PA- Wick polling- Oz +5

[AncestralDemocrat.]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

I wouldn't bother.. he did exactly the same nonsense in 2020 and got clowned upon as a result.

Never learned.

[Calthrina950]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

I wouldn't bother.. he did exactly the same nonsense in 2020 and got clowned upon as a result.

Never learned.

Obviously, he didn't. I personally thought Biden and Democrats would do well in 2020, and certainly have learned from that. But this is why I want for Election Day to come, and to come soon. Hopefully one narrative or the other will be proven.

[wbrocks67]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

No, because there's a thing called outliers that should be treated as such. Should we just take every single poll at face value? If a poll has 80% conservatives in a state that usually has 40, should we just accept it like there's not anything questionable about it?

There's nothing wrong with being smart about polls and calling out things that are clearly off. That AZ poll is certainly favorable for Democrats but I'm going to call it out anyway, refuting your point. Because there are clear issues allover it. Regardless if it has a favorable result that I like.

[AncestralDemocrat.]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

I wouldn't bother.. he did exactly the same nonsense in 2020 and got clowned upon as a result.

Never learned.

Obviously, he didn't. I personally thought Biden and Democrats would do well in 2020, and certainly have learned from that. But this is why I want for Election Day to come, and to come soon. Hopefully one narrative or the other will be proven.

He was even delving into the crosstabs of double digit Biden polls in Florida.. and suggesting they were underestimating his margins.

https://talkelections.org/FORUM/index.php?topic=384252.msg7473987#msg7473987

[soundchaser]

Can we not make EVERY thread into a meta analysis of wbrocks’ look into crosstabs?

[wbrocks67]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

I wouldn't bother.. he did exactly the same nonsense in 2020 and got clowned upon as a result.

Never learned.

Obviously, he didn't. I personally thought Biden and Democrats would do well in 2020, and certainly have learned from that. But this is why I want for Election Day to come, and to come soon. Hopefully one narrative or the other will be proven.

He was even delving into the crosstabs of double digit Biden polls in Florida.. and suggesting they were underestimating his margins.

https://talkelections.org/FORUM/index.php?topic=384252.msg7473987#msg7473987

Why are you obsessed with me? Put me on ignore then. Imagine being so obsessed with someone to pull up old posts.

[wbrocks67]

Can we not make EVERY thread into a meta analysis of wbrocks’ look into crosstabs?

I'd rather it not be, but people seem to be obsessed with me for some reason.

Not just that, but the fact that people on this forum want to act high and mighty and yet they were ready to salivate over an outlier poll, yet they themselves constantly want to dismiss any poll in the other direction. Priceless.

[President Punxsutawney Phil]

How did they end up with a 44% conservative sample?

[GeorgiaModerate]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

Looking at crosstabs is certainly a legitimate method of examining polls, and it can be especially helpful with a pollster with such a meager track record.

You probably won't want to hear this, but you've become just as reflexive in defending polls as those who pick them all apart.

That's because some posters on here will defend polls that are favorable to Democrats, or are less critical of them, but will tear apart any polls that are favorable for Republicans.

I won’t disagree with that. But you shouldn’t assume that there’s nothing wrong with a poll just because they criticize it. They may be right! The cross tabs here *do* look strange. The pollster is very up front saying they adjust their samples to try to compensate for non-response bias, which is a real thing. But it looks like they’ve probably overcompensated for it on this one.

[wbrocks67]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

[Skill and Chance]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

They do an explicit shy R adjustment.

[wbrocks67]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

They do an explicit shy R adjustment.

How does that work with their 2020 recall though?

There's no way you can put together an electorate that is 10%+ more conservative and yet somehow get an accurate 2020 vote recall like they do for each race.

[Alben Barkley]

Well this has been a wild week, but I'm almost ready to do a 180 and pull this one out on Democrats' Senate hopes.

[Alben Barkley]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

They do an explicit shy R adjustment.

How does that work with their 2020 recall though?

There's no way you can put together an electorate that is 10%+ more conservative and yet somehow get an accurate 2020 vote recall like they do for each race.

I'm not going to lie man: Your posts feel like coping. I get it. But still. The best thing that could be said about this race in recent days is that Oz never led. Now he has. As I've learned the hard way, one poll shouldn't be taken as gospel. So I'm not saying Fetterman is doomed, especially not when it's possible Shapiro (in a much stronger position) will carry him over the finish line. But he's been on nothing but a downward trajectory for weeks now. The timing couldn't be much worse.

[President Punxsutawney Phil]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

Not surprising if an Oz lead came from a polling firm, it came from a firm that does this.

[Alben Barkley]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

Not surprising if an Oz lead came from a polling firm, it came from a firm that does this.

Their worst poll so far was overestimating Biden by over 5 points in Ohio.

[President Punxsutawney Phil]

How did they end up with a 44% conservative sample?

You should check out the GA one, it's even worse, at 55% (vs. 38% in 2021 and 40% in 2020)

They do an explicit shy R adjustment.

How does that work with their 2020 recall though?

There's no way you can put together an electorate that is 10%+ more conservative and yet somehow get an accurate 2020 vote recall like they do for each race.

I'm not going to lie man: Your posts feel like coping. I get it. But still. The best thing that could be said about this race in recent days is that Oz never led. Now he has. As I've learned the hard way, one poll shouldn't be taken as gospel. So I'm not saying Fetterman is doomed, especially not when it's possible Shapiro (in a much stronger position) will carry him over the finish line. But he's been on nothing but a downward trajectory for weeks now. The timing couldn't be much worse.

There's a reason I've never had the race at Safe D.
Fortunately, Fetterman and Shapiro should still both win; Oz, this poll notwithstanding, has had a lot of trouble getting over 45%.

[Chancellor Tanterterg]

Junk poll

[Vespucci]

Reminds me of the Youngkin +8 poll in VA-GOV last year. Margin is probably not going to be that high, but this might make some people realize that Oz can win.

[LostFellow]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

If you want to have legitimacy, at least know what a "crosstab" is. The posters before your post mentioned candidate favorability and the governor race topline, none of which are cross tabulations since they are statistics from the whole sample.

The following can all be true:

• The previous posters are making unfounded assumptions (that the governor race can't be close nor can Oz and Fetterman have the same favorabilities) due to their D bias.
• Sometimes D-hack posters do pick apart crosstabs like the Black vote being 20% R or something, without thinking properly about sample size and variance.
• The critiques of the poll that the posters made regarding the governor toplines and favorabilities are more legitimate that critiques based-off cross tabs because those toplines are based off the complete sample of respondents.

[Calthrina950]

Immediately dismissed because of crosstabs, as would be expected here. I'm not surprised, although this pollster also polled Arizona and North Carolina, and found Kelly and Budd leading by perfectly reasonable margins.

If you want to have legitimacy, at least know what a "crosstab" is. The posters before your post mentioned candidate favorability and the governor race topline, none of which are cross tabulations since they are statistics from the whole sample.

The following can all be true:

• The previous posters are making unfounded assumptions (that the governor race can't be close nor can Oz and Fetterman have the same favorabilities) due to their D bias.
• Sometimes D-hack posters do pick apart crosstabs like the Black vote being 20% R or something, without thinking properly about sample size and variance.
• The critiques of the poll that the posters made regarding the governor toplines and favorabilities are more legitimate that critiques based-off cross tabs because those toplines are based off the complete sample of respondents.

Your condescension is not needed and is not welcome.

[Amenhotep Bakari-Sellers]

People Dooming and we haven't even voted yet c'mon people

[Pollster]

• This is a valid poll.
• This poll is an outlier.

Both can be (and likely are) true.

[President Punxsutawney Phil]

• This is a valid poll.
• This poll is an outlier.

Both can be (and likely are) true.

Yeah. This fits in the average, but it's just one part of an average, one that pretty clearly leans towards Fetterman at this stage and where Oz is still facing an uphill climb.

[Hollywood]

Democrats have a warped viewpoint of their overall strength due to bad polls their side has been consistently churning out for the last two years.  The reliance on Registered Voter Methods utilized for surveying and weighting respondents is clearly flawed given the Republican over-performance in PA, OH, FL, WI, MI, etc. exit polls.  In 2020, Republicans outvoted Democrats in PA by 41-40 percent, and Joe Biden narrowly defeated Trump by 1.2 points (80,000 votes).  This is much closer than Democrat Pollsters calculated.  

In the Wick Poll, Biden's margin is 1.5 percent over Trump (48.5-47%) instead of 1.2%.  The problem for Democrats is clearly evident from the number of Trump Voters remaining loyal to the party while the Biden Coalition has significantly fractured.  Some have complained that the Conservative Ideology is too high among the voters, but that is divided into two groups (21% Very Conservative and 23% Somewhat Conservative).  That rings true to me.  PA Populations from Pittsburgh to the Poconos is generally conservative, and I imagine some of the black christians in Philly would count themselves among that population.  

Democrats should have learned their lesson after VA, and the close-call in NJ, after the 2021 Election.  Republicans will get over 90% of the Trump 2020 Voters while Democrats will lose 15-25% of those Biden supporters that were mostly just panicked about Trump.