,
18 tweets,
7 min read
Read on Twitter
Does Polling Accuracy Count Anymore?
(A data string for our readers.)
Many of you are now familiar with our open question to the @DNC as to how they made their choice of "approved pollsters."
We started w/ the ARG 2016 bench-marking analysis -
americanresearchgroup.com/ratings/2016/u…
1/
(A data string for our readers.)
Many of you are now familiar with our open question to the @DNC as to how they made their choice of "approved pollsters."
We started w/ the ARG 2016 bench-marking analysis -
americanresearchgroup.com/ratings/2016/u…
1/
We then layered in sequential number rankings & cross referenced that with the official @DNC list of approved pollsters & sponsors.
The resulting chart is here -
2/_
The resulting chart is here -
2/_
As you can see, only 3 of the 12 ranked in the 2016 Top-10 for accuracy.
9 of the 12 fell below the 2016 Top-10.
3 of the approved firms gave no final 2016 estimate at all.
3/_
9 of the 12 fell below the 2016 Top-10.
3 of the approved firms gave no final 2016 estimate at all.
3/_
Perhaps to compensate for this 'inconsistency' a new fable is circulating in the MSM that pollsters actually got the 2016 race correct because many landed their final Popular Vote estimates within a few points of the actual final winning margin. This narrative is garbage.
4/_
4/_
The vast majority of 2016 pollsters gave Donald Trump no path to victory. Most never showed Mr. Trump leading Mrs. Clinton. Many showed Mrs. Clinton with a sustained & insurmountable 2016 lead & decreased that lead only in their final estimates.
5/_
5/_
All of the charting data showing this phenom is available on the Real Clear Politics website at the link below.
MSM myth-makers would like you to forget where it is. Don't buy what they're selling.
realclearpolitics.com/epolls/2016/pr…
6/_
MSM myth-makers would like you to forget where it is. Don't buy what they're selling.
realclearpolitics.com/epolls/2016/pr…
6/_
Pollsters giving Donald Trump no path to victory in 2016 by never or only once showing him in the lead were not only wrong but grossly mislead their readers. The shock that then reverberated across the country & the media is a lasting testament to this polling negligence.
7/_
7/_
RCP keeps excellent 2016 election records but we poll more frequently than their standard format allows & we keep even more detailed records of our own published work.
The final months of our 2016 presidential work is below & at this link -
rasmussenreports.com/public_content…
9/_
The final months of our 2016 presidential work is below & at this link -
rasmussenreports.com/public_content…
9/_
As you can see, we showed the race lead changing frequently, alerting our readers that the 2016 race was actually close for an extended period that fall.
Only 2 of our polling competitors saw & promoted a similarly tight 2016 race. We detail this here -
10/_
Only 2 of our polling competitors saw & promoted a similarly tight 2016 race. We detail this here -
10/_
Next lets do some comparisons - again using RCP public data.
Here is IBD/TIPP. Pretty close to us until the last survey when they projected Trump would win the Popular Vote by 2 points. Their readers at least knew the race was always close & of course it was.
11/_
Here is IBD/TIPP. Pretty close to us until the last survey when they projected Trump would win the Popular Vote by 2 points. Their readers at least knew the race was always close & of course it was.
11/_
Here is USC/LA Times. Their readers knew Trump was competitive throughout & for this they were mocked mercilessly by the 'polling elite.' Their final estimate also showed Trump winning the Popular Vote but their readers were prepared for him to win, no surprises.
12/_
12/_
Here is ABC News/Washington Post. They mostly featured an unstoppable Mrs. Clinton until, just before their final projection, they signaled a small tremor in the force. It was soon eclipsed by their final projection.
13/_
13/_
Here is @realDonaldTrump latest polling target - Fox News. They showed little doubt that Mrs Clinton had a solid lock on the 2016 race. Their final MOE could get them into the actual final margin, but their viewers likely missed that part.
14/_
14/_
Here is NBC News/Wall St Journal. Not a lot of data points but very reassuring to Camp Clinton. Trump really had no chance.
15/_
15/_
Here is CNN/ORC. What can we call this 2016 election work - A Profile in Courage? At least they showed Mr. Trump up once early on in these final months.
16/_
16/_
Lets wrap up here with the 2019 media's "It-Girl" pollster, that gold-est of the 'Gold Standard' academic sponsor pollsters, aka The Big Q. 2016 was a magnificent testament to how they operate.
17/_
17/_
We'll end this analysis here but others are welcome to continue the pollster accuracy comparisons. The RCP link is above.
But don't forget, the MSM now insists that 2016 U.S. election polling was not at all misleading.
How stupid do they think we all are?
18/End
But don't forget, the MSM now insists that 2016 U.S. election polling was not at all misleading.
How stupid do they think we all are?
18/End
