It's started again, the quadrennial obsession of presidential campaigns, news organizations and, indeed, the entire political punditocracy with polls, especially those in the crucial early voting states of Iowa, New Hampshire and South Carolina.

Just recently, we learned in one survey that Jeb Bush has dropped into the single digits among "likely" Iowa caucus goers, an almost impossible category to gauge at this stage, but leads Hillary Clinton among registered New Hampshire voters.

Jim Messina, fresh from a double triumph of masterminding President Barack Obama's re-election in 2012 and British Prime Minister David Cameron's in 2015, might have overstated the case on MSNBC when he said "most public polling is garbage." But he probably provided an apt assessment of most surveys taken this far from any real decisions by voters.

More coverageOpinion and analysis about the 2016 presidential campaignCartoonsCartoons: The race to the presidency in 2016 CartoonCartoon: Early for 2016?

Messina might have been inordinately influenced by the failure of many polls in both countries to predict the scale of the Obama and Cameron victories. But their predictive value at this point is sufficiently minimal that readers can save a lot of time by disregarding much of what will be breathlessly reported in the next eight months.

In 2011, transient GOP national poll leaders (setting aside ultimate nominee Mitt Romney) included eventual also-rans such as former Texas Gov. Rick Perry, former House Speaker Newt Gingrich, former Pennsylvania Sen. Rick Santorum, real estate mogul Donald Trump and Atlanta businessman Herman Cain.

Unfortunately, national poll numbers are being used to determine participants in the initial Republican debates. That seems unfair to those whose current low poll numbers - unlike in the middle of a general election campaign - reflect name identification or prominence far more than political potential.

advertisement | advertise on newsday

But if the polls are pretty worthless, these other aspects of primary campaigns are worth watching. Here are five things to keep in mind: Televised debates, however participants are determined, will likely be significant. They can establish or destroy the standing of candidates, especially those only superficially known. Just ask Perry, whose 2012 campaign never recovered from his failure in a debate to name the three Cabinet jobs he proposed to abolish. Many believe Hillary Clinton's 2008 downfall began with her fumbling response to a debate question about driver's licenses for illegal immigrants.

Iowa campaign developments between Christmas and the currently scheduled caucus date of next Feb. 1 might have greater impact than most of what happens before. Obama's 2008 Iowa campaign was dead in the water until the Illinois senator's bravura performance at a party dinner less than two months out. Underfunded GOP candidates Mike Huckabee and Rick Santorum emerged as potential winners only in the weeks before the 2008 and 2012 Iowa caucuses.

New Hampshire's pre-Iowa pecking order will likely change dramatically in the eight days between the Feb. 1 Iowa caucuses and the Granite State's Feb. 9 primary. The classic example occurred a generation ago, when Democrat Gary Hart converted his distant second in Iowa and a strong New Hampshire organization into an upset victory over frontrunner Walter Mondale. Hillary Clinton's victory over Obama only developed in the final 48 hours before the 2008 primary.

Sign up for The Point

Go inside New York politics.

Straw polls are even less valuable than early public opinion polls. Last weekend, a poll at the Southern Republican Leadership Conference showed Dr. Ben Carson in first place. He has a definite following, but only slightly more chance of being nominated than Sarah Palin. A generation ago, straw polls often reflected genuine political strength, but more recently they've mainly reflected which campaign bought the most tickets. Decisions by Huckabee and Jeb Bush to bypass next August's Iowa straw poll are hopeful signs. The Democrats had the good sense to ban them three decades ago.

None of the first four tests - Iowa, New Hampshire, South Carolina and Nevada - has a perfect record in predicting GOP nominees. South Carolina had one until Gingrich upset Romney in 2012. While Iowa and New Hampshire can't determine a winner, they can - and usually do - narrow the field.

To be fair about it, poll questions about voters' attitudes are more valuable than the headlines about their candidate preferences. And they're not going away. After all, it's eight months until Iowans cast the first real ballots that will ultimately determine the 45th president.

Carl P. Leubsdorf is the former Washington bureau chief of the Dallas Morning News.