Since I wrote about Sean Trende’s recitation of the case for predicting an Obama victory (conducted mainly by showing how things had changed since the beginning of the year) yesterday, I should note he’s performed a parallel exercise today in support of the strong possibility, at least, of a Romney comeback and victory.

I won’t go through it in detail, if only because it’s late in the day and you can read the whole piece at your leisure. But he places a lot of emphasis on (a) poll averages that do not indicate–at least yet–the kind of persistent Obama post-convention gains that are being shown by individual polls like yesterday’s bombshell from Pew and (b) a refutation of the “no one trailing in September has ever won” argument by counter-arguing that the late dates of the conventions this year are distorting the timeframe, making such comparisons premature. He also notes, accurately, that the marginally better economy is still only marginally better; that Team Mitt’s paid advertising blitz is just getting underway; and that the ultimate impact of Romney’s various gaffes may well be overstated, based on research concerning past gaffes.

Trende’s piece, along with yesterday’s, are useful reading for anyone bold enough to make hard predictions at this relatively early date. There’s not much doubt the last several weeks have provided a lot of good developments for Obama, but questions concerning how much any of them actually matter, and how actual voters absorb the “fundamentals,” provides enough suspense for me.

Ed Kilgore

Ed Kilgore is a political columnist for New York and managing editor at the Democratic Strategist website. He was a contributing writer at the Washington Monthly from January 2012 until November 2015, and was the principal contributor to the Political Animal blog.