PA: NYT / Siena: Fetterman (D) +6

[astrohuncho]

https://twitter.com/Nate_Cohn/status/1587048192079716353
https://twitter.com/Nate_Cohn/status/1587093381603368960

how bad of a reckoning will the polling industry receive after this election?

[Calthrina950]

This actually bolsters the case I make, that NYT/Siena is not inherently more credible than the GOP pollsters who have been decried. No pollster is perfect, and it is best to average them all, and to understand that the only poll which matters is on Election Day.

Great! So looks like Fetterman will be winning by around 4 points, then. Happy we agree.

We don't agree about that. I actually think Oz is going to win now, although I believed for months that Fetterman would. And he sill can, given that the early vote in Pennsylvania looks unambiguously good for Democrats.

"It's best to average them all" (and then completely ignore the average to fit my priors instead)

I just said in this very thread that Pennsylvania is a Tossup and that I wouldn't be surprised if Fetterman won. What I'm saying is that I've changed my prediction over the last few weeks from a Fetterman to an Oz victory. I continue to firmly believe that Shapiro will win without any difficulty.

[Ferguson97]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

[Calthrina950]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

This is not true. Many (such as wbrocks67) would be dividing head first into the crosstabs to find any nugget that would discredit the poll, and assert that Fetterman is faring much better than what it shows.

[Sir Mohamed]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

Yup, that's because polls have a history over underestimating R-support and overestimating D-support. Especially in recent election cycles.

I guess 2016 broke us forever.

[Eraserhead]

The real pollsters have arrived. About time.

[wbrocks67]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

This is not true. Many (such as wbrocks67) would be dividing head first into the crosstabs to find any nugget that would discredit the poll, and assert that Fetterman is faring much better than what it shows.

Sis why are you so obsessed with me

[Amenhotep Bakari-Sellers]

This is over even if we lose NV which we're not it's still 50/50 and Barnes, Ryan, and Beasley can still win and if Rs win WVA we can win FL in 2024 with DEMINGS against Rick Scott whom isn't Latino

DEMINGS looks like she is audition for 24 ANYWAYS

[Orwell]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

This is not true. Many (such as wbrocks67) would be dividing head first into the crosstabs to find any nugget that would discredit the poll, and assert that Fetterman is faring much better than what it shows.

Sis why are you so obsessed with me

You're the best example of his analogy.

[soundchaser]

This is not true. Many (such as wbrocks67) would be dividing head first into the crosstabs to find any nugget that would discredit the poll, and assert that Fetterman is faring much better than what it shows.

Sis why are you so obsessed with me

You're the best example of his analogy.

First of all, this isn't an analogy.

[Amenhotep Bakari-Sellers]

We don't need PA anyways Warnock and CCM solidify the Senate PA and WI, OH, NC and UT are Wave insurance

[kwabbit]

Some issues here (this sample voted for Biden by 4 points) but what's key is that at 44%, Oz being at 50% is outside of the margin of error. He really needs to win over virtually every undecided, moreso than Republicans normally do, which is not an impossible task (especially given that this was conducted pre-debate) but an uphill battle nonetheless.

Is not just trivial to weight on 2020 recall if they have the data? They weighted on 2016 recall in 2020, but perhaps the midterm electorate is too different to do that?

[wbrocks67]

If this showed Oz+6, so many people here would be calling this race a done deal. But because it shows the Democrat leading, it’s treated with skepticism.

This is not true. Many (such as wbrocks67) would be dividing head first into the crosstabs to find any nugget that would discredit the poll, and assert that Fetterman is faring much better than what it shows.

Sis why are you so obsessed with me

You're the best example of his analogy.

I'm not sorry for taking a poll that has Oz at a similar black share of the vote that has happened historically in PA a little more seriously over a poll that has him winning nearly 30% of it.

[Eraserhead]

This has now become a very common and disturbing thing by either the New York Times/Siena or CNN to publish Polls that were done over a week ago.

It's meaningless to publish them. It was also done before Fettermans Debate Performance.

Except if you look at the dates, the poll was conducted partially after the debate. I agree it's a bit dated and yet still some of the newest information we have available about the race.

[Pollster]

Some issues here (this sample voted for Biden by 4 points) but what's key is that at 44%, Oz being at 50% is outside of the margin of error. He really needs to win over virtually every undecided, moreso than Republicans normally do, which is not an impossible task (especially given that this was conducted pre-debate) but an uphill battle nonetheless.

Is not just trivial to weight on 2020 recall if they have the data? They weighted on 2016 recall in 2020, but perhaps the midterm electorate is too different to do that?

Not sure I'm understanding the question right - but a huge issue in 2020 was the turnout surge that made the 2016 recall question somewhat less salient (many pollsters also didn't use it as a weight - the ones that did were somewhat closer to the mark). 2020 polls also significantly overrepresented people who voted by mail which may or may not happen this year given the different circumstances - very hard to tell.

[astrohuncho]

Some issues here (this sample voted for Biden by 4 points) but what's key is that at 44%, Oz being at 50% is outside of the margin of error. He really needs to win over virtually every undecided, moreso than Republicans normally do, which is not an impossible task (especially given that this was conducted pre-debate) but an uphill battle nonetheless.

Is not just trivial to weight on 2020 recall if they have the data? They weighted on 2016 recall in 2020, but perhaps the midterm electorate is too different to do that?

well nate cohn did mention 2020 level nonresponse bias. weighing 2020 recall could have helped

the rcp average at this point 2 yrs ago showed biden +6 in pennsylvania. just saying...

[kwabbit]

Some issues here (this sample voted for Biden by 4 points) but what's key is that at 44%, Oz being at 50% is outside of the margin of error. He really needs to win over virtually every undecided, moreso than Republicans normally do, which is not an impossible task (especially given that this was conducted pre-debate) but an uphill battle nonetheless.

Is not just trivial to weight on 2020 recall if they have the data? They weighted on 2016 recall in 2020, but perhaps the midterm electorate is too different to do that?

Not sure I'm understanding the question right - but a huge issue in 2020 was the turnout surge that made the 2016 recall question somewhat less salient (many pollsters also didn't use it as a weight - the ones that did were somewhat closer to the mark). 2020 polls also significantly overrepresented people who voted by mail which may or may not happen this year given the different circumstances - very hard to tell.

If they’re getting Biden +4 on the recall and believe that the sample is over representing Biden voters, why do they not weight the sample to Biden +0 or whatever is appropriate. NYT already weighted to vote recall in 2020, so why would they stop the practice now? Nate Cohn was so reluctant to publish polling because he feared it would over estimate Democrats, yet he decided to not take this step to partially resolve that issue.

[Orwell]

Failing New York Times poll=add 8 points to MAGA Mehmet to correct it.

This is the most shameful thing anyone has said on this website, maybe ever.

Maga Mehmet is better than Feckless Fetterman though

[Anti Democrat Democrat Club]

Lmao

[Lief 🗽]

Really funny to see the emotional breakdown from right-wingers when they have to step out of their partisan trash poll safe spaces.

[Dani Rose]

I have my suspicions about NYTimes polls having a D tilt (Davids +14 comes to mind), but yeah I'd definitely call Fetterman the favorite.

For what it's worth, NYT actually underestimated her in 2018. Not by much, but it still happened.

In any case, the EV gives me reason to believe this. Oz's number is too low - I completely expect something in the range of 51-49 Fetterman in the end - but I think we've got this.

[Eraserhead]

Really funny to see the emotional breakdown from right-wingers when they have to step out of their partisan trash poll safe spaces.

They've been coddled by them more this cycle than usual.

[GALeftist]

This actually bolsters the case I make, that NYT/Siena is not inherently more credible than the GOP pollsters who have been decried. No pollster is perfect, and it is best to average them all, and to understand that the only poll which matters is on Election Day.

I'm going to take this opportunity to soapbox about this attitude, which I have been seeing extremely frequently as of late. I hope it's not taken as a criticism of Calthrina950 per se, who I think is a good poster and a benefit to the forum.

I think a lot of people are almost traumatized by the polling misses in 2016 and 2020 and are overcorrecting by giving pollsters like Trafalgar and even Big Data far more weight than they ought to be given. Now I'd like to preface this by saying that the polls could obviously be systemically biased against Republicans. (The same could be said in favor of Democrats, though I think it's less likely.) But accepting polls like Traf and Big Data is not an effective prophylaxis against that. Why? Because their methodology is just not sound. Take Trafalgar, for example. It was recently revealed that Cahaly has a very "unique" method of avoiding nonresponse bias: he chooses a sample of voters which he thinks are representative and just spams them with calls until he gets a sufficient number of responses. This is why his response rates are so similar: he's actually targeting a rate with each poll he conducts. As a result, his polls are, in my view, reflective of almost nothing except Cahaly's educated guess as to the outcome of a given election. Sometimes it will happen to be more accurate than a polling aggregate, but this is not the slightest bit indicative of any methodological legitimacy – if the polling bias disappears or changes, which could happen very easily, poof, its accuracy is gone. Big Data is even worse in this regard.

Now, I think the response of the aforementioned traumatized would be, "well, I care about accuracy, and the supposedly methodologically sound polls clearly aren't accurate anymore." I think this is a defensible position, and my advice would be to simply throw polling out altogether if you believe this. Trafalgar and Big Data won't save you; there's no reason why Robert Cahaly or Rich Baris can produce better educated guesses than you can on your own.

[Amenhotep Bakari-Sellers]

https://twitter.com/Nate_Cohn/status/1587048192079716353
https://twitter.com/Nate_Cohn/status/1587093381603368960

how bad of a reckoning will the polling industry receive after this election?

Yeah Shapiro winning by 10 has something to do with Fetterman winning lol this is PA it's not Alabama

We are gonna win WI and PA because Fetterman and Barnes are still LT Govs Oz and Vance and Budd aren't office holder that's why OH, NC, Wzi and OA are gonna elect Ds that are office holder

Rubio is favored because he is Sen

[Stand With Israel. Crush Hamas]

Just not relevant anymore, but I do think Fetterman is still slightly favored.

No one is going "Well, Fetterman probably isn't fit to serve, so I guess I'll vote for Oz.". Their politics are too different. They're shifting the vote calculus in their head to Oz vs. Josh Shapiro's Senate appointment.