I disagree with one part of the Silver headline.
For the last 80+ years (I'm using 1928 as the cutoff since 1924 was skewed by a strong Independent, and the strength of the Democratic party from 1860-1928 wasn't as strong as that of the Republican party today), there has been a steady pattern of parties gaining support, peaking and losing support until the other party wins. And the cycle repeats.
Obviously, the next election could be an exception due to a variety of circumstances (a political realignment, a poor Republican candidate, an especially favorable environment for Democrats, an exceptional Democratic campaign, a weak environment in 2012 skewing results for that election, etc.)
All that has to happen for the formula to continue is for the 2016 Democratic nominee to get less votes than the 2012 nominee. It's still possible to win under those circumstances. FDR peaked in 1936, and slowly lost support in 1940 and 1944, but he still stuck in the White House. Republicans peaked in 1984, but Bush managed to stick around in 1988.
