PieFedDonald Trump's Chances of Winning Election Are Declining
www.newsweek.com/donald-trump-polling-data-five…
Polls are crap
It’s a little bit notable how they change over time, but mostly it is just a bunch of poor methodology noise
But, I do feel it’s relevant to point out that even in polling land Biden has been gaining for the last few weeks and it’s notable which media outlets are pretending it is the opposite of that (like severely the opposite) and to wonder why that is
62 Comments
Comments from other communities
Personally, I don't like to trust commentary on FiveThirtyEight's tracker from anyone other than FiveThirtyEight. They do a fantastic job of tempering expectations and not putting too much stock in momentary changes in the tracker. The current numbers are basically the same as they were on May 11:
It hasn't changed much, even after all the wild shiz that has happened since May 11. Furthermore, the polls usually work on a delay, so they haven't yet factored in the assassination attempt, or the RNC. And often, the tracker stays at a dead-even heat largely because their statistical model factors in uncertainty from the 4 months between now and Election Day.
All of that is to say: the race is still essentially dead-even. Like Chris said in the OP, the most important thing to do is vote and encourage others to vote, no matter what the polls say.
The RNC has 6 Million LESS viewers than 2016. That's quite a swing!
Republicans still show up to vote and reliably vote party line. The only way to beat them is motivating people to vote against them in battleground states.
Less viewers because fewer people are engaged or because people already know who they’re voting for?
Republicans vote Republican. 2016 was an anomaly because Trump was an unknown factor and a voice for change in the RNC. I doubt many people’s opinions on Trump have shifted since 2020 and the Republican candidate was a foregone conclusion.
Exactly, I think because races have been so close lately, and the probabilities are ending up close to 50% often, people sometimes unintentionally conflate them with poll numbers. 53% to 46% would be a massive poll lead. For probabilities though in this situation it's the same as saying they have even odds of winning. Look at those massive 95% confidence intervals, the race is in a statistical dead heat. It's kind of remarkable how steady it has been despite all the wild events that have happened.
538 has an explanation on their site about why it's not moving. Essentially they're so far out they heavily discount current polls and events. We're effectively seeing their baseline result with some previous races and economic measures tweaking things a little.
Yes, that's true. The poll averages themselves haven't moved much either though. And the reliance on the fundamentals forecast has me nervous, but they definitely do it for a reason. When they developed the models and looked at poll history the pattern they found was the fundamentals had a big influence on what the polls would look like closer to the election and the eventual result. Polls closer to the election are more predictive than the fundamentals. Polls farther away from the election less so. There's at least some reason to think things have changed enough maybe the fundamentals aren't as fundamental for this race, but I guess we won't know until afterward.
It's a new model this year, as Nate Silver took his with him when he left 538. The new one seems to put a lot of emphasis on "the fundamentals" this far out, that is, it "thinks" that the general environment and economy and such is pretty good for the incumbent and that the polls might move in that direction by the time election day comes along. And since it's fitted to historical data, it's also implicitly assuming that this election will be similar to past elections (like, say, including a competent campaign by a candidate who can get out there and effectively communicate accomplishments and a plan for their term).
I personally think those assumptions are pretty clearly wrong *this year* and so I'm more inclined to base my perception of the race on pure polling averages, which are looking quite bad for Biden.
538's win projections aren't really based on current polling or really current events at all, that's why nothing's moved despite crazy things happening in the race. They've got some other fundamentals in the model that makes them think it's all going to revert to norm and Democrats will win, but it's just their own guesswork, not anything particularly well supported.
Assassination attempts don't correlate with wins per se. Neither does debate performance.
Popularity doesn’t even correlate with wins. There are people who will do whatever they can to prevent Trump from being elected, there are people who will do whatever they can to prevent the radical liberal cabal from being elected, and there are those who won’t vote (for one of a multitude of reasons). Everyone else is statistically insignificant.
🤖 I'm a bot that provides automatic summaries for articles:
Donald Trump's chances of winning the November 2024 presidential election are declining, according to a new aggregation of polls. It comes after a Presidential Voting Intention poll of 3,601 swing state voters by Redfield & Wilton Strategies, found that Trump's margins over President Joe Biden have narrowed since June in two key swing states: Florida and North Carolina. In a post on their website, FiveThirtyEight said their latest predictions can be explained by the flurry of swing-state polls that show Biden is encroaching upon Trump's lead. In the week before the shooting, national polls had Trump as the favorite to win the White House, leading narrowly in the six key battleground states of Arizona, Georgia, Michigan, Nevada, Pennsylvania, and Wisconsin. An ABC News/Ipsos/Washington Post poll released Thursday found Biden and Trump are tied among registered voters. Biden's fitness for office has been called into question since his shaky debate performance against Trump in June in which he muddled his words and appeared to lose his train of thought. Saved 75% of original text.Click here to see the summary
It is July. There is August, September and October.
Do not believe the polls! Prepare to VOTE! Tell others to VOTE!
There is no leaning back or relieve until November 5.
Do not believe the medias polls. Especially not Newsweek.
GO OUT AND VOTE!
100% agree
Just requested my absentee ballot yesterday.
If you can vote by mail, do it. Avoid the intimidation at the lines if you can.
But if you can't, go in groups. Take as long as necessary, do not be deterred
Just vote, doesn't matter the polls.
*Oh but Biden / Trump has enough vote to win*
So? Add another vote.
Vote in numbers too large to manipulate.
Even if your presidential vote does not matter, you also vote for down ballot candidates. They are the ones that have the biggest influence on your life. Your state senator can write laws to protect you or harm you. Your town's sheriff gets to decide if your local police officer should be fired and who should be hired instead.
Vermin Supreme 2024!
I refuse to go out and vote.
Because I'm getting a mail-in ballot.
I would strongly suggest voting in-person if you happen to live in one of the swing states that has been clamping down on absentee ballots.
Thankfully, I'm not. But that's good advice.
I live in a mostly Republican district that will inevitably vote Trump, I am still voting against Trump no matter what, regardless of what any polling says. American democracy as we know it ends if Trump wins
There's a fair possibility that lots of "mostly republican" districts are relying on voter apathy to stay mostly republican. Good on ya for doing your part!
Don't forget the voter suppression efforts! Make sure you're registered, because the fascists are going to challenge as many as possible.
There are about 5 polls that actually matter in this election.
Pennsylvania (whoever wins this will likely take all), Michigan, Arizona, Wisconsin, and Nevada.
Georgia, North Carolina, and Minnesota might matter but I don't see much chance of them voting Blue, Blue, or Red respectively.
Wherever you are, just go vote to make the House as blue as you can. Mark down the democratic nominee for president while you're there.
And even in those places, Trump supporters are known for not wanting to admit to pollsters that they support him, so they're not super reliable
On the other hand, Biden supporters are known for not answering phone calls from unknown numbers which makes them even less reliable.
Yeah, maybe it evens out, maybe not.
That is one of 10 various reasons why the polls are not reliable in various directions
Don’t care. Just vote.
Yup, only one thing affects his chances of winning, and that's voters in swing states. If all the ones that don't like him vote for Biden, he'll lose easily. If they don't, he has a good chance to win.
Everyone that stays home or votes for Trump because Biden is too old, will be complicit in the ending of American Democracy and the deaths of thousands. Don't think it can't happen.
Why? Does he have some bad qualities?
Idk... someone asked him recently, and he could only rattle one off the top of his head.
Jesus Christ
No of course not. He's an angel who respects all people and the law.
Obviously the only poll that matters is the actual election, and that's months away, and anything can happen, and everyone needs to do their part no matter what.
That said, I would expect Trump's numbers to slide a bit from time to time as we get closer to the election. The closer we get to November, the more we'll see independent and apathetic voters start to pay a little bit of attention, and there is just so much not to like with Trump. The MAGA-dominated GOP is incredibly myopic about how their views will be seen by the average American living outside the right wing echo chamber.
And as much as it would help them, I don't think Republicans will be able to pretend to be moderate or sensible, even for the sake of the election. There is a perverse incentive to be as extreme as possible in order to gain support from other extremists and win influence on the right, but that behavior drags the party as a whole down when it is seen by anyone who isn't already part of the base. For example, MTG gets national support and tons of media attention despite being a representative for a little rural district that no one gives a fuck about, all because she is a loud and controversial extremist. Anyone who moderates their tone or stops themselves from saying the quiet part out loud will lose influence to those who proudly proclaim their awfulness.
None of this is to say that victory is assured. Hell, I'm not even saying it's likely (nor am I saying that it's unlikely). I'm just saying that I think that there's good reason to expect Trump's poll numbers to dip from time to time.
Polls for a close race are never going to be great. That’s just how statistics works.
America is fairly split down the middle nationally, and winning or losing can come down to who gets people to turnout on election night in a couple purple states.
Even with Clinton V Trump, the polls were not wrong. They showed a close race, and 538 literally said that it was a 4 sided die roll with Trump on a side.
Florida is a swing state? Ha, good one!
Polls currently have it closer than it's been in a while. Biden might not win there, but more competitive battlegrounds means more voters, and more voters is good for democracy.
The anti abortion people figured this out 40 years ago, and they kept pushing for 40 years, and look they got their victory, against impossible odds.
The fairly normal things the left wants to do would be a *lot* easier to achieve if we could display the same level of stubbornness.
It's hard to be stubborn about politics when the closest thing we have to an ally shares most of its campaign financiers with the enemy.
Biden: A trillion dollars spent on climate change, student loan forgiveness, huge increase in corporate tax to pay for all of that, income inequality is dropping, support for unions, basically the country moving 1% in the right direction for the first time since IDK Carter or something
Trump: Let’s shoot all the Mexicans
Lemmy: They’re the same picture
Lets compare apples to apples, shall we?
With friends like these, who needs enemies?
Just like a non stop avalanche of bad faith bullshit from you guys, huh
I dealt with the immigration issue at a lot more length back in my comments from a while ago; I don’t really want to dig it back up and copy paste it, because what’s the point. But short summary, you’re wrong.
Lots of people are tired of culture war bullshit. Not the diehards, but the moderates. All they see are library book bans while their home insurance prices going up 5x.
I always believe that these articles are written by Republicans, in order to convince people that Biden doesn't need their vote.
https://mediabiasfactcheck.com/newsweek/
I can never believe that news has biased people and goals. It's pathetic, as news should be reporting on issues and stating facts
At the start of the year, I gave 95% odds that Trump would lose the election. So far, the biggest shake up has been the assassination attempt which didn't give him a boost in the polls. I think the polls showing Biden losing are flawed because those showing him doing so more than the margin of error have 18 to 20-somethings voting more for Trump. However, I also believe that the way the polls change can inform us about changes to the political climate. (Systematic sampling bias should still show changes that have a global effect in the intended dataset.)
Furthermore, Biden *already won* against Trump before Jan 6, which showed every single warning about him was true. Right now, our biggest threat is complicity and voter suppression. Register to vote, encourage other people to register to vote, call everyone online who tells you not to vote for Joe Biden a Russian Bot, and everyone in meat space whatever name will hurt them the most (traitor is a good one, fascist is another.) Register, organize, encourage, *show up.* Vote early if you can. I have arrangements with people in my area who are able to taxi people to their polling place for free; consider doing the same.
Gave up on FiveThirtyEight after 2016. Once again, it's all about turnout.
2016 was what gave FiveThirtyEight credibility. They’d given the highest chance to Trump, IIRC, out of the aggregators.
If I am remembering correctly then fivethirtyeight gave him a ~33% chance of winning, which was absolutely within the realm of possibility
I think it may have been around 20%, but 1 out of 5 is still reasonable.
I only trust the casinos at this point. What are the current odds?
Betus is giving 2:1 odds on the Democrat, I.e. they think the Democrats have a 33% chance to win.
Lol, although they're probably as accurate as any poll.
Casinos are only worried about the juice, they could care less about the outcome of a bet.
Damn it, you are right. Casinos only goal is to make profit from the bets. The current odds only indicate bets already made.
Honestly, casinos have a reason to try to get it right. When pollsters get it wrong nothing really happens. I more or less agree with the logic of just looking at the casino odds as as much of a thing to trust as anything else.
Also, it’s apparently illegal to bet on the presidential election in the US, which I didn’t know.
This is the best summary I could come up with:
Donald Trump's chances of winning the November 2024 presidential election are declining, according to a new aggregation of polls.
It comes after a Presidential Voting Intention poll of 3,601 swing state voters by Redfield & Wilton Strategies, found that Trump's margins over President Joe Biden have narrowed since June in two key swing states: Florida and North Carolina.
In a post on their website, FiveThirtyEight said their latest predictions can be explained by the flurry of swing-state polls that show Biden is encroaching upon Trump's lead.
In the week before the shooting, national polls had Trump as the favorite to win the White House, leading narrowly in the six key battleground states of Arizona, Georgia, Michigan, Nevada, Pennsylvania, and Wisconsin.
An ABC News/Ipsos/Washington Post poll released Thursday found Biden and Trump are tied among registered voters.
Biden's fitness for office has been called into question since his shaky debate performance against Trump in June in which he muddled his words and appeared to lose his train of thought.
The original article contains 683 words, the summary contains 168 words. Saved 75%. I'm a bot and I'm open source!
With all the propagandists on higher alert and posting fervently- I suppose it’s a bit of a relief to see that even though the polls can’t be believed, that don’t ALL show the same bullshit attempts to cash in on clicks regardless of the damage.
Remind me what 538 was predicting the last time Trump won
Thanks for link! I'll save others a click:
Site’s still up
Trump had a near 30% chance per their analysis, which is pretty significant
Edit: so yeah a higher chance this time wouldn’t be great
Deleted by author
This one?
https://www.natesilver.net/p/nate-silver-2024-president-election-polls-model
Show Biden behind, but not down to 20%. Why lie about that?
That page looks to only show polls and has the actual model paywalled?
Deleted by author
That's from July 3rd...
Deleted by author
The same source shows that Biden is up 2% since the 3rd, and a bit more over the past 10 days.
Deleted by moderator
The results are on a paid blog. At least I'm not aware of it being available anywhere for free. He posts on formerly Twitter, though.
Oh wow, I didn't know he left. I'm going to trust 538 much less.
Just take 538 in aggregate.
Nate Silver still forecasts. You're not losing anything.
They don't though, they only care that the money bet equals the money paid out minus the house cut (juice). When the line changes, it's not because the casinos believe one side has a better chance. It's because the money being bet on one side is uneven. Unlike their customers, casinos don't like gambling their money.
?
That’s not how it works
You’re betting with the casino here, not directly with other users. When you bet with the casino about who will win the election, the casino’s “house cut” is baked into the non reciprocal nature of the odds on the two sides (plus whatever fees). If they have the odds wrong enough, then their expected value can go negative. As you said, they don’t gamble, so they do their damnedest to make sure that doesn’t happen.