Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
The goal of polls are to be accurate. With a 3% margin of error the expected results were to be Biden winning by as little as 4.2 points and as much as 10.2 points, the final result was +4.5 for Biden. That means that the polls were correct in terms of the overall vote. When you look at the swing states, Ohio was the only one that was outside of the margin of error.
Given that there is over 8 months for things to happen, to say for certain who is going to win is foolhardy. I am not even willing to depend on past precedent to predict what is going to happen. Given all of the things that should have happened in 2022 and in special elections, the only thing that we should be expecting is the unexpected.
Regarding the 2022 polling.
Quote:
The problem for using the red wave that wasn’t as a template for 2024 is that Democrats outperformed narratives and expectations more than polls in 2022.
There was very little in the available polling on election night 2022 that suggested Republicans were on the verge of a 2010-style midterm romp. In fact, the final generic ballot polling average on the much-maligned RealClearPolitics was that Republicans would win the national House vote by 2.5 points. They ended up winning it by 2.8.
Polling as an industry and an art has taken a bludgeoning since 2016. And it is true that some individual state polling was improbably off in 2016 and 2020, largely but not exclusively in the Midwest. But everyone seems to have forgotten that the national polls in 2016 were close to as good as they’ve ever been, with Hillary Clinton winning the popular vote by 2.1 percent against a final RealClearPolitics average lead of 3.2 percent. Nationally, Trump beat expectations produced by forecaster certainty as much as he outperformed his actual polls. And in 2020, Trump ran ahead of his national polling average by a mere 3 points and once again came agonizingly close to victory in key battlegrounds like Wisconsin and Arizona. https://news.yahoo.com/three-comfort...104500704.html
And...
Quote:
Let’s give a big round of applause to the pollsters. Measuring public opinion is, in many ways, harder than ever — and yet, the polling industry just had one of its most successful election cycles in U.S. history. Despite a loud chorus of naysayers claiming that the polls were either underestimating Democratic support or biased yet again against Republicans, the polls were more accurate in 2022 than in any cycle since at least 1998, with almost no bias toward either party.
Of course, some pollsters were more accurate than others. And today, we’ve updated the FiveThirtyEight pollster ratings to account for each pollster’s performance in the 2022 cycle. https://fivethirtyeight.com/features...ling-accuracy/
I was hearing that Trump was up by 30 points and looks like she is losing by 20. Again I'll say that Trump supporters are ignoring the potential trouble that this is showing. Despite the certainty that she would lose, nearly 40% of the people who turned out rejected Trump, this is similar to what happened in Iowa (Haley and Desantis combines) and New Hampshire. There were too many shenanigans in Nevada to really count that.
Trump has tried to sell it as Haley votes are just Dems that are trying to stop Trump, but in NH and SC, Dems had high turnout for their own contests so it is unlikely that a significant portion of Haley votes are coming from Dems. Even if you want to stay that 25% of Haley votes are Dems trying to keep Trump off the ballot, that still leaves 30% of Republican and Independent votes who actively turned out to reject Trump. How many of those you think are going to away in November or vote third party? Again, even if you are generous and stay another 25% will stay home or vote for someone else, that will still leave 20% of Republicans voters that will vote for Biden in a effort to keep Trump out. That could be enough to tip several swing states to Biden even with the loss of previous supporters.
I was hearing that Trump was up by 30 points and looks like she is losing by 20. Again I'll say that Trump supporters are ignoring the potential trouble that this is showing. Despite the certainty that she would lose, nearly 40% of the people who turned out rejected Trump, this is similar to what happened in Iowa (Haley and Desantis combines) and New Hampshire. There were too many shenanigans in Nevada to really count that.
Again, even if you are generous and stay another 25% will stay home or vote for someone else, that will still leave 20% of Republicans voters that will vote for Biden in a effort to keep Trump out. That could be enough to tip several swing states to Biden even with the loss of previous supporters.
I'd be concerned that Ted Cruz beat Trump at AK, CO, IA, ID, KS, ME, OK, TX, UT, WI, WY...
Let's hope Cruz's supporters fall in line with Trump in the General Election... or will they vote for Hillary to keep Trump out?
Now, my point isn't the numbers (outside of Wisconsin) aren't within the margin of error. My point is Trump was behind in all but two of the 2020 battleground states on election day.
Yet he overperformed by at least 2.5 in all but AZ and GA. And AZ and GA where he underperformed it was less than 1%.
Come this Nov. if Biden is behind by say 2.5 in the battle ground states I don't see how he make up the votes given Trump mostly outperformed in 2016 and 2020.
Wisconsin
Trump overperformed 7.2
Pennsylvania
Trump over preformed 2.6
North Carolina
Trump overperformed 2.7
Michigan
Trump overperformed 3.7
AZ
Trump underperformed 0.5.
Georgia
Trump underperformed 0.7.
You can't say that numbers are within the margin of error and that he over preformed at the same time. The Margin provides the range of numbers the final result is expected to fall within. It's not like the line in a football game where if you lose by less than expected it's really a win.
Also in addition to the margin of error, the pollsters are probably using a 95% confidence interval which means that they are expecting that margin of error to be correct 95% of the time.
Finally, state level polling is a tricky thing since it's not done as often a national polls due to costs. I would definitely not trust any state level estimates that are based on a excerpt of a national poll.
That fair, I only heard the 30 points number from one source so I wouldn't be surprised if they weren't quoting the average.
I wasn't saying that Trump was at risk of losing a state like SC in the general, I am just saying that Haley staying in exposes the size of the lack of support for Trump.
Trump needs to project the impression that the entire Republican party is behind him, it would be easy to dismiss 10 or 20% of voters not choosing him, but 40% is harder. Haley staying in until Super Tuesday is potentially going to expose how weak Trump is when the swing state results come in. If that 40% number holds up in Michigan, North Carolina, and Virginia, that's bad news for Trump, I also see Haley also staying in until GA to get another swing state on record. I don't see Trump willing to try to win over Haley voters, after all they already weren't loyal to him, so at best they will stay home and at worst they will vote for Biden, given the margins in the swing states, losing her supporters will cost him the state.
I'd be concerned that Ted Cruz beat Trump at AK, CO, IA, ID, KS, ME, OK, TX, UT, WI, WY...
Let's hope Cruz's supporters fall in line with Trump in the General Election... or will they vote for Hillary to keep Trump out?
Sample of one: I supported Cruz in 2016 and I may write in his name in 2024. Not Trump, though I voted for him in 2020. Not Biden, or whoever the Democrats nominate.
You can't say that numbers are within the margin of error and that he over preformed at the same time.
O.k. Let's stick with numbers and forget under/overpreformed. Trump's numbers in the battle ground states were mostly better than the polls predicted in both 2016 and 2020.
So, come this Nov. if Biden is trailing Trump by say 2.5 or more points in most of the battle ground states. I don't see Biden making up the difference given his final numbers in a majority of the battleground states weren't as good the 2020 polls showed
Biden needs to be ahead of Trump in those states this Nov. If he's not... he'll very likely lose.
Quote:
I would definitely not trust any state level estimates that are based on a excerpt of a national poll
Yet, 538 has stated, "Despite a loud chorus of naysayers claiming that the polls were either underestimating Democratic support or biased yet again against Republicans, the polls were more accurate in 2022 than in any cycle since at least 1998, with almost no bias toward either party." https://fivethirtyeight.com/features...ling-accuracy/
What else is she going to do? She has bills to pay, she has No job. She is getting paid to get beat. Boeing doesn't elect it's board members until the end of Q3 anyway.
That fair, I only heard the 30 points number from one source so I wouldn't be surprised if they weren't quoting the average.
I wasn't saying that Trump was at risk of losing a state like SC in the general, I am just saying that Haley staying in exposes the size of the lack of support for Trump.
Trump needs to project the impression that the entire Republican party is behind him, it would be easy to dismiss 10 or 20% of voters not choosing him, but 40% is harder. Haley staying in until Super Tuesday is potentially going to expose how weak Trump is when the swing state results come in. If that 40% number holds up in Michigan, North Carolina, and Virginia, that's bad news for Trump, I also see Haley also staying in until GA to get another swing state on record. I don't see Trump willing to try to win over Haley voters, after all they already weren't loyal to him, so at best they will stay home and at worst they will vote for Biden, given the margins in the swing states, losing her supporters will cost him the state.
Not really. I watched a bit of her speech tonight. She mentions she will continue to run because she doesn't believe Trump can win against Joe B. Which makes little sense given Trump is polling better than any Rep presidential candidate since Bush in 2004.
I also notice she's not really "attacking" Trump (which is smart on her part) but saying "we can do better." Fine. I also watched a bit of the Trump speech and didn't hear him mention Haley.
If she has has the funding and wants to continue so be it. Trumps best play is to ignore her.
O.k. Let's stick with numbers and under/over preformed. Trump's numbers in the battle ground states were mostly better than the polls predicted in both 2016 and 2020.
So, come this Nov. if Biden is trailing Trump by say 2.5 or more points in most of the battle ground states. I don't see Biden making up the difference given his final numbers in a majority of the battleground states weren't as good the 2020 polls showed
Biden needs to be ahead of Trump in those states this Nov. 1. If he's not... he'll very likely lose.
Again polling does not work that way. Numbers within the margin of error are essentially the same. If Biden was predicted to win a state by 7.5 points with a 3% margin of error, that means if he wins 10.5 or by 4.5 that is expected. The only way to say that Trump overperformed is if he lost by less than 4.5 points.
So if Biden is trailing by 2.5 points and the margin of error is 3 points, it is entirely possible for him to win those states. Polls are just snapshots in time and while pollsters try to select respondents that will allow them to accurately predict, sometimes there is no accounting for the human factor.
Quote:
Originally Posted by john3232
Yet, 538 has stated, "Despite a loud chorus of naysayers claiming that the polls were either underestimating Democratic support or biased yet again against Republicans, the polls were more accurate in 2022 than in any cycle since at least 1998, with almost no bias toward either party." https://fivethirtyeight.com/features...ling-accuracy/
I said that you shouldn't trust state estimates based on data from a national poll because the numbers are too small to make state level estimates accurate. Most polls are based on a sample size of 1000 to 1200 respondents to get about a 3% margin of error. You need at least 385 respondents to get a 5% MOE. A national poll is most likely going to proportionally interview people by state size, so at best in a national sample a mid size state like Michigan will have a handful (10 to 12) of respondents. You can't make statewide estimates on that sample size.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
