Michael Bennet won a Senate race he should have lost. Can Democrats copy his strategy?

Ken Buck's campaign for U.S. Senate in Colorado is already fading into trivia, another example of how the Tea Party screwed up a Republican win. It's being forgotten too quickly. On paper, Buck's loss to Sen. Michael Bennet made no sense.

Bennet, a soft-spoken bureaucrat who had never won an election before being appointed to the Senate, never cracked 50 percent in a public poll. Hours before the election, InTrade.com gave Buck a 60 percent chance of winning it. Larry Sabato, the election calculator that walks like a man, marked down Colorado as a probable Republican gain. And so on.

How did Bennet do this? Well, that's easy. He had 1,900 volunteers on Election Day, he reached out to Hispanic voters, and he raised more money than his opponent, even when you added in the $5.1 million that American Crossroads spent against him. But some of this can be said about the Democrats who went down in the wave, too.

I wrote about the Senate race in Colorado in mid-October, when the White House hadn't bailed on Bennet—Bill Clinton and Michelle Obama both stumped for him—but also when Buck was stubbornly ahead in polls and handicappers had moved on to discussing California and Connecticut as the Democrats' "firewall" to hold the upper house. I was surprised by Bennet's win. The win prompts a question: Can Democrats learn anything from the upset victory of an incumbent in a purplish state? Can they reverse-engineer this election and use the good parts to re-elect Barack Obama?

"Huh," says Mike Hamrick, the chairman of the Democratic Party in Arapahoe County, part of the Denver sprawl that voted for Bush in 2004, Obama in 2008, and Bennet in 2010. "That's a tough question. I think the president and the DNC need to toot their own horns more about what they're accomplishing. We did that here. But it's easier to manage locally."

It sure is. "Colorado was one of many states where the Republicans nominated a fresh face who wasn't ready to have everything he said on the stump be recorded," says Lori Weigel (no relation), a pollster for Public Opinion Strategies in the state. "That's helpful if you're the incumbent and you need to make the alternative to you look unacceptable. Bennet did that."

He had to. According to the exit poll, there were really no reasons that voters would re-elect their Democratic senator. Sixty-two percent of voters said that Congress' next priority should be cutting taxes or reducing the deficit. Fifty-one percent of them wanted the Affordable Care Act repealed. Fifty-one percent disapproved of the job Obama was doing. Right now, that's a good snapshot of the rest of the country that Obama, presumably, wants to win over in two years.

So what can Democrats learn from Bennet? Can they copy it? Mostly, no.

The "Crazy" card can work. The number that mattered in the Colorado exit poll was this one: 54 percent of voters called Buck's views "too extreme." A whole lot of money and media went into making that word, and that attack, stick to him. Bennet was so shameless and assiduous about loonifying Buck that one of his ads even clipped footage of Buck, frustrated, sarcastically asking "I'm extreme? I'm extreme?" Quoth the narrator: "Ken Buck asked the right question."

But the crazy card didn't work for the Democrats in dozens of House races, including two races they lost in Colorado. What was different in this race? The diversity of the "craziness," for one. Bennet's rote attacks on Buck for entertaining the ideas of privatizing Social Security and implementing a national consumption tax were added to a powerful attack on his opposition to abortion in all cases and to some forms of birth control. That, according to FreedomWorks president Matt Kibbe, "divided and distracted" voters. Nationwide, Republicans erased the Democrats' advantage with female voters. Buck lost female voters by 17 points.