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

see my post just a short time ago.

 

Looking at the poll right now and for some reason they oversampled Republicans. Doesn't make sense to me

 

You are just cherry picking 5 polls to make a fake point. 538 analyzed over 120 final polls from each pollster. Do all 120+ polls then compare. Trafalgar has the second lowest rate of picking winners… because they say republicans are going to win just about every race. Trafalgar sucks because they are intentionally bias. NYTs is likely higher rated because their misses are split equally and average out.

From what you posted, while both pollsters were equally bad, the NYTs was 6 for 6 at picking winners. Trafalgar was 1 for 6… the results speak for themselves…

 

ummm no you did.
You pointed out traflager and 528

I pointed out that NYT who per 528 is the best pollster was actually off by not only the states you said where it matters but they also missed on the GE vote as well versus traflager.

The reason this is important is because if NYT continues their bias with how they’re weighing their polls, this is going to be a crushing defeat for Harris.

 

Yeah, I mentioned Trafalgar's low rating on 538 based on the 120+ polls they analyzed. You countered with 6 cherry picked polls that they called incorrectly but were marginally closer than the NTYs in most cases ...

The thing about non partisan pollsters is that they just want to be right, so NYTs probably adjusted their methods to account for more Trump voters ... partisan pollsters, like Trafalgar, only care about showing their guy is winning, and will adjust their methods to make that look true.

You are welcome to believe RCPs polling averages if you want, and maybe they are right if the hidden excitement for Trump is still there, but I wouldn't go "all-in" on believing a model that is influenced by partisan pollsters by a factor of 2 to 1...

 

The 4 most recent national polls posted on RCP this morning have Trump winning the popular vote. Someone explain to me how in the world Kamala wins this race when Biden needed to be >4% ahead of Trump in the popular vote just to barely skate by and win the EC in 2020? These polls would literally have to be 6 to 7 points off in Trump’s favor just for her to have a realistic shot.

 

Thanks surfn. You may be right, but I don’t think you or I could know this without seeing the entire analysis that led to the rankings. Eg it’s possible that their rankings include 18 elections going back 20 years, and the results of this one election weren't representative of the larger signal. It is also true that these kinds of rankings are always going to include some level of art, so disagreement among analysts doesn’t necessarily imply any intentional wrongdoing.

 

There have been a couple articles about how Republicans have lost their EC advantage to Democrats because of the shift in voters with Trump leading the Republican Party. If you google that, you can probably find the articles.

 

Yeah. I don't have a position on it, but have heard this discussed as well. An article for those interested...

https://www.washingtonpost.com/politics/2024/10/17/campaign-moment/

That Trump benefits from the electoral college has long been an article of faith. He, after all, was elected in 2016 despite losing the popular vote by more than two points. Trump then in 2020 lost the popular vote by more than four points, but lost the “tipping-point” state — Wisconsin — by less than a point. (The tipping-point state is basically the state in the middle of the results of all 50 states that gives the winner their 270th electoral vote.)

As The Washington Post’s Lenny Bronner noted a couple months back, Republicans benefited from a nearly three-point electoral college bias in 2016 and a nearly four-point one in 2020 — the largest such advantages since the World War II era. That bias measures the difference between the winner’s margins in the popular vote and the tipping-point state.

***

Why might the electoral college edge be narrowing for Republicans? Bronner pointed to Democrats appearing to bank fewer votes in large states, which pulled down their share of the popular vote but didn’t really impact the electoral college. The Times’s Nate Cohn similarly noted that Trump seemed to be doing better than he previously had in noncompetitive states where Republicans made some of their bigger gains in the 2022 midterm elections.

***

That doesn’t mean it won’t matter; in a very close race like this one, even a small electoral college bias could mean the popular-vote loser is elected for the third time since 2000 (remember Bush v. Gore?). But at least for now, Democrats’ popular-vote promoters such as Walz don’t seem to have quite as much to fear from the electoral college.

 

Im also not sure that the Republican EC advantage will disappear this year, but I know that it will disappear some year. And to my lights, that will be a welcome development indeed.

 

RCP posts polls in the order that they are released with the most recently released polls listed first. 538 post polls with the most recently sampled listed first. There is a difference.

 

Ehh, so you want me to believe a couple of left-leaning sources that the Trump electoral college advantage, which actually grew from 2016 to 2020 in Trump's favor has now all of the sudden vanished? I'm sure you realize that 4.4% of the total vote in 2020 was ~7 million voters. California and New York have lost some voting age population, but it's not that stark and we know Republicans aren't moving to those states in droves. The only reason the EC advantage exists today is via Trump, not the GOP. Trump has been able to flip states like PA, MI, WI and OH into high % chance of victory states. PA, MI and WI are traditionally blue states that have Dem governors and Dem state congressional majorities. Other GOP may have tried, but have largely failed to win in these states.

Bottom line: if Trump splits the popular vote with Harris 50/50 he inevitably wins all 3 of WI, MI and PA and thus the election in a EV landslide.

 

Ok, let's say hypothetically it's 2.0% now. It was 4.4% in 2020, but let's say it's gone down two and a half points. All four of the aforementioned national polls have to be off 4 to 5 points, outside the margin of error, in Trump's favor just for Harris to tie. The Nate Silver's of the world are clearly factoring in a significant EC advantage to Trump this election. Perhaps not as big as 4.4% but big enough to have a very significant impact on the final result.

 

So being more accurate in not only the GE but most of the battleground states means nothing huh?

Look, could NYT have changed some of their Methodology and be more accurate this time around, of course. Their cross tabs have been analyzed by others who know much more about this than I do and suggest they have not.

If that’s true, than Harris is in trouble as it’s only the left leaning polls like NYT that are showing a much closer race this time compared to their 2020 and 2016 polls.
Also, the RCP aggregate is very good but obviously not perfect.

Right now the GE on RCP is Harris + .2

Biden was + 8.1 and Clinton + 5.5 same day those election cycles.

 

I guess it depends on how you define 'more accurate' ... they were marginally more accurate percentage wise on a couple polls in one election -- but called 5 of the 6 wrong. That hardly tells the story. I already said several posts ago that Trafalgar and Rasmussens claim to fame is that they look like geniuses when republicans outperform. They tout those moments and live off it for the next few cycles. They look like idiots the other 75% of the time.

With a 50% correct pick ratio, Trafalgar polling is no better than random guess ... or a monkey flipping a coin to pick winner and losers. Insider Advantage, at 47%, is worse than random guess .... think about that.

 

Oh sure. I think it would be pretty surprising if Harris won the popular vote by 1.5% and won the election. Of course, this result will happen one day in the future, and whatever day that is, it will likely be surprising.

 

Yeah, I mean you can still see the advantage in the EC in favor of Trump this election, because if let's say Trump and Harris tie in the national popular vote, that means Trump would have inevitably (perhaps I should say much more likely than not) have won all 7 battleground states. Which would give Trump 312 electoral votes to her 226 in the event of a 50/50 tie in the popular. That tells you all you need to know.

Frankly, kudos to Trump for figuring out how to win the presidency with such a lop-sided loss in the nationwide popular vote. Bush did it, but he did so barely and with not nearly the same amount of sheer exploitation of market inefficiency. Perhaps Trump is smarter than many want to give him credit for.

 

Just a reminder in the 2022 midterms the Democrats significantly outperformed the polls in all three of those states and even in WI where Ron Johnson managed to squeak by in his bid for reelection he still underperformed the polls. And let's not forget the 2023 Wisconsin Supreme Court election.
Wisconsin's recent blue-red voting trends solidify in spring 2023 election

 

Two high quality and local surveys in Michigan by MSU and PA by Muhlenberg College have come back looking really good for Harris.

Northhampton county (Allentown PA) has voted for the winning president for 100 straight years. Muhlenberg College has Harris over 50% and +3 over Trump there

MSU has Harris at 52% and +4 over Trump in MI.

7th Congressional District Survey - October 2024 | Muhlenberg College
MSU Survey: Harris Leading in Michigan | IPPSR