• This is a political forum that is non-biased/non-partisan and treats every person's position on topics equally. This debate forum is not aligned to any political party. In today's politics, many ideas are split between and even within all the political parties. Often we find ourselves agreeing on one platform but some topics break our mold. We are here to discuss them in a civil political debate. If this is your first visit to our political forums, be sure to check out the RULES. Registering for debate politics is necessary before posting. Register today to participate - it's free!

Wikileaks Email Shows Democrat Playbook to Rig Elections Through Oversampling

Internal poll that are weighted due to oversampling dont benefit the Democrats until theyre released to the general public.

First of all, no one trusts internal polls put out by campaigns, and they don't make it into fivethiryeight or RCP projections.

And you're pretending to be or ARE ignorant of why internal polls will often be "oversampled." If Hillary is planning an ad buy in VA, they need to know exactly where to spend that money, how many are independents, what issues are important, etc. They can't get that from a broad statewide poll, so they 'oversample' key regions, demographics, etc.

At least there zero evidence to confirm that Hillary and the DNC haven't been working with major news outlets.....:roll:

https://www.google.com/amp/observer.com/2016/08/wikileaks-reveals-mainstream-medias-coziness-with-clinton/amp/?client=ms-android-hms-tmobile-us

Right, and Republicans campaigns don't do all of that? You can't be that naive.
 
As is usual, you are wrong.

https://ig.ft.com/sites/brexit-polling/

46% leave, 48% stay, final polling.

So a 2% margin, with 6% still undecided in a binary yes or no question, is NOT a virtual tossup? LMAO....

It's not QUITE a coin flip, since stay has a slight lead, probably within the polling margin of error, but damn close. It is certainly not evidence that we can ignore polling altogether because BREXIT!!
 
Yay! Even more people (referring to the OP and the rest of those whining about "rigged polls") who don't understand how data analysis works, and don't care to know.

Actually we do. which is why when we see skewed polling samples heavily favoring democrats then people should be calling foul.
a better sample size would be an even number democrats and republicans and a slightly higher number of independents that can go either way.

I have a 1000 person poll and I want to sample and I have 20% or 30% more people in group A than group B and 10% more than the control group.
there is an issue with my sample.

I need to account for that in my bell curve or I need to redo the sample size.
 
Actually we do. which is why when we see skewed polling samples heavily favoring democrats then people should be calling foul.
a better sample size would be an even number democrats and republicans and a slightly higher number of independents that can go either way.

I have a 1000 person poll and I want to sample and I have 20% or 30% more people in group A than group B and 10% more than the control group.
there is an issue with my sample.

I need to account for that in my bell curve or I need to redo the sample size.

It's not an issue with your sample if there are more of group A than group B. More people do identify as Democrats than Republicans. Conservatives are relatively more likely to identity as Independents than liberals are.
 
Actually we do. which is why when we see skewed polling samples heavily favoring democrats then people should be calling foul.
a better sample size would be an even number democrats and republicans and a slightly higher number of independents that can go either way.

I have a 1000 person poll and I want to sample and I have 20% or 30% more people in group A than group B and 10% more than the control group.
there is an issue with my sample.

I need to account for that in my bell curve or I need to redo the sample size.

Are you saying that the pollsters should not pick their polling subjects at random?
 
First of all, no one trusts internal polls put out by campaigns, and they don't make it into fivethiryeight or RCP projections.

And you're pretending to be or ARE ignorant of why internal polls will often be "oversampled." If Hillary is planning an ad buy in VA, they need to know exactly where to spend that money, how many are independents, what issues are important, etc. They can't get that from a broad statewide poll, so they 'oversample' key regions, demographics, etc.



Right, and Republicans campaigns don't do all of that? You can't be that naive.

Your'e the one pretending to be or IS ignorant.

Tom Mazzie is recomending oversampling " for our polling so we can maximize what they get out of our media polling.".

He then goes on to " highly recomend " over sampling of hispanics, african americans and native americans.

This is not some objective pollster who's trying to determine how much to spend where. Thats ridiculous.

Oversampling is the last thing you would want to do when it comes down to spending on Political adds. It would mean they're over spending.

No, oversampling in this case is to, AGAIN like it says in the friken email you didnt read, is to maximize what they get out of their media polling.

Dont tell me you just barged into another thread without reading the Op.
 
It's not an issue with your sample if there are more of group A than group B. More people do identify as Democrats than Republicans. Conservatives are relatively more likely to identity as Independents than liberals are.

if you want to do that then you have to expand your margin of error.
if I do a sample size showing 50% of them democrats 20% of them republicans and 30% independents and go Clinton is up 14% in the polls.
is that a fair poll? not really.
 
if you want to do that then you have to expand your margin of error.
if I do a sample size showing 50% of them democrats 20% of them republicans and 30% independents and go Clinton is up 14% in the polls.
is that a fair poll? not really.

I don't think "fairness" has anything to do with polling
 
Are you saying that the pollsters should not pick their polling subjects at random?

why cant' you actually address what I say instead of something I never said.

if your sample size is skewed then you have to account for that.
 
if you want to do that then you have to expand your margin of error.
if I do a sample size showing 50% of them democrats 20% of them republicans and 30% independents and go Clinton is up 14% in the polls.
is that a fair poll? not really.

It's a fair poll if 50% of people identify as Democrats, 20% of people identify as Republicans, and 30% of people identify as independents. Now that isn't true, but it also isn't true that an equal number of people identity as Republicans and Democrats. Any pollster weighting that way would have an inherent error.

(Good pollsters don't weight by Party ID at all because it is a fluid statistic.)
 
I don't think "fairness" has anything to do with polling

actually they are. polls are supposed to be objective in their nature.
meaning they are supposed to give even weight across everything. If there is a skew in who is asked
then that has to be accounted for.

if it isn't accounted for then you end up with a bad poll results.

Wikileaks has proven that dishonest polling numbers are being used and the media is supporting it.
 
actually they are. polls are supposed to be objective in their nature.
meaning they are supposed to give even weight across everything. If there is a skew in who is asked
then that has to be accounted for.

if it isn't accounted for then you end up with a bad poll results.

Wikileaks has proven that dishonest polling numbers are being used and the media is supporting it.

And anagram keeps saying that good pollsters don't weight polls by party I.d.
 
It's a fair poll if 50% of people identify as Democrats, 20% of people identify as Republicans, and 30% of people identify as independents. Now that isn't true, but it also isn't true that an equal number of people identity as Republicans and Democrats. Any pollster weighting that way would have an inherent error.

(Good pollsters don't weight by Party ID at all because it is a fluid statistic.)

the only way they know if someone is republican or democrat or independent is to ask them.
again there is evidence of oversampling to skew results.

it doesn't surprise me at all.
 
the only way they know if someone is republican or democrat or independent is to ask them.
again there is evidence of oversampling to skew results.

it doesn't surprise me at all.

Since when has Oversampling been controversial?
 
2 percentage points seperating the two is hardly decisive.

Just to be clear.

The results of brexit was close, with plenty of polls showing that it was a toss-up between leave or stay. The General election polls have consistently given Clinton a slight lead

Perhaps a 2% lead? My point is, polling is not an exact science. It has been wrong in the past and will be wrong in the future. Those that lean on it as concrete data are foolish or seek to influence voting. Polling data has been off rather notoriously of late, as evidenced by the article of the poster I quoted.
 
Since when has Oversampling been controversial?

Since the latest podesta emails showed it was being done to maximize what the Democrats could get out of their media polling.
 
the only way they know if someone is republican or democrat or independent is to ask them.
again there is evidence of oversampling to skew results.

it doesn't surprise me at all.

There is not evidence of that. That's not what oversampling means. It has to do with sampling more of certain smaller demographics to get an accurate breakdown of them but re-weighting them in their correct proportion.

And yes you ask people, which is how we've found out more identify as Democrats.
 
There is not evidence of that. That's not what oversampling means. It has to do with sampling more of certain smaller demographics to get an accurate breakdown of them but re-weighting them in their correct proportion.

And yes you ask people, which is how we've found out more identify as Democrats.

But thats not the reason given IN THE FRIKEN EMAIL.

Tom Mazzie, formally of MoveOn.org is " highly recomending " oversampling of minorities to " maximize what they get out of their media polling ".

Its not about internal polling or re-weighting oversampled demographics.
 
But thats not the reason given IN THE FRIKEN EMAIL.

Tom Mazzie, formally of MoveOn.org is " highly recomending " oversampling of minorities to " maximize what they get out of their media polling ".

Its not about internal polling or re-weighting oversampled demographics.

They're doing it so they can get accurate subsections for advertising. Why he wants it broken down by market/region.
 
But thats not the reason given IN THE FRIKEN EMAIL.

Tom Mazzie, formally of MoveOn.org is " highly recomending " oversampling of minorities to " maximize what they get out of their media polling ".

Its not about internal polling or re-weighting oversampled demographics.

And if you need more substantial evidence you can just look at the fact that the email was from 2008 and the polls in 2008 were generally accurate.
 
And anagram keeps saying that good pollsters don't weight polls by party I.d.

the only way they know if they are democrat or republican is to ask.
otherwise there is no way of knowing.

unless they account for the skew in their weighting then their sample size is wrong.

If call 1000 people and 805 of them turn out to be democrat and publish the poll then the poll
is heavily skewed unless I account for it.
 
if you want to do that then you have to expand your margin of error.
if I do a sample size showing 50% of them democrats 20% of them republicans and 30% independents and go Clinton is up 14% in the polls.
is that a fair poll? not really.

What should the numbers be?

Democrats______ %
Republicans ______%
Independents ______%

Fill those in to reflect what would make a "fair" poll.
 
Back
Top Bottom