It was like watching a mechanical rabbit spring onto the rail of a greyhound track: Over the weekend, the Des Moines Register reported the results of an October poll showing Democratic Senator Hillary Clinton pulling away from her two main competitors for the Hawkeye State’s January caucus vote. And the nation’s pundit corps rallied to the betting windows.

“There’s no question that Iowa’s outcome will be hugely influential for both parties’ nominees, even more so for the Democrats, whose Iowa winner will get a huge blast of momentum,” burbled Time magazine’s Mark Halperin, whose column sometimes seems like a random search engine for warmed-over Beltway conventional wisdom. Mr. Halperin offered up some pro forma caveats concerning the erratic reliability of Iowa polling—but the frenetic “buzz” surrounding the Register results was something he could scarcely contain.

Senator Clinton’s fellow tenants in the Democratic top tier, Senator Barack Obama and former Senator John Edwards, must catch up soon, he writes, “to halt her methodical march to the nomination, since Clinton’s current strength in New Hampshire and in national polls mean she must be stopped in Iowa.” In other words, she must be stopped to, um, be stopped.

Such hyperventilating is an early symptom of the great rolling dysfunction known as the presidential primary season. And now, three months ahead of any actual ballot-casting, the brain damage is beginning.

“Polls are meaningful to campaigns, so that they can adjust what they’re saying to appeal to more voters,” says John Haskell, a senior fellow with Georgetown University’s Government Affairs Institute, and author of Fundamentally Flawed: Understanding and Reforming American Primaries. “They don’t mean anything in predictive terms. The media, if they were doing their jobs, should point out these types of returns just don’t mean much.”

Look no further, Mr. Haskell said, than the Democratic polling on the eve of the last election cycle.

“It was amazing back in Iowa in 2004—the polls did change a lot the last few days, so that it went from Dean and Gephardt at the front to Kerry and Edwards.”

The latter two placed first and second in the caucus vote, and went on to be the party nominees; the former, all early polling to the contrary, finished third and fourth.

Campaigns well know the extra mileage that the press’s subliterate love affair with primary polling can win them. Hillary campaign operatives “are smart to play the inevitability card”—campaign shorthand for impermeable front-runnerness—“and play it hard,” says Mark Blumenthal, who ran the Mystery Pollster blog before heading up the Disclosure Project, a cross-poll collation of the screens used to filter poll results. “Because that’s what drives coverage.”

And so a positive feedback loop sets in: The front-runner, once anointed, supplies the presumptive motivations to all other comers in the field—every move they make gets interpreted tactically, as a ploy to unseat the field leader. The loop then spreads easily to take in voters’ perception of the wider field. “To the extent that these stories continue saying, ‘It’s Hillary, it’s Hillary, it’s Hillary,’ you see a different dynamic,” Mr. Blumenthal said.

That’s no doubt welcome news to the Clinton campaign, which has long been fighting the notion that she’s too “polarizing” to prevail in a national vote. But it’s also another odd displacement: suborning actual voter preferences to the standard theatrical fluff of pundit speculation. Mr. Blumenthal notes that closer inspection of poll results show that, after all, voters aren’t ultimately motivated by the second-order speculations of who’s the most “electable” candidate in a general election—the sort of wifty reasoning that (combined with boatloads of cash) landed the truly weak candidate John Kerry the party’s nomination in 2004.

“The thing that [Des Moines Register pollster] Ann Setzer did in this poll that I really like was near the end she had this question: ‘If you could choose a candidate to be president without thinking about whether they have a chance to win,’ or words to that effect—‘If you were assured your party were to win in 2008,’ the results were barely different. Hillary picked up maybe a point. That proves to me the point I get out of all this—you aren’t going to get all that many voters going on electability.”

Of course, pundits don’t write about such poll results—illustrating as they do the manifest irrelevance of punditry. Instead, they will just loudly rattle well-worn campaign clichés like talismans, in order to ward off the uncomprehended power of actual participants in the primary balloting. Here, for example, is how Mr. Halperin sizes up a key “weakness” in Edwards’ poll numbers: “Edwards’ expensive haircuts and hedge fund ties have muddied his image as a champion of the middle class.”

This, presumably, is in reference to a question in the poll about whether Edwards’ “wealthy lifestyle … undermines his credibility in speaking out against poverty.” But the question makes no mention whatsoever of “the middle class” or its championing—and of course omits the hedge-fund ties and tonsorial excesses of the senator, so Mr. Halperin’s own mention of them is but a somewhat hysterical—though of course enormously telling—insistence on their own deep relevance here, because, one assumes, that’s what Mr. Halperin wants his readers to believe.

What’s more, the poll presented the “wealthy lifestyle” issue as an explanation of Edwards’ off-puttingness among voters who had already chosen another candidate—and even then the media’s much-loved “wealthy lifestyle” meme won assent from just 50 percent of Edwards’ nonsupporters—23 percent deeming it a “major factor” and 27 a minor one.

Of course, the reason campaigns exist in the first place is to change the profile of likely support, and hence the dynamics of an election. Such is the reasoning of the Obama strategists, who are making unprecedented outreach to younger voters in defiance of the traditionally senior-heavy demographics of caucus turnout in Iowa.

“If somehow you could pick out a sample that typically make up the 100,000 who turn out for caucuses, and then you can define 10,000 or 15,000 of your supporters and you convince them [to participate], well, that could be enough to turn the election,” Mr. Blumenthal points out.

“I’m not sure it’s possible to perfectly model all this,” he cautions. “But if you grant that Obama finds this 10 to 15,000, they’re probably not going to turn up [in polls] until about two weeks out.”

In other words, Mr. Blumenthal says, “the most fair and reasonable conclusion is that the Iowa race is wide open.”

Unless, of course, the pundits finally bludgeon the Iowa electorate into submission.