Wire Newsletter!

The Trump Card...

muttly

Veteran Expediter
Offline
Thank you. I see at the top of the chart, the final summary of the polls shows Clinton up by 2.1%, which is EXACTLY the margin she won the popular vote by. As I said above, the national polls do not account for the Electoral College. They measure the popular vote. By this measure, the 2016 final-week polls, taken as a whole, were spot-on accurate.
The final result is what the popular vote result was. Their average at the end was off just over one point. The National polls tightened considerably at the end.
 

ATeam

Senior Member
Retired Expediter
Offline
4 years ago in Michigan poll :
We can pick individual polls to support one point of view or another regarding poll accuracy. My approach is to cite the aggregate of all polls reported by FiveThirtyEight. Within that universe there are polls that will prove to be close to or exactly on the mark, and those what will be further from or way off the mark. If you want something to be true about polls, you can certainly find a poll to cite as evidence. Given the nature of polling itself, that seems like an unproductive approach.

For the purposes of discussion, I assert that, in general, and on average, and taken as a whole, polls reasonably reflect the voter intentions at the time the poll was taken. I also assert that the closer the polls are taken to Election Day, the more likely it is that the polls will accurately project the Election Day results.

It is not uncommon for polls to say something, and then something happens on the ground that produces a major shift in sentiment, and the next polls show that sentiment shift. Does that mean the previous polls were wrong? No. It does not. It means they were right at the time they were taken, and right again when the sentiment shifted.

Can you find polls that were taken at the same time but disagree? Yes you can. That's where the conversation shifts to the methodology and prior accuracy of the pollster in question. FiveThirtyEight accounts for that by rating the past accuracy of the polls it lists.

I get it, Muttly. You really want the current polls to be wrong and really want Trump to win. You are of course free to pick individual grains out of the greater sandbox to support your desire. I'm not going there for the reasons stated above. Polls are general in nature, not specific. In general, they are accurate, and always, you can find exceptions to the rule. But a small number of exceptions do not disprove the rule. The exceptions are included in the universe of all polls. The exceptions do not disprove the rule because the greater number of polls were accurate or close to accurate.
 
Last edited:

Turtle

Administrator
Staff member
Owner/Operator
Offline
I remain hopeful that civility will improve.
I remain very skeptical. Liberals (and the anarchists) have been practicing all summer for the election. People are people and they're very predictable. While history doesn't exactly repeat, it does rhyme. People still, right now, believe Trump stole the election from Hillary with Russia's help, and the Dems and the media have pounded into the brains of the malleable that the only way Trump can win this time is by cheating. Biden just said the same thing two days ago. That's been a thing for a while with Democrats as it is. Stacey Abrams even today maintains that the only reason she's not the governor of Georgia is because the Republicans stole it from her, mainly by voter suppression, despite Georgia having the highest voter turnout, and the highest black voter turnout in the state's history.

Trump was impeached because, in the words of Democrats, if he isn't impeached, he will steal the 2020 election with the help of Russia.

People believe the Fine People Hoax. They believe Trump is a white supremacist, and that justifies violence against him and his supporters. He's been branded as literal Hitler, which also justifies violence. He's been labeled a dictator, which justifies violence. Biden leads in the polls by an insurmountable margin, and if he doesn't win, the only conclusion that you can draw from that is Trump cheated to win the election.

The polls will almost certainly tighten up the week before the election, but the people described in the above paragraph are not going to be paying any attention to that.
 
  • Like
Reactions: muttly

coalminer

Veteran Expediter
Retired Expediter
Offline
Thats just great, neither one of the candidates are going to accept the results, we are screwed.....

I already filled out my form and sent it in, I didnt vote for either of them, I went libertarian, not that I believe it would be a good thing to have a libertarian president, but im taking a page from the 2 big party playbook, a means to an end....

 
  • Like
Reactions: Turtle

dalscott

Expert Expediter
Offline
Just like golf.

These fellow citizens that you're so concerned about have physically attacked children and seniors for wearing a MAGA hat. These are the same citizens who Tweeted utter joy when Trump contracted the virus, and wished for his death, and got genuinely mad when he didn't die. If lying to pollsters causes democrats to waste massive amounts of money on campaign contributions, so he it. If lying to pollsters causes overconfidence to the point where they don't bother to go vote because Biden has this in the bag, just like Clinton did, so be it. Any chance of civility in politics went out the window with the Russia Collusion Hoax coup attempt, and the Fine People Hoax. Lying to pollsters, by comparison, is a meager pushback to "by any means necessary."

If you don’t think that stuff happens both ways, i have a five-mile long bridge to sell you.


Sent from my iPhone using EO Forums
 

skyraider

Veteran Expediter
US Navy
Offline

muttly

Veteran Expediter
Offline
We can pick individual polls to support one point of view or another regarding poll accuracy. My approach is to cite the aggregate of all polls reported by FiveThirtyEight. Within that universe there are polls that will prove to be close to or exactly on the mark, and those what will be further from or way off the mark. If you want something to be true about polls, you can certainly find a poll to cite as evidence. Given the nature of polling itself, that seems like an unproductive approach.

For the purposes of discussion, I assert that, in general, and on average, and taken as a whole, polls reasonably reflect the voter intentions at the time the poll was taken. I also assert that the closer the polls are taken to Election Day, the more likely it is that the polls will accurately project the Election Day results.

It is not uncommon for polls to say something, and then something happens on the ground that produces a major shift in sentiment, and the next polls show that sentiment shift. Does that mean the previous polls were wrong? No. It does not. It means they were right at the time they were taken, and right again when the sentiment shifted.

Can you find polls that were taken at the same time but disagree? Yes you can. That's where the conversation shifts to the methodology and prior accuracy of the pollster in question. FiveThirtyEight accounts for that by rating the past accuracy of the polls it lists.

I get it, Muttly. You really want the current polls to be wrong and really want Trump to win. You are of course free to pick individual grains out of the greater sandbox to support your desire. I'm not going there for the reasons stated above. Polls are general in nature, not specific. In general, they are accurate, and always, you can find exceptions to the rule. But a small number of exceptions do not disprove the rule. The exceptions are included in the universe of all polls. The exceptions do not disprove the rule because the greater number of polls were accurate or close to accurate.
I posted Detroit News poll, but also the RCP average of polls for the three swing states from 4 years ago and current average if polls. The polls 4 years ago clearly didn't capture closely what the results would actually turn out to be. I don't know much about polling. Just putting it out there. My memory of polling in Michigan for years ago didn't show Trump winning in any polls except maybe one. I don't recall the name definitively. Could have been Trafalger. But it was considered an outlier. The only other thing that I thought may make the race closer that it was polling were the Trump yard signs from 4 years ago. I still though he would lose because of the tv ad advantage Clinton had back then and almost every poll saying she would win handily. I notice the same thing this year with the signs, only difference is there is a bit more Biden signs this time. But more Trump signs this time in a Democratic leaning county. So it should be very close here in Michigan overall. Because the rural areas are going heavy for Trump.
 

Ragman

Veteran Expediter
Retired Expediter
Offline
 

muttly

Veteran Expediter
Offline
Just a coincidence but Trafalger just came out with a poll showing Trump up by one point in Michigan.
 

ATeam

Senior Member
Retired Expediter
Offline
Last edited:

ATeam

Senior Member
Retired Expediter
Offline
As readers of this thread know, I'm posting poll results week by week as Election Day approaches. My purpose is to document the poll results on certain dates so we can see how accurate or inaccurate the polls were after the votes are counted.

There is a wrinkle, however. It's the unusually high volume of early voting. This Forbes article provides the numbers and an informative discussion. With early votes being returned in high numbers weeks before Election Day, it may not tell us much about poll accuracy. The votes that will be counted will not be cast just on Election Day but in the weeks before. We can look at the polls published a day before Election Day and know those numbers. But it appears a whole lot of ballots will be cast long before then.

If the early voting "turnout" reported by Forbes is an indicator, the Democrats have the edge. But that may mean little. We have no way to know if the Republicans will show up in force to vote in person on Election Day or close to it.

I think the Republicans may be hurting themselves by encouraging their voters to vote in person. Long lines at voting locations are already commonplace. It they continue, that may discourage many people of all parties (or no party) from going to vote. With Democrats clearly outvoting Republicans by wide margins where early-vote ballots are now being turned in, it will be more important than ever for Republicans to physically show up, or to quickly change their minds about participating in early voting as the Democrats are.

Another consideration is the invalid ballots cast by early voters. A certain percentage of those ballots will be disqualified for lack of a signature, failure to follow the instructions, etc. How many? I don't know, but this article about the rejected Florida primary mail-in ballots is instructive.

Is this a possible development? A significant number of Republicans who planned to go in person to vote on Election Day hear news of long lines. They've also heard news, consistently in preceding weeks, of Biden's commanding leads in the polls (produced in part by Republicans who lied to pollsters and said they would vote for Biden). Feeling discouraged and not motivated to stand in a long line to vote for a a losing candidate, they conclude their vote won't matter and they stay home.

Note: that last paragraph is pure speculation. I'd be interested to hear speculation about a Trump victory that explains how the Republican vote will be strong enough to pull that off.
 
Last edited:

muttly

Veteran Expediter
Offline
Coincidental with what?

See this page to track Michigan polls and the quality ratings FiveThirtyEight assigns to them.
Michigan Polls

You are right about the Trafalger Michigan poll. Is there a reason you do not mention the six other Michigan polls shown here?
RealClearPolitics - Election 2020 - Michigan: Trump vs. Biden
Coincidence that I mentioned that Trafalgar was an outlier from 4 years ago in my previous post and then they coincidently just came out with a poll recently that again appears to be an outlier.( But is it an outlier? Lol They were the only pollster to predict a Trump victory here in Michigan) I just thought it was funny that I had just mentioned it and unknown to me they would come out with their poll showing Trump with a slight lead in Michigan. Just like what happened four years ago.
No there is absolutely no nefarious reason why I didn't mention the other polls. I figured it was common knowledge that Trump was behind in the polls in Michigan. My only input that I have is what I'm seeing on the ground here in Democrat leaning Macomb County. Specifically just the yard signs. I'm seeing a decent amount of Biden signs, but more Trump signs. One area on my route there are 6 houses in a row with Trump signs in their front yards. Just saying there is a robust amount of support for him in a county that has picked the winner of the Presidency the last 4 elections at least. Michigan is a tossup. Not a 6 or 7 point difference for Biden.
 

muttly

Veteran Expediter
Offline
Since 1912, the President who has faced little primary challenge has been elected 11 for 11 times.
 

Turtle

Administrator
Staff member
Owner/Operator
Offline
Note: that last paragraph is pure speculation. I'd be interested to hear speculation about a Trump victory that explains how the Republican vote will be strong enough to pull that off.
I will speculate that 2020 will at least rhyme with 2016.
 
  • Like
Reactions: muttly
Top