Another fake Democrat poll exposed

All good polls over-sample Democrats because more people identify as Democrats than Republicans. This is illustrated by the 3% Hillary Clinton lead in the 2016 polls which proved accurate.

A good poll is representative of the national population (of likely voters).

This. If there are more Democrats than Republicans, more Democrats will appear in the sample. That's how polling works. Apparently, this is another skill that the Trump Cult is lacking.
 
All good polls over-sample Democrats because more people identify as Democrats than Republicans. This is illustrated by the 3% Hillary Clinton lead in the 2016 polls which proved accurate.

A good poll is representative of the national population (of likely voters).

Hillary Clinton never had the lead in 2016. Donald Trump was elected President. Perhaps you didn't notice.
 
This message is hidden because Into the Night is on your ignore list.

Don't worry, I'll ignore the other you as well.
 
All good polls over-sample Democrats because more people identify as Democrats than Republicans. This is illustrated by the 3% Hillary Clinton lead in the 2016 polls which proved accurate.

A good poll is representative of the national population (of likely voters).

That didn't "prove accurate." Democrats have to cheat with fake absentee ballots, "early voting" scams, ballot harvesting, and automatic voter registration for illegal immigrants to even STAY IN THE GAME. :bs:
 
The 'methodology' is a lie much of the time as well. It has to be a verifiable methodology, an it has to show how bias was removed during data collection, the raw data itself must be available, the entire set of questions used, whether its 'data' or 'made up shit' and how that can be verified, the variance must be declared and justified, the margin of error calculation must be shown, and selection by randN must be done up on the data. Anyone quoting a poll as the future is bogus. Statistical mathematics does not have the power of prediction any more than probability mathematics, and for the same reason.

Lies, damn lies, and statistics. :nodyes:
 


Lies, damn lies, and statistics. :nodyes:

People lie with statistics so easily because most people don't know any statistical math, hence the common phrase. Governments and government agencies are no different.

Statistical math has its uses, but it is not capable of predicting the future, has VERY stringent requirements for the raw data and variances used, and requires the calculation of the margin of error (and published) along with the mean and the raw data set used itself and the variance used and its justification. If any one of these are missing, the summary is meaningless.

One good use, for example, is in conducting a census count. Another is conducting an analysis of a pattern of infection to help determine its vector. None of it can tell the future. Statistical math uses random numbers for picking and normalizing the data. Like probability math, it loses the inherent power of prediction in mathematics because of the importation of random numbers, which are based on the modulus binomial, a one-way function.

Statistical math can be a useful tool for summarizing past or present data, but it cannot predict the future. It's not a crystal ball or a Holy Entrail.
 
obama was also around the 40s in his first term around this time, in fact recently trump was even higher than obama was in this particular point in their terms. approval ratings aren't a 1:1 translation to who people will vote for, because it's not trump vs. no one. it's trump vs. someone else. case in point, trump won in 2016 with approval ratings in the 40's. free lesson from me, the poll expert.
Trump is losing voters he badly needs, 65+ and women.
 
Hillary Clinton never had the lead in 2016. Donald Trump was elected President. Perhaps you didn't notice.

Leftist Moron rebuttal: BUT, BUT, she won the POPULAR vote!!

giphy.gif
 
Back
Top