Maybe the Pollsters Got It Right
The current President of the United States has a disease that cannot be cured by public pressure or smart people telling him he’s full of shit. Donald John Trump is in a farce bubble. Compounding the symptoms are his enablers who live in fear that he will explode and bring the whole structure of government and their existence down to the ground. They are all failures in government and will be marked for life as losers and liars. Did Trump win the election? Why would he say he did? Well, he’s also a liar.
Vox.com features excellent reporting and writing. This was their headline the weekend before the election: “Biden has a big lead in the polls, but can we trust them? You mostly can rely on the polls, but that doesn’t mean Trump can’t win.” I love how some people who are paid to predict the future always have something in their pocket for a grand excuse when they lose.
This is a good time to fill in some detail about my background. For more than thirty years I was an international radio broadcasting consultant, specializing in music formats. A major part of that job was designing, conducting, and analyzing research polls to determine what the public wanted to hear when they turned on the radio. We probed why a particular station was their favorite and why they listened to others. We studied the disc jockeys, songs and formats with high audience appeal and listenership. If we found the client station had low appeal with listeners, we made wholesale and dramatic changes. The job involved big risks taken with the potential of a huge rewards, like placing Howard Stern on the air at client properties. The basis of success in music radio is totally driven by pollsters at radio ratings companies.
There is a notion in statistics called the “margin of error.” It’s typically a small amount that helps users of a study understand the possible impact of errors, such as miscalculation or change of circumstances. Now one might ask, why the “miscalculation” Larry? I once knew a researcher named Larry, thus the reference. Larry was a straight shooter. He would answer that question with lots of words about demographic distributions, geographical weighting and probabilities. You see, research based on a sample of respondents is only as good as the composition of the sample. For example, if you are trying to understand the perceptions of African American voters, you need a strong representation of Black people in the sample of those you survey. You must balance the sample so it contains the same total percentage of groups that exist in the population you are surveying. If your sample has more or less of the type of participants you wish to measure, you must have a scientific method of adjusting the results. If you’re thinking there’s a lot that could go wrong, you are absolutely correct. Good research involves strict, scientific discipline with a reliable margin of error.
In the old days, Gallup used 3,000 people to determine what the whole country thought about everything from cigarettes to carports. But as America diversified and changed, that kind of simple sample became unstable and unreliable. Yes, the pollster got the 2016 election wrong, but they’ve gotten other things less and more important equally wrong. First, people don’t always tell the truth. Next, the method of contact has a major effect. In the old days of radio listener research, a survey began with recruitment of respondents using a home phone. Because it was easier, many people from the same household were used for the same survey. It could be that Dad filled out the survey for all the kids, meaning their use of the radio was not measured at all. Researchers knew this, of course, and factored the probability of such possibilities into the margins of error of their radio listenership reports, “the ratings.”
In today’s world, the internet is not a reliable instrument for recruiting survey respondents. It’s too easily manipulated by evil forces. We all know that. So, with each call that a surveyor makes in a study to find out who the American voters will choose, there is an assumption made about the probability of the person on the phone actually voting how they say they will. Of course, they could be speaking the truth on the call but later change their mind before the vote. Stuff like this happens all the time and researchers adjust their margins of error accordingly.
In 2016, former poker player and famous pollster Nate Silver and his FiveThirtyEight website gave Hillary Clinton a 72 percent chance of winning the election. In 2020, the same group forecasted that Donald Trump had a 10 in 100 chance of winning while Joe Biden had an 89 in 100 chance of winning: So, FiveThirtyEight didn’t get it wrong. As I am writing this, Biden has 77,242,061 votes while Donald Trump has 72,096,110 votes. Biden got 290 electoral votes to Trump’s 217.
I thought the videos of Trump supporters in states where he was leading chanting “Stop the Vote!” while in states where the President was behind, they were screaming, “Count all the votes!” were asinine. In baseball, we don’t stop the game in the fourth inning because our team is ahead, and we certainly don’t add innings after we’ve lost. This is the most dramatic demonstration of “population density” that I have ever seen. When I say density, I mean STUPID SHITS.
When the pollsters told Joe Biden’s campaign that Georgia looked like it was moving toward the Democrats, Joe, Kamala, Barack and many other notable Democrats got off their butts and travelled to the Peach State to pitch their guy. This was like Trump in 2016 seeing that Michigan, Wisconsin and Pennsylvania were in the play for him. He went there and it made a difference. The Democrats simply used the orange man’s playbook and it worked. Hey, if you can’t read and correctly interpret the numbers, you’ll never know how to work them in your favor.
When analyzing research, there are times you see what you want to see. You see big numbers, but ignore the margin of errors. When you see a possible error ranging from plus-four to minus-four, you need to take four off your guy’s score and add four to the other guy. That is the correct way to study the numbers. I always knew this election would be close because I kept my eyes on the margins of error. When trying to judge baseball teams against each other at the end of the season, the team with the most errors is not a good one. Large errors matter.
Did they say Biden would win? Yes. Did Biden win? Yes. So, before everyone goes on TV to yell at the pollsters, stop and think. Polling is but one of the many “sciences” that openly admit they may have errors. When you hear the new coronavirus vaccine has a 90% effectiveness, look at the fine print. There is a margin of error. For one, the sample doesn’t include the extremely ill or the very young. No babies were given the serum. Remember, it’s the sample that most deeply affects the results of any study. And not to be a bit negative, but a vaccine with 90% effectiveness means that 10% of the sample would not be protected and could even die.
I know these are hard times for people to understand. The emotional upheaval of COVID-19, people losing jobs and the political vitriol, make it easy to blame pollsters for not getting it exactly right? They were conducting these surveys under high pressure during a terrible time in the history of America. They got it right, even though not perfect. Just like America.
Brand New Book for the Recovery
How to Hire Great People: Tips, Tricks and Templates for Success
Great companies hire great people. This short, easy-to-read book will help you recruit, review and refocus your new workers into the style and culture of your company. Motivating people to do great work will manage turnover and keeping good workers at your company will maintain your success. Employee inspiration makes a positive difference in our competitive world. HOW TO HIRE GREAT PEOPLE covers everything, including testing, training, tricks and tips. Follow this guide and you’ll assemble strong teams with smart workers, and you’ll learn some time-tested techniques about how to keep them.
Kindle and Paperback Click Here
The book that tells it like it is…
Gold, God, Guns & Goofballs: If you only read one chapter of this book, try “Take a Knee for America” and think about our never-ending conflicts between minorities and the police. I’m not asking you to take a stand but having a deep and honest conversation about why some people think the way they do would be productive. This is a book for the moment which seeks to start a conservation about peace. And if you are worried about social media, you really should check out the chapter called “Social Media Menace.”