Friday, August 17, 2012

The Persuadables

Buzz and I have come to a saddening conclusion.  This year's Presidential election will be decided by people who really don't give a damn about Barack Obama or Mitt Romney.

Unlike Buzz, me, and all of our readers, the people who will decide this election just aren't engaged in politics now or probably ever. 

Buzz begins his day by pouring over all the latest polls and punching that data into the Univac 3000, checking for subtle changes which might give him even the smallest insight into electoral trends.   He then takes this material and produces a multitude of charts, diagrams, and graphs to supplement the OVC Daily Briefing memo.  All of this is presented at the morning staff meeting.

I begin my day by checking the latest WNBA news and setting my lineup for my WNBA Fantasy Basketball team.  I don't mean to brag, but I've won the league seven years in a row, but I digress.

Poll after poll in the last few months have consistently shown two things - each candidate has a floor percentage of which they seldom, if ever drop below, and a ceiling, of which they never rise above.  Most changes in the individual polls can be attributed to statistical sampling variations.  In other words, poll results vary greatly depending on who you sample.

The trick to good polling is to tailor your sample to represent the voting public as a whole.  If you have too many Democrats in your sample, your poll win inflate Democratic numbers, and vice versa if you include too many Republicans.  So what the pollsters do is massage the numbers to reflect past voting trends and put forth a calculated guess as to how the raw data collected translates into the numbers they release in the poll.

Some polling outfits are run by Democrats or liberal leaning folks, and some are run by Republican or conservative folks.  These polls may favor the political leanings of the folks conducting them, or contrary to what would seem to be common sense, may tend to overcompensate for the pollsters' biases and lean in the opposite direction of the pollster. 

After all, pollsters do not make money from the general public on their polling, they make money by the candidates or organizations who commission the polls.  The more accurate the polling results are, the more chance the pollster has of getting work.

With all of this said, Buzz tells me that if we averaged out all of the polls, a sort of poll of the polls, we'd be able to minimize the individual prejudices and biases of the pollsters.  I told Buzz that RealClearPolitics has already done this.

The RCP average of polls is an excellent source of information, and believe it or not, it has been a very accurate predictor of election results over the past ten years.  RCP does include what the pollsters refer to as an outlier (a poll that seems to be out of whack with all the others,) but do to the averaging of all the recent polls, an outlier won't throw the numbers off that much.

Over the past few months, there has been very little variation in the numbers.  Obama's numbers have ebbed and flowed, as have Romney's. Neither has dropped below the middle to low forties, and neither has jumped above the mid to high forties. 

What this means is that both of the major party candidates are guaranteed to get somewhere in the neighborhood of 45 percent of the vote come November. The trick is getting at least 51 percent of the remainder.

These undecideds are what this election is all about.  They are referred to as the swing voters, but Buzz and I like to call them the Persuadables.  Whoever wins the majority of these voters will win the election.

This presents two issues as Buzz and I see it.  First, you gotta identify these voters, and number two, you gotta convince them that you deserve their vote. 

We'll talk more about how this works later.  Gotta get to the morning meeting.

1 comment:

Local News PA said...

I read your blog..all are nice post..Thanks for sharing this post...