Less than 48 hours later, the Des Moines Register's page-one banner said: "Big shakeup: Cruz soars to lead."

Well, maybe they're all right. Perhaps Trump is, as The Times poll found, gaining ground nationally. Maybe that's true despite, as The Journal poll determined, voters don't like his plan for dealing with Muslims. And it could be that Iowans have, as the Register reported, a shifting view.

At best, polls provide a snapshot of how people feel at the moment the poll is taken. At worst, they mislead us. Regardless, they are like drugs to which news editors and cable commentators have become hopelessly addicted.

Writing on the New Yorker magazine's site, John Cassidy provides a useful history lesson:

"At this stage in 2003, Howard Dean was leading John Kerry in the polls by 8 percentage points. In mid-December, 2007, Hillary Clinton was leading Barack Obama by 18 points, and on the Republican side Rudy Giuliani had a 5-point lead. On December 11, 2011, the Real Clear Politics poll average showed Newt Gingrich with a 12-point lead over Mitt Romney...which suggests the national polls shouldn’t be taken too seriously."

If only Cassidy had stopped there. Instead, he went on at length about how polls showing Trump far in front might actually be correct. Of course, that was 24 hours before the Iowa poll showed Trump tanking. So, who really knows and who really cares?

There are so many polls that sub-contractors such as Real Clear Politics are needed to sort through them. This produces what commentators love – a single number – by averaging many polls. A good slogan for the process would be: We don't know if it's right, but it fits in a headline!

At least the major polls utilize the best science available, which isn't saying much, but it's a lot better than what's offered by bottom feeders like Frank Luntz, the GOP polling consultant and Fox News regular. Luntz favors televised "focus groups," and the one he staged recently was a doozy.

He rounded up 29 people who said they voted for Romney in 2012 and were now leaning toward Trump. Then he invited news media to observe as he questioned them. The coverage, across the spectrum of newspapers and TV networks, was stunning. Not what was said, mind you: that was mostly blather. Rather, it was the way in which responsible media were drawn to this silly exercise like poll-craving addicts.

The Luntz sideshow was even cited by John Cassidy in his New Yorker piece, as if 29 people, performing for cameras, matter even slightly in evaluating national voting trends.

Sean Trende, the elections analyst for Real Clear Politics, says "there is a very good chance" that no one will win the GOP nomination via the primaries, and that the selection will be brokered at the convention in July.

Yet, that won't stop Trende's company and countless media outlets from obsessing about meaningless polling data for the next seven months. The real clear likelihood of that is 100%.

(c) Peter Funt. Distributed by Cagle syndicate.