If this happens, take a breath. It won’t be the first time there was a tie in Iowa. That’s sometimes the most accurate way to describe the outcome of the caucuses—and it is often obscured by the desperate need of the news media to declare a winner.

Here’s what happened in the 2016 Democratic caucuses:

Hillary Clinton wound up with 49.84 percent of “delegate equivalents.”

Bernie Sanders wound up with 49.59 percent.

Does this mean Clinton got more votes from a narrow plurality of caucus goers? Nobody knows, because before this year, Iowa Democrats never reported actual votes. Maybe Sanders' votes were concentrated in fewer precincts, or maybe Clinton’s barely perceptible margin of victory came from supporters of candidates who didn’t survive the first round. The 2016 contest between Clinton and Sanders was, by any reasonable measure, a tie. To say Clinton “won” is at best a reach; and it would have been journalistic malpractice to proclaim her a “badly wounded front-runner” had Sanders prevailed by two-tenths of 1 percent of a “delegate equivalent" formula almost no one understands.

Nor was this the first time Iowa’s caucusgoers deadlocked. Four years earlier, the GOP battle between Mitt Romney and Rick Santorum came down to a literal handful of votes. Late into the night, the Republican Party announced that Romney had won by eight votes. Sixteen days later, a final tabulation showed that Santorum had actually prevailed, with 29,839 votes to Romney's 29,805, a 34-vote triumph.

If you screen the footage of the coverage, you will witness CNN frenetically searching for the last handful of votes to determine who “won,” as if that fact would decide something consequential. The fact is, nobody won. A race that comes down to 34 votes among almost 60,000 is by any reasonable calculation a dead heat.

Keep this in mind when the 2020 totals are reported. It’s possible that a few percentage points will separate the top candidates. If a few hundred, or a few dozen, votes separate the leading contenders, it will make little sense to draw a distinction between “a surprisingly strong third” and a “disappointing, perhaps fatal, fifth.”

But if past is prologue, that is more than likely to happen. So why the desperate hunt for a winner?

Elections Need Winners. Presidential Primaries Don’t.

In November, the voters actually determine who winds up holding an office, or, in the case of the presidency, which electors get to cast votes for the president.

And it’s true that in a November election, there has to be a winner, no matter how close the margin. In 2000, George W. Bush won Florida by 537 votes out of some 6 million cast, giving him all 25 electoral votes and thus the White House. Eight years later, the Senate race between Norm Coleman and Al Franken in Minnesota led to a months’ long series of recounts and court battles. In the end, Franken was declared the winner with a margin of 312 votes out of nearly 3 million ballots cast.

But presidential caucuses and primaries are different. No one has to win them. They are steps in a process of delegate accumulation that leads to a convention. They aren’t contests that must produce victor and vanquished.

But the news coverage of these contests too often focuses on a “winner” as if coming in first in Iowa or New Hampshire has the same decisive consequences as it does in the fall.

Once upon a time, this made some sense, because the candidate who came in first place won all the delegates from a state. When George McGovern won the 1972 California primary with 43.5 percent of the vote to Hubert Humphrey’s 38.5 percent, he won 271 delegates.

But after 1972, Democrats—unlike Republicans—abolished that winner-take-all rule. Now, “winning” a state tells you little about how many delegates the “winner” has corralled. When a talking head tells you who “won” a state, wait for the delegate count. It’s not impossible that a statewide runner-up ends up with more delegates than the popular-vote “winner.”

No One Knows if Iowa Matters This Year

In the 1970 movie Little Big Man, the character Old Lodge Skins, played by Chief Dan George, goes to an Indian burial ground to die, offering incantations to the Great Spirit. After a while, the very much alive Old Lodge Skins observes: “Well, sometimes the magic works. Sometimes it doesn’t.”

That’s the story of Iowa. Sometimes, as with Jimmy Carter in 1976, John Kerry in 2004, Barack Obama in 2008, it propels a candidate to the nomination.

But sometimes, as with George H.W. Bush in 1980, Bob Dole and Dick Gephardt in 1988, Mike Huckabee in 2008, and Ted Cruz in 2016, it doesn’t.

The principal impact of Iowa has been to end the candidacies of long shots—a fate that Amy Klobucher and Pete Buttigieg are surely eager to avert. There’s no reason to jump to sweeping conclusions from results that are produced by a relatively tiny number of participants in a famously unrepresentative state.

Forty years ago, after George H.W. Bush upset Reagan in Iowa, one of NBC’s most prominent analysts declared, “I should like to suggest that Ronald Reagan is politically dead.”

Those words should be laminated and handed out to anyone looking to offer judgments about how what happens in Iowa will affect the rest of the 2020 presidential election.