I’m a poll junkie. I’ve worked with some of the best pollsters in the country. I’ve devoured hundreds of polls.

Every one of them came with a caveat about methodology. It’s printed clearly at the beginning. A recent example:

“This poll…has an accuracy of +/- 4.0 percent at a 95 percent confidence level.”

Now, the media always reports the 4 percent “margin of error.” (I never worry about it. I tell pollsters they have no margin of error.)

But nobody pays attention to the 95 percent. It means that five percent of the time the poll will just be wrong.

New Hampshire Tuesday was one of those five percent. For 100 percent of the pollsters.

Now the TV bigmouths – this means you, Chris Matthews – are blaming the pollsters for making them look bad.

Why did the polls miss New Hampshire? (For the Democrats, that is. The polls were on target with McCain.) Let’s round up the usual suspects:

• Racial bias? I’m skeptical. Especially among Democratic primary voters.

• Voters lied to pollsters in the exit polls? Could be. But did they all tell the same lie? Did they secretly get together beforehand and decide to play stump-the-pundit?

• The polls stopped polling too early? Some of them, maybe, but surely not all of them. And it’s not clear that late-deciders voted much differently from early-deciders.

• A big turnout by women for Hillary? Bingo. The girls ganged up for her after the boys ganged up against her. She let the mask drop, and we had that rare thing in politics: a glimpse of a real person. Just what she needed, as it turned out. Plus, Hillary’s campaign – with a late boost from the legendary ground-master Michael Whouley – apparently out-organized the others.

Don’t blame the pollsters. They warned you.

Click Here to discuss and comment on this and other articles.