Not exactly a sexy topic, but an important one…
If polls are to be believed, then Marine Corps recruits at both Parris Island and San Diego would be taught that Hillary Clinton is the top dog in their chain of command.
Suffice it to say, I’m not the only conservative who questions the veracity of polls, especially since the 2016 presidential election.
Unfortunately, no one seems to have informed the media. In particular, the Fox News Channel. Suffice it to say that the liberal Establishment Media still conducts themselves as the Ministry of Propaganda for the Democrats.
Nonetheless, case in point would be that nearly every one of the FNC news anchors and field reporters breathlessly telling the world just how badly President Trump is getting his ass handed to him by Joe Biden. With that said, I’m going to use the latest Trump vs Biden FNC poll (July 19, 2020) for illustration purposes.
Be it known that I’m the kind of guy who more often than not doesn’t take anything at face value. Just because someone “heard it on TV” or “saw it on line” doesn’t make it the truth.
Watching five minutes of CNN or MSNBC, or reading five minutes worth of the New York Times or the Washington Post will tell you that.
Anyhow, just a few fine points I’d like my fellow conservatives to take as peek at, especially when it comes to steering folks into **believing** how we collectively will vote.
Side note: Nearly as bad as liberals, many conservatives have the very nasty habit of believing any given headline (regardless of how outrageous it is) is Gospel truth. If you see a headline claiming Nancy Pelosi has been impeached or Ilhan Omar has been deported, it’s total bullshit. Go to the article itself, you’ll more often than not will see that no half-way reputable news service is cited. The whole thing is just made up fake news clickbait.
Anyhow, back to the topic at hand.
1). Does it really make a difference who conducts any given poll?
One would think this is a no-brainer. Saying this in all objectivity, both Left and Right have their own biases. That’s simply human nature. I would no more trust a CNN poll as I would a Newsmax poll.
Even polls that are supposedly “right down the middle” (such as Fox News or Pew Research), simply aren’t. A bit deeper into this article, I’ll point out how.
2). Beware the MoE!
Rarer than a transvestite Yeti on roller skates, just try finding a poll that doesn’t have an MoE (margin of error).
All the MoE does is to allow the pollsters the opportunity to shift the verdict as they see fit, period, end of sentence, full stop.
The cited FNC poll states their inquiry has “a margin of sampling error of ± 3 percentage points.” Keep in mind that I’ve seen MSNBC and CNN polls with MoEs as high as 7-11 points.
Example, if 10 or 100 or 1,000, or billions upon billions of people were asked “Will you vote for Candidate A or Candidate Z?”, the least amount of legitimate responses are the following;
- Candidate A
- Candidate Z
- Not sure yet/someone else
- Prefer not to answer
Just for illustration purposes, let say that 100 people were polled regarding Candidates A and Z with an +/- MoE of 6.
- 55% said A
- 44% said Z
- 1% said Not sure yet/somone else
- 0% prefer not to answer
It sure seems simple enough to say that Candidate A is crushing Z by double digits (11 points), right? Well, not really.
This poll can arbitrarily change the numbers simply by implementing their MoE of 6.
- Candidate A has now magically dropped by 6 to a new percentage of 49, now in second place.
- Candidate Z has now magically taken the lead by adding the MoE of 6, ensuring that Z is cited as receiving 50 percent of those polled prefer him.
Seriously, how hard is it for a pollster to add up the totals where only four responses were possible?
3). It really is important how many people were polled?
This may be presumptuous of me, but in a nation of roughly 330 million people, you’d think that FNC would pose such an important question to more than a miniscule 1,104. Right?
You would think.
4). The voter status matters.
- “Likely Voters” are the beast gauge of people who really will cast their ballots. Nothing short of Nuclear Winter will keep these folks away from the ballot box
- “Registered Voters” (what Fox News did) is code for slack-assery. Just because your 30-something Starbuck’s barista with a Master’s degree in Lesbian Dance Theory was registered to vote when he got his driver’s license doesn’t mean a damn thing.
- “Adults” is the worst category of people to poll. For all the pollster knows, they could have on the phone a stoned 18-year-old illegal alien.
5). “Political Identification” and what it means.
For whatever bizarre reason, in the FNC poll, they polled more Democrats than Republicans.
Here’s the real kick in the avocados… they knew exactly who they contacted due to Fox simply going off of whatever party the voter in question registered as. Here’s the numbers of who Fox polled;
- 46 percent Democrat
- 42 percent GOP
- 12 percent Other/Independent
If you find this particularly outrageous, FNC admits that they have quite the history of pumping out bullshit numbers.
Example, they note that from 23-26 Feb. 2020, they polled 50% Dems, 41% GOP, 9% Indy/Other.
Also, during 27-30 Oct. 2019, they polled 49% Dem, 41% GOP, 11% I/O. How in the world is this fair?
6). MoE for “Subgroups” is where pollsters REALLY lie to you.
Be advised, BEFORE pollsters starts skewing thing with their stated MoE of 3%, they FIRST monkey around with “subgroups”.
Deep in the weeds… in the very small fine print buried deep in the bowels of the poll is where we find out just how bad the lying has been (page 19 of 39).
Before FNC kicks in the overall MoE of 3%, they factor in the dozens of listed “subgroups”.
For instance, here are just a handful examples of how FNC (supposedly a non-partisan poll) paints the picture they want;
The meat and potatoes of the Fox poll was Question 8 (page 5 of 30: “If the presidential election were today, how would you vote if the candidates were: Joe Biden or Donald Trump or Other/Wouldn’t vote/Don’t know”
IMPORTANT! FOX, and no one else for that matter, ever cites the raw numbers… but let’s just throw out a few possibilities.
In regards to Question 8, what if Trump had an overall raw number of 55, Biden – 40, everyone one else, 5.
Fox gets those raw numbers, then REALLY monkey’s around with them (page 19 of 39).
For white RV (Registered Voters) overall, regardless of sex, age, income, education, etc., both Trump and Biden supporters specifically each have a +/- MoE of 5%. Please keep in mind that this can shift the numbers a total of 10 points. So technically, they can recalculate the raw numbers to Trump dropping down to 50, Biden rising 45.
Hmmm… interesting, huh?
But wait, there’s more. As soon as they’re done screwing with the numbers for Trump and Biden supporters, THEN they further shift the All RV’s by an additional 3 percentage points to play with. They can lower Trump again… raise Biden by 3.
Another example, non-white women (a Democrat stronghold), regardless of age, income, education, etc., give Biden supporters a whopping 9.5% MoE. Trump supporters are listed as N/A. In other words, Trump is considered to be so unpopular among non-white women, he’s given a 0% rating.
But we’re not done yet regarding the non-white woman demographic. Once the pollsters change up the Biden supporters numbers by 9.5%, THEN they further change things up with All RV non-white women overall with ANOTHER MoE. This time, it’s another eye-popping 8%.
By the time the pollsters get done shifting the raw numbers to what THEY think we should be told, THEN they kick in the final +/- MoE of 3%.
Ever get the feeling we’re being lied to?
So do I.