Here’s a beautiful graphical interface for analyzing the election at the NYT:
http://www.nytimes.com/interactive/2012/11/02/us/politics/paths-to-the-white-house.html
Wow, the MSM is sure going to town with these polls! And I can’t say I blame them, though it’s actually the fault of the polls themselves. As you may have seen, our man has seemingly been doing worse in the polls ever since Sandy hit.
All the lefty-pundits are now pointing to these polls and saying, “Look! Look! You can’t explain them away without systematic bias!” (when in fact, I think they mean “systemic bias”).
Here’s two classic examples:
http://krugman.blogs.nytimes.com/2012/11/03/reporting-that-makes-you-stupid/
http://fivethirtyeight.blogs.nytimes.com/2012/11/03/nov-2-for-romney-to-win-state-polls-must-be-statistically-biased/
But in fact, that is exactly what there is: systemic bias!
Almost every single pollster is under-counting the expected Republican turnout (Rasmussen and Gallup excepted). It turns out that most pollsters are using roughly the same turnout percentages as 2008. This is absurd, period. Some have even amplified the 2008 R vs D difference further towards the Democrats(!?)
There’s no way on G-d’s green earth that the D+6 or D+8 percentage difference (or delta) of 2008 will hold this year.
We must hope that Republicans maintain their enthusiasm of 1 −2 weeks ago on faith, against the tide of the most recent news about the polls. These pollsters are, in effect, collaborating. This an “unspoken conspiracy.” Right now the Pollsters, acting through the MSM, have the opportunity to do what no amount of advertising can do, and that is to sway what happens on Nov 6. They know exactly what they’re doing. But they view it as a way to “right the wrong of Florida” — they simply think to themselves that they are “stealing an election back for the home team.”
Look at the “about me” page for the company, PPP (which regularly grants Obama a huge projected +D turnout):
In Numbers, We Can Make a Difference
Here's what others have said about Public Policy Polling's track record:
Note what they are highlighting: “Make a Difference” — make a difference??? How about just “be accurate” — no, they know well that politicians will hire them to “do a job” — and that job is to slant the polls by upping the expected +D margins. Then when they release a hugely slanted result, the press, always looking for a story, will publish that result. This is what they mean when they say, “Now… PPP is driving national coverage.”
Outfits like this can affect voter turnout. If enough Republicans think “it’s hopeless” to vote in order to affect the “inevitable Obama victory” — precisely why so many stayed home in 2008 — it becomes self-fulfilling.
But, if Republicans and so-called independents come out in expected numbers, we will win. Independents have been shown now to break for Romney. They break for the challenger historically (despite what that schmuck Nate Silver says) but this time even more so. So it’s all in the turnout.
— FF
P.S. Does anyone know why the Wall Street Journal has joined with the clearly partisan Marist/NBC polling organization? I hope they break that relationship up after this election.
----- A Texan responded:
Ya, I like our chances this year. We just have to show up.
Its funny how in 2008 the pundits made white voters feel that if they vote for O that it would move us beyond race, but now in 2012 if they don't vote for him it must mean that they went back to being racist AGAIN!
And as a resident of Texas, I still have trouble accepting the fact that Virginians consider themselves "southern."
-----
On Oct 30, 2012, at 8:57 PM, Frumious Falafel wrote:
Every one on the Right says the following in one way or another — that many of the polls out there — especially those polls which RealClearPolitics averages into its mix — are tilted towards an overly high Democrat turnout. Whether it’s +8 or +6 — most argue that that number won’t hold given the enthusiasm of the Republicans this time around.
And while Sandy provides the president with opportunities to look “presidential” (Charles Krauthammer wonders why he didn’t do this earlier with Benghazi?), poor weather, or frankly any difficulty in the logistics of getting to the polling station is a benefit to Republicans for what should be are obvious reasons.
Then you add that Republicans are highly motivated this election, and Sandy is really a little push from the L-rd himself for us. G-d helps those who help themselves — so given that push, we now need to run the last mile on our own.
In any case, part and parcel with all of the above is an apparent flaw in the model of “Team-0” in their prediction of the percentage of white/independent-voter turnout. Gallup and others have indicated with their polling breakdown data that the huge advantage of Romney in the popular vote (the beauty contest — which does matter if it’s highly imbalanced) is largely due to white/independent voters (especially in the south)
That advantage is now responsible for a 5 − 6 point lead in the Gallup National poll.
But the Democrats come back always with, “but that’s not going to make a difference in the battleground states” the implication being, and it’s correct on its face, that the president is elected by electoral votes. And presumably, if one follows RCP and their “electoral” map, Obama still wins (due largely to the battleground states going towards him).
So, yes, while electoral and battleground states are certainly where we must all look, the question is, are many of the pollsters exhibiting a bias with their +x Democrat turnout model and/or their assumptions on what white/independent turnout will be this year? If so, these slanted results are being constantly averaged into the RCP result (which is influential).
The answer is yes on both. There are a number of articles claiming that the +D number is too high — that many polls are assuming this number is far higher than it will be (and the “best” pollsters for not doing that are Rasmussen and Gallup). Then there’s this interesting article analyzing the definitely motivated white/independent southern vote — a further clue that the polls will be pretty far off this year.
http://battlegroundwatch.com/2012/10/29/the-folly-of-david-axelrods-turnout-model/
Indeed, if both the +D offset and the expected white/independent voter turnout are being under-estimated, then Romney will win, and win large this election.
-- FF
http://www.nytimes.com/interactive/2012/11/02/us/politics/paths-to-the-white-house.html
Wow, the MSM is sure going to town with these polls! And I can’t say I blame them, though it’s actually the fault of the polls themselves. As you may have seen, our man has seemingly been doing worse in the polls ever since Sandy hit.
All the lefty-pundits are now pointing to these polls and saying, “Look! Look! You can’t explain them away without systematic bias!” (when in fact, I think they mean “systemic bias”).
Here’s two classic examples:
http://krugman.blogs.nytimes.com/2012/11/03/reporting-that-makes-you-stupid/
http://fivethirtyeight.blogs.nytimes.com/2012/11/03/nov-2-for-romney-to-win-state-polls-must-be-statistically-biased/
But in fact, that is exactly what there is: systemic bias!
Almost every single pollster is under-counting the expected Republican turnout (Rasmussen and Gallup excepted). It turns out that most pollsters are using roughly the same turnout percentages as 2008. This is absurd, period. Some have even amplified the 2008 R vs D difference further towards the Democrats(!?)
There’s no way on G-d’s green earth that the D+6 or D+8 percentage difference (or delta) of 2008 will hold this year.
We must hope that Republicans maintain their enthusiasm of 1 −2 weeks ago on faith, against the tide of the most recent news about the polls. These pollsters are, in effect, collaborating. This an “unspoken conspiracy.” Right now the Pollsters, acting through the MSM, have the opportunity to do what no amount of advertising can do, and that is to sway what happens on Nov 6. They know exactly what they’re doing. But they view it as a way to “right the wrong of Florida” — they simply think to themselves that they are “stealing an election back for the home team.”
Look at the “about me” page for the company, PPP (which regularly grants Obama a huge projected +D turnout):
In Numbers, We Can Make a Difference
Here's what others have said about Public Policy Polling's track record:
In Numbers, We Can Make a Difference
Here's what others have said about Public Policy Polling's track record:
Talking Points Memo 8/22/11 How PPP became the 'It' Democratic Pollster
"Just
a few short years ago, Public Policy Polling was an obscure Democratic
outfit, mostly focused on local polling in Raleigh, North Carolina.
Now, ten years after its founding, PPP
is driving national coverage with an unmatched supply of polls on
everything from the Republican primaries to God's approval rating.
Note what they are highlighting: “Make a Difference” — make a difference??? How about just “be accurate” — no, they know well that politicians will hire them to “do a job” — and that job is to slant the polls by upping the expected +D margins. Then when they release a hugely slanted result, the press, always looking for a story, will publish that result. This is what they mean when they say, “Now… PPP is driving national coverage.”
Outfits like this can affect voter turnout. If enough Republicans think “it’s hopeless” to vote in order to affect the “inevitable Obama victory” — precisely why so many stayed home in 2008 — it becomes self-fulfilling.
But, if Republicans and so-called independents come out in expected numbers, we will win. Independents have been shown now to break for Romney. They break for the challenger historically (despite what that schmuck Nate Silver says) but this time even more so. So it’s all in the turnout.
— FF
P.S. Does anyone know why the Wall Street Journal has joined with the clearly partisan Marist/NBC polling organization? I hope they break that relationship up after this election.
----- A Texan responded:
Ya, I like our chances this year. We just have to show up.
Its funny how in 2008 the pundits made white voters feel that if they vote for O that it would move us beyond race, but now in 2012 if they don't vote for him it must mean that they went back to being racist AGAIN!
And as a resident of Texas, I still have trouble accepting the fact that Virginians consider themselves "southern."
-----
On Oct 30, 2012, at 8:57 PM, Frumious Falafel wrote:
Every one on the Right says the following in one way or another — that many of the polls out there — especially those polls which RealClearPolitics averages into its mix — are tilted towards an overly high Democrat turnout. Whether it’s +8 or +6 — most argue that that number won’t hold given the enthusiasm of the Republicans this time around.
And while Sandy provides the president with opportunities to look “presidential” (Charles Krauthammer wonders why he didn’t do this earlier with Benghazi?), poor weather, or frankly any difficulty in the logistics of getting to the polling station is a benefit to Republicans for what should be are obvious reasons.
Then you add that Republicans are highly motivated this election, and Sandy is really a little push from the L-rd himself for us. G-d helps those who help themselves — so given that push, we now need to run the last mile on our own.
In any case, part and parcel with all of the above is an apparent flaw in the model of “Team-0” in their prediction of the percentage of white/independent-voter turnout. Gallup and others have indicated with their polling breakdown data that the huge advantage of Romney in the popular vote (the beauty contest — which does matter if it’s highly imbalanced) is largely due to white/independent voters (especially in the south)
That advantage is now responsible for a 5 − 6 point lead in the Gallup National poll.
But the Democrats come back always with, “but that’s not going to make a difference in the battleground states” the implication being, and it’s correct on its face, that the president is elected by electoral votes. And presumably, if one follows RCP and their “electoral” map, Obama still wins (due largely to the battleground states going towards him).
So, yes, while electoral and battleground states are certainly where we must all look, the question is, are many of the pollsters exhibiting a bias with their +x Democrat turnout model and/or their assumptions on what white/independent turnout will be this year? If so, these slanted results are being constantly averaged into the RCP result (which is influential).
The answer is yes on both. There are a number of articles claiming that the +D number is too high — that many polls are assuming this number is far higher than it will be (and the “best” pollsters for not doing that are Rasmussen and Gallup). Then there’s this interesting article analyzing the definitely motivated white/independent southern vote — a further clue that the polls will be pretty far off this year.
http://battlegroundwatch.com/2012/10/29/the-folly-of-david-axelrods-turnout-model/
Indeed, if both the +D offset and the expected white/independent voter turnout are being under-estimated, then Romney will win, and win large this election.
-- FF
