National poll by Quinniepiac: Rubio jumps ahead

[Former Democrat]

http://www.tampabay.com/blogs/the-buzz-florida-politics/national-poll-rubio-jumps-to-front-of-gop-field-beats-clinton/2226596

[MT Treasurer]

Clinton: 45%
Rubio: 43%

Clinton: 46%
Bush: 39%

[Former Democrat]

they release full numbers tomorrow

[DS0816]

You must be logged in to read this quote.

This may be the poll's problem.

[Gallium]

You must be logged in to read this quote.

This may be the poll's problem.

An improvement on their last national presidential poll that sampled 610 Democrats and 707 Republicans...

Edit: That was in November. Their more recent March poll was 29% D and 32% R.

Their sample sizes are wack.

[Landslide Lyndon]

You must be logged in to read this quote.

This may be the poll's problem.

Wow! This maybe even worse that the CNN poll.

[Mr. Morden]

You must be logged in to read this quote.

This may be the poll's problem.

Not sure why that's a problem.  They're presumably oversampling both Democrats and Republicans so that they can get a decent sized sample for their primary polling.  It doesn't mean that Democrats, Republicans, and Independents are all going to be weighted the same in the general election result.  They poll different subsamples by different ammounts, but then do demographic weighting to get a sample of "registered voters".

[New_Conservative]

Wow, interesting! I really like the campaign Rubio is trying to run so far.

[DS0816]

You must be logged in to read this quote.

This may be the poll's problem.

Not sure why that's a problem.  They're presumably oversampling both Democrats and Republicans so that they can get a decent sized sample for their primary polling.  It doesn't mean that Democrats, Republicans, and Independents are all going to be weighted the same in the general election result.  They poll different subsamples by different ammounts, but then do demographic weighting to get a sample of "registered voters".

It is a sampling that, intentionally or not, assumes an almost even self-identification between Republicans and Democrats. And the margin of error is possibly as much as 8.2 percentage points. Between the presidential elections of 2008 and 2012, there were approximately an average of 130 million votes cast. An 8.2-point margin, as a possible level of error, is as much as 10.66 million votes. That is not good.

[Mr. Morden]

It is a sampling that, intentionally or not, assumes an almost even self-identification between Republicans and Democrats.

No it doesn't.  That's not how polling works.  They poll people and get different response rates from different groups, but then apply weighting in order to get the sample to match the demographics of "registered voters".  In a case like this, where they're also doing Democratic and Republican primary polling, they also oversample those groups in their calling, to get a decent sample for the primary numbers, but then reweight things for the general election poll.

You can't just say: "Gee, 567 Republicans from a total sample of 1353.  I guess that means they're assuming that 567/1353 = 42% of registered voters are Republican."  That's wrong.  When they calculate the numbers for the general election, they downweight those responses to something more reasonable.  You can see this in their previous polls.  Here's their recent New Jersey poll:

http://www.quinnipiac.edu/news-and-events/quinnipiac-university-poll/new-jersey/release-detail?ReleaseID=2219

You must be logged in to read this quote.

444 Republicans, 539 Democrats, and 1428 total, which I guess leaves 445 Indies.  "Well, derp de doo, I guess that means they assume 37.7% Dems, 31.1% GOP, and 31.2% Indies, right?"  Wrong.  If that's what they were doing then why is it that when you try to reconstruct, say, Christie's support in the Christie vs. Clinton matchup, you can't get it by adding up the contributions from the party samples like so?:

86% * 0.311 + 6% * 0.377 + 37% * 0.312 = 40.6%

40.6% is nowhere close to the 36% they actually list for his support in that matchup.  That's because they oversampled Republicans in order to get a decent sized sample for their GOP primary poll, but then downweight those votes in the general election matchup to something more reasonable.

[Tender Branson]

That's not possible ... Hillary is inevitable !

IceBear told me so.

[IceSpear]

That's not possible ... Hillary is inevitable !

IceBear told me so.

LOL

1) I never said Hillary was inevitable in the general
2) Weren't you just saying a couple days ago Hillary was actually ahead 5-10 points while you were desperately trying to discredit the CNN poll? Please at least try to stay consistent and end your cherry picking
3) You really shouldn't be judging the prognostication of others with gems like these:

So, Hillary already dropped 5% or so this week compared to previous polls and she'll likely drop another 5% right after she announces her run.

Creating a tied race already.

And in the end she might lose too (maybe even in a landslide).

I told you Hillary-fetishists early about this, but you were not listening ...

I'll ask you again how you feel about her chances in a few months when she has announced her candidacy and already has negative favorable ratings and trails Bush and Walker in the polls ...

I'm sure they'll be tied any day now! LMFAO! Just look at that Bushmentum!

[Gallium]

Clinton 45%
Christie 40%

Clinton 46%
Paul 42%

Clinton 47%
Huckabee 42%

Clinton: 45%
Rubio: 43%

Clinton: 46%
Bush: 39%

Clinton: 46%
Walker: 41%

Clinton: 48%
Cruz: 41%

http://www.quinnipiac.edu/news-and-events/quinnipiac-university-poll/national/release-detail?ReleaseID=2221

[King]

Why is this thread title Rubio jumps ahead when he's not even ahead?

[Tender Branson]

Why is this thread title Rubio jumps ahead when he's not even ahead?

Maybe because Rubio jumped ahead in the primary ...

[henster]

The honest & trustworthy result is brutal for Hillary, I don't think it got that bad for Obama at any point.

[IceSpear]

Sorry, dude, but ANYTHING can happen in politics, even Hillary losing the election

Again, when did I ever say Hillary couldn't lose the election?

As for Maryland, nearly everyone who followed politics got that one wrong.

[King]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

[Attorney General, LGC Speaker, and Former PPT Dwarven Dragon]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

Leading by 2 points nationally over a year before the election does not give you a 100% chance of victory. And don't point to that CNN or ABC  Poll. Polls of all adults are junk, and remember that many 2014 losers: Pryor, Landrieu, Crist, Coakley, Hagan, etc.: all led by double digits in parts of 2013/early 2014. Early polls are close to useless.

[IceSpear]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

Sorry King, but you're way too optimistic even for me. I've never thought she was guaranteed to win the general, and if anything my confidence that she will has gone down lately due to the all out, universal, and perpetual media assault on her over the past few months. I'm starting to think they'll do everything in their power to keep her out of the White House, and I'm not sure if she can overcome that. Her biggest obstacle to winning the WH is not Scott Walker or Jeb Bush, but the overtly hostile media.

[Landslide Lyndon]

Why is this thread title Rubio jumps ahead when he's not even ahead?

Maybe because Rubio jumped ahead in the primary ...

The title says that he beats Clinton, which is obviously not the case.

[IceSpear]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

Leading by 2 points nationally over a year before the election does not give you a 100% chance of victory.

That would be cherrypicking. The aggregate still has her leading all the Republicans by 8-13 points.

[Pandaguineapig]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

Leading by 2 points nationally over a year before the election does not give you a 100% chance of victory.

That would be cherrypicking. The aggregate still has her leading all the Republicans by 8-13 points.

Most recent polls have the race in single digits(with the potential exception of Ted Cruz) The CNN poll is a joke

[IceSpear]

Disappointed in IceSpear.

I'll keep on saying it. Hillary can't lose this general election.

Leading by 2 points nationally over a year before the election does not give you a 100% chance of victory.

That would be cherrypicking. The aggregate still has her leading all the Republicans by 8-13 points.

Most recent polls have the race in single digits(with the potential exception of Ted Cruz) The CNN poll is a joke

Yes, the aggregate has the "serious candidates" with an 8-9 point deficit, and the others with a 10-13 point deficit.

[King]

I'm not optimistic. I like Rand Paul more than Hillary just for intrigue factor. But this is Hillary's election. No other way to slice it.