Looking at the numbers, you could see why Kay Hagan was one of the most vulnerable incumbents at the beginning of the year. While NC is a purple state, Romney carried it by 2 points in what was a pretty good year for Democrats. Once you factor in Democrats dropping off of midterms, you’re looking at an electorate that decisively rejected Barack Obama. Now, some candidates have crossover appeal, even though it looks like that didn’t come through at all for Landrieu and Pryor. But Kay Hagan started off with none.

An incumbent generic Democrat, running in reddish-purple North Carolina in a midterm, with no significant crossover appeal, has a very, very difficult path to reelection. To be reelected, Hagan would have to do several things: first, eviscerate her Republican opponent and make him entirely unacceptable to the electorate. Considering the makeup of the voters, that’s quite a task. Second, run a near-perfect campaign. And finally, cross her fingers and hope there’s not a Republican wave. Because if there is, none of the other stuff matters.

Turns out, Hagan did run a near-perfect campaign. She had a few missteps, but none of them (like missing out on the last debate) were decisive in the outcome of the election. She pounded Tillis with ads throughout September and turned the Senate contest into a school board race. She positioned herself as a moderate – the most moderate in the U.S. Senate – and pretty much made Tillis and the North Carolina legislature, not Obama, the subject of the campaign. In September, pollsters started seeing incredible percentages of North Carolinians saying public education was their top issue. Basically, Hagan’s campaign briefly convinced voters that Republicans were destroying education in this state, and this destruction was so severe that Tillis was absolutely disqualified from being a U.S. Senator.

Polls taken at the end of September had Hagan around 45%, Tillis around 40%. But worryingly for her, the education attacks really didn’t increase her support significantly, rather it had the effect of lowering Tillis’s support and pushing his voters into the undecided column. Still, the Republican looked wounded. The party establishment had clearly overestimated the House Speaker. Going into October, he looked like a sure loser, and for a time it looked like the NRSC was not going to compete in North Carolina – an incredible fact, something no one would have believed at the outset of the race, and a testament to the success of Hagan’s strategy.

The Tillis campaign found their groove in October by focusing on foreign policy and Hagan’s stimulus mini-scandal. Hagan’s biggest mistake, it turns out, wasn’t anything related to her campaign, but to her job. Skipping that ISIS briefing to attend a fundraiser was an extremely potent line of attack for the Tillis campaign. Combined with the stimulus scandal, Hagan’s personal image took a beating. Voters, who have amazingly short memories, stopped worrying about education. Bickering about textbook funding seemed remarkably petty when confronted with ebola and attacks from ISIS. Tillis got a reprieve from the NRSC, who decided they were going to play ball here. North Carolina Senate was competitive again.

Still, it looked like too little, too late for Tillis. The polls showed him cutting into Hagan’s lead, getting oh so close to her in the surveys, but oh so far. It appeared Hagan’s September attacks had simply wounded Tillis too much for him to prevail. Republicans braced themselves for heartbreak. Kay Hagan, the one who got away.

But while observers continued to describe Hagan as “resilient”, there were ominous signs in the polling for her that few picked up on. For one thing, Sean Haugh was still getting an unusual amount of support, and there were still a lot of undecideds. That added a lot more uncertainty to the outcome.

And yet, Hagan maintained a small, but persistent lead in the polling averages. In early voting, Hagan’s ground game looked strong. A lot of people were overly rosy about the early voting totals for Democrats, but all indicators pointed to a significant improvement over 2010.

We now know the real story. Haugh voters either stayed home or backed Tillis (he did get an unusual amount of support for a Libertarian, but still underperformed his polling averages). The undecideds went overwhelmingly for Tillis. They had been weighing their options all year. They strongly disapproved of Obama and were extremely skeptical of Hagan. Her claims of being a ‘moderate’ didn’t wash with them. But they didn’t like Tillis either and had strong concerns about the direction the General Assembly was taking the state. In the end, they said: “Oh, screw it. I’ll vote for Tillis. I really don’t have a choice.” Turns out, Hagan was not an option for these voters. Just like in 2010, no Democrat running for U.S. Senate was going to win in North Carolina. We just didn’t know that until last night.

And the Democrats’ big advantage in early voting? Turns out, having a big lead in early voting isn’t an advantage if you’re just “cannibalizing” your Election Day supporters. Perhaps the most surprising thing of all was the low turnout – 44%. That’s hardly up from 2010. Tillis and many other observers were predicting 50% turnout. $100 million spent and the electorate was virtually unchanged from 2010.

Hagan ran a better campaign than Tillis. But you don’t get bonus points for running a great campaign. It brings to mind an old joke told by those in the medical profession – “What do you call someone who graduates last in his class at med school? Doctor.” In the same vein, what do you call a Senate candidate from North Carolina, running in a GOP wave year, who leads an unpopular state legislature, who has trouble raising money, who is prevented from campaigning because of work-related obligations, who gets defined early on as intent on destroying education and giving the savings to yacht owners, as hostile to women and minorities, and suffering more negative attacks than any Senate candidate in U.S. history?

You call him Senator.