Trump vs. Harris - Please put all new polls here.

I am not a fan of the betting sights quoting odds. Just give me the data. lol

 

His model is supposed to prevent large swings causes by "flipping" close swing states, because it isn't all or nothing. He proportions votes, so if a swing state with 20 ECs is 55-45%, then he assigns 9 votes to one candidate, and 11 votes to the other, for example. If that state flips to 45-55%, then it only changes the model by 2 ECs. That's why his model has fractional EC votes. It's supposed to keep the model stable. I agree that Trump should never have been favored at +20%. I think the real reason for his drop is that Silver was applying a "convention bounce remover", taking something like 3 or 4% off Harris's tallies for weeks after the convention. I think that wore off about the time the post debate polling swung to Harris's favor.

 

Thanks war, that makes sense.

 

Silver is not even pricing in an anti-Trump polling bias that has been clearly present in the past two elections. I understand why he can't do that, but I also understand it's risky to assume pollsters who underestimated Trump's numbers in 2016 and 2020 are suddenly going to nail it on the head in 2024. And with all of the media hype and kid gloves with Harris, you almost have to assume it's going to happen again. This makes it especially difficult for analysts like Silver, because they are at the mercy of imperfect polling. Harris is way behind where both Clinton and Biden were at this stage of the race in polling data, which is why models like Silver's can show a >60% chance for a Trump victory. I'm willing to accept his lower chance of victory as of today for Trump. I mean, this guy gave Trump a 20% chance to win the day before Election Day 2016. A 48% chance in September is definitely not a death knell. Especially when he had Trump with a 65% chance just a week ago.

 

i’ve explained in the past why it’s risky to employ intuitive corrections to algorithmic models. It was the case that Trump well over performed his polling in 2016, but it’s not necessarily the case that will always be so or by the same amount. i Don’t think I would do such correction either, but prediction markets will of course be able to put such a correction in. oddly, they seem even more optimistic about Harris’ chances than silvers model does, so I’m not sure so has a real problem here.

 

Some of the comments in this thread are hilarious. Somebody actually thinks Harris is closer to flipping Florida than Trump is to flipping Wisconsin. Do yourselves all a favor and lay off the MSNBC for a while. That joke of a cable news network eats the literal brain cells of its viewers.

And to all my nutclingers out there, Biden’s still got this one, right?? LOLOLOL

 

It wasn’t just 2016, homie. She’s needs about 6 points in the national polls consistently to feel good about winning this and right now she’s at about +2. And that’s based on realclearpolitics aggregate of polls, which overestimated for each of Trump’s opponents in 2016 and 2020.

 

Say, what happened to your daily update of the betting odds? What happened to your daily update from Nate Silver? Post his graph of how his win probability has changed. I want to see the crash.

 

Kamala is going to win. Madame President.

 

Again, if all of this so obvious, why are the betting odds in Harris’ favor?

Vox has a nice article describing the challenge of using past election results to predict the future. By trying to solve the last problem, the correction you apply might actually cause your polling error, as the world changes.

The phrase I heard most in my conversations was a worry about “solving for the last problem” or “fighting the last battle.” In other words, lessons have been learned, but will those lessons apply this time around?

In 2016, for example, pollsters addressed some of the reasons they overestimated Mitt Romney’s performance in 2012 but missed that state-level surveys were overrepresenting college graduates. That miss ended up artificially boosting Hillary Clinton’s support, especially in the Midwest battleground states that proved decisive.
​

It should also be noted that pollsters are actively trying to fix these recent problems:

At the same time, pollsters seem to have learned from the polling misses of the past. They are more conscientious about reaching the hardest-to-reach voters, have reconfigured the way they run their operations, and feel pretty good about capturing a snapshot of a political sentiment in time. They point out that the alternative to a poll-filled world is one where we’re all left trying to define “vibes.”​

I’m not saying that Trump won’t overperform the polls in 2024 but just that this result isn’t guaranteed simply because it happened in 2016 and 2020. The variables are too many. Did the changes in polling methodology fix the problem? Are Trump voters still hard to contact? Do Harris voters have a similar issue? Does the 2020 census make weighting of subgroups more accurate? Might there be new subgroups we’ve not yet seen?

 

@okeechobee Is digging themselves a hole so deep it's almost sad

 

Today's betting odds and if I recall our good friend @vegasfox has made the argument that the betting markets and especially Polymarket are more accurate than the polls although noting that he made that argument when the betting markets favored Trump over Biden and later Harris.

 

Harris +5 from a highly rated pollster.

Poll: Newly popular Kamala Harris builds momentum, challenging Donald Trump for the mantle of change (nbcnews.com)

Harris +4 from another highly rater pollster

Harris shows some gains and economic views brighten a bit — CBS News poll - CBS News

 

The 'Trump voters are harder to reach excuse' is major sus. How do they know they're not able to reach them? Don't they have to reach them to find out if they support Trump in the first place? Seems to me the only method they could use to combat this faux issue is to sample the proportionate amount of likely voters based on party registration. Perhaps filtering out respondents who didn't vote in both 2016 and 2020 and then matching the state's share of each party and independents. So, I don't buy that excuse. If they're "guessing" who that person they're trying to call was going to vote for Trump or Kamala, then they're injecting prejudice. I don't see how prejudice helps a poll be more accurate, at all.

There were several factors at play in 2016 and 2020 which are still likely present in 2024 and perhaps even some new biases which exist this cycle that did not before:

There are a small percentage of polling respondents who lie about not voting for Trump. More often these individuals state they are "undecided" and then when they vote, they vote for Trump. That's the pitfall of the media narrative that Trump is Satan. It is what it is. You have to account for that. And of course, the media (Vox, whomever) will never admit this to you, because then they'd have to admit bias and they've never do that.

Also, because Trump is such a known quantity (this being his 3rd election), any "undecideds" out there are people who are at least open to the idea of voting for Trump and because his is the known quantity and generally has better numbers on the economy, he'll do well with those undecideds. If you're never Trump or abortion is your top issue, you've likely already made up your mind about Trump. So Kamala doesn't stand to gain much and her vagueness on answers isn't doing her any favors with undecided voters. The more interviews she has where she literally gets combative as she is evading a direct answer is just going to make it worse.

Last but not least, there's been an ongoing effort by the media to prop up Biden and now Harris beyond what constitutes ethical journalism. So, it's reasonable to conclude the vast majority of pollsters want Kamala to win and it's really hard for journalists with such biases to give Trump credit for anything, let alone good polling numbers.

As far as the betting sites, Trump has outperformed them more than he has the polls. At last check they are still giving him a 46.3% chance to win which is way higher than he had with the gambling sites in 2016 and 2020. He is clearly, if nothing else, in far better position now than he was in 2016 and 2020 from a polling and oddsmaker standpoint. That much is undeniable. 46.3% is still a lot of people to risk money on a bet. That would never happen if he had no hope of winning. And again, just two weeks ago, he was in front. Given all that's happened, if you're a Trump supporter, you should be very pleased right now. He always closes the gap down the stretch.

 

When you see a poll like Harris +5 in PA and the same polling outfit told us in October of 2020 that Biden was +13 in PA, only a Neanderthal wouldn't question that poll. Especially as said poll is an outlier. Especially in a state that has been steadily closing the gap of registered voters in favor of the GOP for the past 8 years. That race is a virtual deadlock right now, which is not good for Harris. If you don't understand these dynamics, it's not that you can't, you simply don't want to.

 

And if you think that 2024 is like 2020 or 2016 you're the one who is delusional. Trump is much weaker than he was in either 2016 or 2020 and no matter how much you and your fellow Trumpers try to demean her Kamala Harris is a stronger candidate than either Joe Biden or Hillary Clinton. For a better analogy I would suggest that you look at Barack Obama's candidacy in 2008.

 

Another highly rated pollster pulls a +4 for Harris in Pennsylvania, post debate polling.

Harris, Casey lead Trump, McCormick in Pennsylvania poll • Spotlight PA

 

Also, Pennsylvania polling looked pretty good in 2020, the averages being perfect ... this why I trust the averages a lot more than a single poll. Also worth noting that the consistently republican bias Trafalgar and Insider Advantage missed about the same on the low side as the most optimistic Biden poll on the high side, but still accounted for a better picture of the final result when averaged in, which I guess is a reason why Nate Silver opposes banning them completely.

 

The three major differences since 2020:

1) Trump is no longer the incumbent

2) January 6th happened - clearly incited by Trump

3) Trump cost the GOP the 2022 mid-terms

I really don’t think it will be all that close. She’ll win by more than Biden did.

 

I would add a fourth difference between 2020 and 2024,in 2020 the Supreme Court had yet to issue the Dobbs decisions overturning Roe v. Wade. Reproductive rights wasn't even on the ballot as an issue. Dobbs was a major factor in the Republican under performance in the 2022 midterms. In virtually every state referendum on reproductive rights the so called pro-life advocates lost.
Abortion rights have won in every election since Roe v. Wade was overturned

While an extremely anecdotal example Virginia's off off-year state legislative election in 2023, may foretell what could happen in the presidential election. Republican candidates in swing districts made crime their prime issue while Democrats in those districts ran almost entirely on the issue of reproductive rights. The Democrats ended up with majorities in both Houses of the State legislature flipping the state assembly the lower house from Republican to Democrat. The February 2024 special election in the New York third district to replace George Santos could also be telling. The Republican candidate ran on almost entirely on immigration. Tom Souzzi the Democrat rather than running from immigration emphasized that the Republicans at the direction of Trump blocked enactment of the most consequential border security bill in decades. Souzzi won by an eight percentage point margin.
Unpacking NY-3 results: Takeaways from and reaction to Suozzi's win

At the same time, he attempted to go on the offensive on an issue that Republicans often claim as their own: immigration. He criticized the GOP for effectively killing the bipartisan border deal negotiated by senators. “Suozzi showed other Democrats the way to engage on immigration, which was not to ignore the issue but to talk about it in a way that I think was probably appealing to the middle of the electorate,” said Kyle Kondik, managing editor of Sabato’s Crystal Ball.​