Why I Don’t Like Pollsters

Pollsters like to remind us as to how accurate they are. If they were half as good as they say, we could avoid the whole election craziness; the pollsters would tell us the week before the first primary or caucus who the ultimate winner was going to be and we could just get on with our lives.

I took multiple statistics and associated classes in college, so I have at least a basic understanding of sample size, margin of error, and scientific wild-ass guesses. In the first place, I challenge their statistical sample because it doesn’t include people who match my demographic. Why? Because people like me refuse to talk to pollsters; we do not answer any telephone call that doesn’t come from a known number. However, I do vote in every election.

Next, someone or some organization is paying for these pollsters. In the event of murky data, whose side so you think a polling company will lean toward?

Finally, let’s be honest—predicting the future is not just tough, it’s pretty much impossible. Look at meteorologists. They tell me that what makes it such a great job is that if they’re right 30% of the time, they’re considered a genius.

Case in point—last weekend, as Hurricane Matthew headed east, out into the Atlantic Ocean, the meteorologists predicted my area would get about two inches of rain. Remember—the National Weather Service has sensors all over the country. They launch weather balloons to check the weather at higher altitudes. They have radar, Doppler radar, and whatever is the newest whiz-bang technology. They have satellites with cameras, infrared detectors, and Ouija Boards. They have weather spotters on the ground. They have computer models. Besides the meteorologists at the National Weather Service, every television station has a platoon of their own meteorologists.

How did my two inches of rain turn out? Try 10.86 inches of rain and tropical force winds that knocked down trees and power lines accompanied by flooding.

Advertisements

3 responses to “Why I Don’t Like Pollsters

  1. Hola Steve: I don’t like agreeing with you all the time, but I must agree with you about polls. First, have you ever noticed that those damned “robo-calls” always ask questions about our “position” on an issues…rather than our “beliefs” on things? Asking us what we believe about “how the world works,” for example, is more insightful, informative, significant, enduring, and predictive than asking someone to “guess” how they might vote on an issue. Not only that, but polls are not only for gaging who’s voting how, polls are also for “changing people’s position” on issues and candidates. Yet, it’s changing what people BELIEVE that is important. Believing is more powerful than agreeing.

    And there’s something else about polls. They cannot take into consideration human nature: The human will and spirit–plus the energy of MOMENTUM. BREXIT is example and proof. Against all odds, the people of England came out of their homes in droves from towns and country-sides to VOTE OUT the European Union–and the old-guard establishment too.

    No, Steve, I won’t quote your favorite Inverse Square Law, yet I will remind you history is replete with battles and wars won by virtue of the power of the human will and spirit–and the moment by moment increased force of momentum which is measured in OUTCOME…and polls be damned.

  2. I agree that poling tends to be flawed, however, how do we square that with the idea that aggregate polling data (at least) tends to be relatively accurate?

    • I actually wrote a reply a few minutes ago, but it disappeared. I’ll try again.
      Samuel Clemens said, “There are lies, damned lies, and statistics.” I agree. In college, on the first day of my first statistics course the professor announced that by the end of this basic course we would be able to prove that a particular piece of land should be exclusively zoned for high priced residential homes, or that the neighborhood qualified for poverty relief.
      1. The most accurate polls are very general; an actuary can predict how many people of a certain age will die next year, but not who and not on which day.
      2. Political polls are a few questions; accurate polls tend to be 50+ questions that are cross referenced to weed out false responses.
      3. Many political pollsters are not interested in an honest representation. They want to produce an outcome that supports their client. The first question they ask will determine which way a respondent is leaning and whether the interview should continue or to hang up the phone.
      4. When reporting on polling accuracy, they may not cherry pick, but I’m sure that the results of every type of poll is included.
      So, was my answer:
      (a) Amazingly brilliant
      (b) Insightful and intelligent
      (c) Well thought out and clearly stated
      (d) All of the above

      All the best!
      Steve

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s