The New York Times website, and its interactive graphic of votes by Council District, revealed that, with a third of precincts in Bloomberg was only up a point.

Talking heads exploded. On NY1, our local city-centric channel, commentators were caught off guard, with one even making the obscenely incorrect leap that perhaps Bloomberg’s opponent (whose name no one really knew) was riding the coattails of a candidate for Comptroller. Others were quick to discuss What This Means™ and how and why it came about.

I’ve run a few campaigns in my day, making me officially an old-timer – but why the numbers were close was obvious. Bloomberg’s strongest areas were reporting 33% of their precincts in; Thompson’s were at 75 and 80%. In other words, there were still a lot more Bloomberg votes still to come in.

Bloomberg, of course, won by five percentage points – less close than it seemed for about an hour last night, but still light years away from the dougle-digit stomping most of the city expected.

Which was the Mayor’s problem. People turned out to vote against him. Those who supported him, having received, on average, 45.6 million pieces of mail, and seeing polls showing the imminent blow-out, didn’t bother. Bloomberg, in a big mistake, didn’t portray the race as being close in the final weeks – perhaps thinking such a prophecy would be self-fulfilling.

The other problem was one of timing – if Bloomberg’s 5% lead had been consistent all night, instead of fluctuating with the vagaries of what came in when, the story would have been about closer-than-expected, not ohmigod-imminent-apocalypse-and-pollster-suicides.

And there was another factor at play: race.

What follows is a contrast between the final Mayoral results (Thompson is blue; Bloomberg, red) and a map of the ethnic breakdown of New York City (as of 1996, but it hasn’t changed an inordinate amount).

As the AP reported, Thompson crushed Bloomberg among Latinos and African-Americans. Bloomberg did well with Whites and Asians (so much for the John Liu for Comptroller coattails theory).

There is much more dissection to be done on this race, by much more qualified people than myself. But here’s a takeaway – in 2005, during Bloomberg’s first re-election, his opponent garnered 503,219 votes. Last night, his opponent got 506,717. Meanwhile, votes for Bloomberg dropped from 678,444 to 557,059 – an 18% drop. Combine that with getting crushed in two core communities, and you end up with last night.

Update: Here’s another telling breakdown.

It was so close that at around 9:30 last night NBC called the election for Bloomberg and then had to reverse the call when it became too close to call. The NYT.com waded into similar murky territory, first running a banner that said “Bloomberg Wins Third Term as Mayor” and then changing it to “Bloomberg Projected to Win 3rd Term.”

It’s safe to say not a lot of people anticipated this outcome, least of all the Dems. I can’t be the only one who saw those retractions and wondered if someone named Bill Thompson was going to be our next mayor and whether I should be concerned about that. Lessons? $100 million can by you an election, but barely. Also, how many Dems are kicking themselves today (looking at you, Anthony Weiner) for not bothering to throw a hat in the NYC mayoral ring. Apparently the only people not buying into Bloomberg’s multi-gazillion dollar ad campaign were the voters. Why the close call? No doubt plenty of people will spend the day deciphering. But let’s go to Jon Stewart, shall we, who speculated the other night that “money bags” Bloomberg might just be a leetle too out of touch with the voters (video below):

“It’s an important election. But it all starts here, at your kitchen table.” Jon Stewart: “It’s a powerful message. Undercut by the fact that no one in New York City has a kitchen table.”