One last election wrap-up post to finish off a grim week. Dave Weigel went back and looked at the national electoral map generated by one poll-aggregation site as it stood on October 3, the day of the Mile High Massacre. Result: It looked … exactly like the eventual electoral map did. If you’re using 10/3 as your point A and 11/6 as your point B, you’re plotting a straight horizontal line.

But the polls, per RCP, did move a lot in those 33 days. Check out how Romney’s national numbers soared:

Who cares about the national numbers, though? They were wrong, weren’t they? Well, here’s what happened after the debate to the state polls in Virginia:

And in Florida:

And in Ohio:

Electoral analysts will be debating for years why Romney’s momentum stalled, whether it was debates two and three that cooled him down or whether debate one was a simple bounce like any other that was destined to fade or whether it was the, er, hurricane. But look again at the Ohio graph; he never quite grabbed the lead there, did he? In fact, in several key swing states, Romney never led in a single state poll after August, even during his post-debate surge. He never led in Nevada, never in Wisconsin, and never in Pennsylvania except for that one Susquehanna poll sponsored by the state Republican Party. In Michigan, he led exactly once. Even if he had maintained his momentum in the three states featured above — Ohio, Virginia, and Florida, worth a combined 60 electoral votes — and ended up winning them, he still would have lost the election very, very narrowly. That’s the sort of advantage that Obama had. In fact, let me share with you three of the more frightening paragraphs I’ve read this week:

The starting electoral map in 2012 offered Democrats 431 ways to win and 76 for the Republicans. This includes Wisconsin as a swing state, which I think we have learned it never really was, campaign bluster aside. Discounting that state, it was 230 ways for the Democrats to the Republican’s 26 (and that includes ties!). From the first day, 2012 was played entirely on the Republican side of the 50 yard line wether they wish to acknowledge it or not.

Since 1992, New Hampshire has only gone red only once, in 2000. Similarly, Iowa has only gone for red once in 2004. Of the total 60 electoral votes during six election cycles, Republicans have only won 10 of them and it wasn’t even during the same election. Does ‘swing state’ really apply then, when you lose 85% of the time? With the base of states that Democrats have won in 6 straight elections of 247 and adding in New Hampshire and Iowa, it appears Democrats can routinely count on starting with 257 electoral votes, just 13 shy of a win.

Such a beginning electoral map would leave Republicans with 3 paths to the White House and one of them involves winning both Colorado and Nevada, states Democrats have now won twice in a row by fairly large margins.

This is what people mean when they say the GOP needs to “expand the map.” Not all of this is structural; the Examiner makes a good case that some of the reason Obama’s won two elections comfortably is Obama himself. Maybe young voters and black voters won’t turn out as much for a less charismatic, non-historic nominee. Maybe the Democratic ground game of 2016 won’t be quite as sophisticated as Obama’s once his staffers are no longer involved. My point, though, is that not everything has to go right for them. The Democrats have such a huge margin of error now that even if they had somehow fumbled away the big three above, they still win. Daunting.

For what it’s worth, GOP “insiders” polled by National Journal say they favor Marco Rubio in 2016. (Democratic insiders favor Jeb Bush.) Chris Christie finished fifth among them, behind even Rick Santorum, even though he leads in New Hampshire according to PPP’s first 2016 poll. After consulting extensively with my very sophisticated model, I project Christie’s odds at the nomination as being … low.