Ah, the opinion polls, with their oft-trumpeted 4% margins of error.

Well, they missed the Governor’s race by a lot more than that, didn’t they?

The consensus, such as it was, gave Governor Shumlin a 12-point edge. Right now, the Associated Press has him at 46.4% and Scott Milne at 45.4%. Feel free to check my math, but I think that’s a margin of one percent. 

The polls were off by almost 11 percentage points.

The difference? Virtually every undecided voter went for Scott Milne. Which is unheard-of; usually, the undecideds don’t all go stampeding in one direction.

Plus, the Associated Press is reporting that Vermont had a record low turnout. The Democratic GOTV machine just couldn’t overcome the broad disaffection with the current administration, and the widespread belief that this election wasn’t close, which made it easier to stay home.

So, Milne got a larger chunk of a smaller electorate.

Let’s take the most recent Castleton Polling Inistute survey, reported on Oct.12.

From Oct. 12 to last night, what happened? Governor Shumlin lost a sliver of his support while convincing no undecideds. Scott Milne gained a whopping ten percent by nabbing all the undecideds and poaching nearly two-thirds of Dan Feliciano’s supporters.

What does that say? It says that Governor Shumlin lost the middle, in spite of all his triangulating. And he lost ALL of the middle. And, I suspect, a fair bit of support on the left, who either sat out the Governor’s race or made a protest vote for Milne or a write-in. (Doug Racine, anyone?)

Or just stayed home, not feeling motivated to re-elect Shumlin and feeling (falsely) secure in the knowledge that their absence wouldn’t make much difference in what was thought to be a Democratic cakewalk.