The historical trend I

[homelycooking]

You do recognize that the three largest differences between popular vote percentages among reelected presidents cited in your table were largely caused by the presence of third party candidacies which did much to depress the vote percentage in the initial election, right?

[homelycooking]

You do recognize that the three largest differences between popular vote percentages among reelected presidents cited in your table were largely caused by the presence of third party candidacies which did much to depress the vote percentage in the initial election, right?

So, everything is explained by third party candidates?

Hmm.

Everything? Not at all.

But do you believe that those candidates (Wallace: 13.5%, Anderson 6.6%, Perot 18.9%) had no effect on the popular vote percentages of the Democrats and Republicans running in those elections?

[homelycooking]

Having a substantial third-party candidacy introduces a great deal of uncertainty to the statistical measure you lay out in your original post. (All of the values therein are written out to two decimal places despite this.) You have a significant uncontrolled variable that I thought was worth pointing out, that's all.