Just ignore all those polls that say Hillary Rodham Clinton has the lead in the race for president in 2016. Polling on presidential preferences this far out is meaningless.
The pollsters who test hypothetical matchups are hanging a lot on their opening line: "If the election for president were being held today ..." But the truth is, the election is more than three years away, and a poll now cannot tell you how people will vote in the future.
Three years from now, the results of those polls might look like they were on point. Then again, they might look as silly as Gallup's July 2005 finding that Rudy Giuliani would beat Hillary Rodham Clinton in the 2008 presidential race by 50 percent to 43 percent.
Even though people are willing to express preferences when pollsters ask for them, there's no way to know if those preferences will hold up through three years of campaigns, news and our own ever-evolving lives. Mostly, head-to-head matchups now measure partisanship, whether a person has heard of either candidate and whether the voter likes them.
So it stands to reason that the better use of polls is to measure what people know and feel about those who might run for office. Although voters' current impressions of the candidates don't tell you much about how they might vote three years down the road, pollsters start to collect that kind of data now to establish a baseline.
The first time Quinnipiac University measured voters' impressions of Barack Obama in February 2006, 59 percent of them said they had no idea who he was. Fast-forward 980 days and 53 percent who cast a ballot actually voted for him.
The story of what happened in between those two reads on Obama came partly from pollsters who asked the same questions at multiple points in the campaign and looked to see how things changed.
As they did in 2006, Quinnipiac recently conducted a poll to gauge voters' starting-line views on a slew of political figures who could run for the presidency in 2016.
Of the 16 people who may harbor aspirations for the White House that were tested in the poll, Clinton was best known, with 98 percent offering an opinion, and also one of the most positively viewed. Joe Biden is also widely known, with 92 percent offering an opinion on the current vice president.
From there, the public gets fuzzier, and therefore more susceptible to persuasion about a possible candidate.
Among leading Republicans, about a quarter said they didn't know enough to have an opinion on Biden's opponent for the vice presidency in 2012, Rep. Paul Ryan of Wisconsin. And 1 in 5 are unsure about New Jersey Gov. Chris Christie or former Florida Gov. Jeb Bush. Three in 10 say they're not sure about former Pennsylvania Sen. Rick Santorum, who made an impressive run at the Republican nomination for president last year.
Majorities say they haven't heard enough to form an opinion about Democrats like Massachusetts Sen. Elizabeth Warren, Virginia Sen. Mark Warner, New York Sen. Kirsten Gillibrand and Maryland Gov. Martin O'Malley; or Republicans like Louisiana Gov. Bobby Jindal, Texas Sen. Ted Cruz, Wisconsin Gov. Scott Walker and New York Rep. Peter King -- nearly all familiar to those following prognostications about the 2016 campaign to come.
As that campaign gets going, we'll know which candidates are gaining traction outside the political establishment, whom they're attracting and whom they're repelling because of just this kind of early polling, and in a more meaningful way than if all we had were hypothetical vote preferences.
Quinnipiac University poll: http://www.quinnipiac.edu/institutes-and-centers/polling-institute/national/release-detail?ReleaseID=1932
Jennifer Agiesta is the AP's director of polling.
Follow Jennifer Agiesta on Twitter at http://twitter.com/JennAgiesta