2016 polling trends are eerily similar to 2012's (relatively)

[Adam Griffin]

I looked at RCP's running average for 2012 and 2016, wanting to see if there were any real similarities between the two elections. Unfortunately, it wasn't very easy at first, since the rolling averages' embeds on the website are displayed at two different resolutions. In order to compare them, I had to proportion the two to one another.

With a simple and straightforward measurement, measuring from July 1 of the year before the election to June 28 of the election year for both 2012 and 2016, you get the following result:



It tells us nominally where each candidate was during a relative portion of the election cycle, but it's a bit messy and doesn't tell us much of anything else.

However, I was more interested in broader patterns that would not necessarily be so obvious with a nominal measurement. This was where separating the two from one another once again was needed, but it was also necessary to keep their horizontal lengths/x-axes identical (since the two RCP embeds are different widths) for a valid one-year comparison. The end result is one that restores the vertical proportions/y-axes to their original proportions on RCP while maintaining identical horizontal proportions; this means that the margin between Trump/Clinton on the graph below is scaled down to one-third of 2012's. That yields the following:



I drew a basic polygon based on the 2016 election patterns for this one-year period, and then overlaid it onto the revised 2016 graph. While not perfect, they actually correspond pretty well: a period of contraction where the Republican gains ground, followed by several months of relatively stable polling, followed by an expansion where the Democrat gains ground, followed by another contraction where the Democrat gains ground, followed by either a stabilization or reversal (yet to be seen in this cycle) leaning in favor of the Democrat.

[Virginiá]

Fantastic catch Griff!

[Ljube]

Of course they are similar. Democrats have the popular vote advantage and have had it since 1996.

[Adam Griffin]

The pattern continues...

LOL

[Dr. Arch]

The difference in the overall max %s the main party candidates are getting are indicative of either a significantly disaffected number of voters or a strong third-party run.