This is a guest post by Judd Legum, who previously served as Research Director for Hillary Clinton's 2008 presidential campaign and was a creator of "Think Progress" at the Center for American Progress. This post originally appeared at Huffington Post and is cross-posted with permission.

There were 23 debates during the Democratic primary and part of my job on the Hillary campaign was to monitor the post-debate reaction in the media. I watched hundreds of self-described political experts instantly declare who won and who lost.

Here's what I learned: the pundits are full of it. They don't know any more than you do and many of them have a vested interest in tilting the scales one way or another. After the debate ends, if you want to know who won, turn off the TV. You can figure it out for yourself.

The first thing to understand is that the winner of the debate isn't the person who makes the best arguments. If it was, Al Gore would be finishing up his second term. The winner of the debate is the person who moves votes to their side. You can figure out who that will be by focusing on these three factors:

1. 30 seconds are more important than 90 minutes. Although tens of millions of people will watch the debate, most everyone will forget the bulk of it immediately. The lasting impression of the debate for most voters will be the two or three exchanges -- usually less than 15 seconds long -- that are replayed, discussed, and analyzed over and over again. More often than not, whoever gets the best of these moments wins the debate.

For example, in the Des Moines Register debate in mid-December, Obama was asked a pretty tough question: How he could rely on so many former Clinton advisers and still represent a break from the past? Hillary laughed and said, "I want to hear that!" Obama flashed a smile and shot back: "Hillary, I'm looking forward to you advising me as well." It was a pitch-perfect response and catnip for the media, which played the exchange repeatedly for days. Overall, Hillary turned in a very solid performance and demonstrated an impressive command of the issues. But it didn't matter. Obama had won the key 15 seconds and it gave him a critical boost just days before the Iowa caucus.

John Edwards was generally regarded as an excellent debater. So why was it that the debates never seemed to help him much in the polls? He never really did anything memorable. (Quick: name one line Edwards said in a primary debate.) His answers were always smooth, coherent and on message. It didn't do him any good.

Identify who got the better of two or three most memorable exchanges between Obama and McCain and you'll be a long way toward identifying the winner.

2. Mistakes matter, but only some of them. Probably the worst mistake in the Democratic primary debates was Hillary's famous non-answer to a question about drivers licenses for undocumented immigrants. But it wasn't a big mistake because people care deeply about the issue. (When is the last time you've heard driver's licenses mentioned on the campaign trail?) It was important because it fit into a pre-exisiting narrative about Hillary that had been developed by her opponents for some time. Namely, that Hillary is politically calculating and dishonest. Since it reinforced a pre-exisiting narrative it caused Hillary immense damage and sent the campaign into a tailspin from which it never fully recovered.

During the next debate in Nevada, Obama was asked a similar question about drivers licenses for illegal immigrants and gave a similarly meandering answer. Yet, he paid no political price. The reason is simple: no one believed at the time that Obama was dishonest or politically calculating. So a mistake that was debilitating for Hillary was a non-issue for Obama.

In this debate, a mistake on an economic issue will be more damaging to McCain because there is a pre-existing narrative that he isn't knowledgeable or engaged on the economy. Similarly, a mistake on foreign policy would be more damaging to Obama because there is a pre-existing narrative that he may not have the experience to be commander-in-chief.

3. It is a popularity contest. At the end of the day these candidates are trying to get voters to like them. As a result, in many instances, what the candidates say is far less important than how they say it.

During the spring and summer, Obama struggled to gain traction in debates because the delivery of his answers were perceived as detached and professorial. In other words, the things he was saying were smart but he wasn't making friends. In an August debate, Hillary won a lot of admirers when she said with a smile: "For fifteen years, I have stood up against the right-wing machine and I've come out stronger. So if you want a winner who knows how to take them on, I'm your girl!"

The person who is the most relaxed and getting some laughs is usually the winner.

The reason why much of the punditry that follows the debate is inaccurate or irrelevant is that many of the people involved are far more interested in shaping the outcome of the debate than reflecting it. It usually doesn't work, but most give it a shot anyway.

You can do a lot better by thinking for yourself.

-- Judd Legum