So I made a graph to disprove the logic of the "but muh 5 out of 6 popular vote wins!"
This graph compares the margin of victory (as defined by Democratic PV% minus Republican PV percentage) on the x axis to the mean of the past six PV margins of victory.
It starts from 1880, as that alllows the past six elections to go back to 1856, the first D vs R election.
The r^2 value is 0.007 (which is basically zero for all intents and purposes). So I'm pretty confident that the successes of Ds in the past few elections doesn't tell us anything meaningful (in and of itself, at least) about their 2016 chances.
But you might protest "1800s are irrelevant for present-day elections! What about recent elections only?" Well, I'm afraid it doesn't get any better. If you were to restrict it to say, post-WWII elections (1948+), the relationship becomes negative. (Of course, that's likely meaningless as well).
Nice graph. It is probably also useful to look at the volatility of the vote, perhaps by plotting the D-R margin, the 6 cycle average, and the standard deviation of the D-R margin during 6 cycles over time.
