Coverage of Michigan’s Democratic primary revealed a significant divergence between data journalists and those reporters who focus on storytelling.

The data journalists were convinced the uncommitted vote was not terribly significant, while the storytellers saw it as nearly earth-shattering.

Anyone who watched TV news primary night or read a newspaper the following day got a clear sense of the breathless coverage storytellers offered.

The geeks differed.

Nate Silver, a father of data journalism, concluded, “Uncommitted didn’t do well by any reasonable benchmark in Michigan…If anything, a little bullish for Biden insofar as it suggests that the protest vote over Gaza might not be all that large.”

Kyle Kondik, of University of Virginia professor Larry Sabato’s Crystal Ball, also took issue with the story tellers, “…the coverage went a bit overboard, particularly because ‘Uncommitted’ didn’t really get that big of a share of the vote.”

G. Elliot Morris, formerly of The Economist and now leading ABC/538, laced his reaction to the coverage with sarcasm, “y’all have fun over-interpreting the Michigan results…. It looks like the ‘Uncommitted’… is going to be ~2-3 percentage points higher than in 2012. I am sure everyone will cover this responsibly and nothing will get blown out of proportion!”

Which brings us to our first empirical fact: After a substantial campaign, “uncommitted” did about 2.5 points better than it did against Barack Obama in 2012, when no one was urging voters to cast an “uncommitted” ballot.

“But,” protest the protest voters and the story-telling journalists, “100,000 people supported ‘uncommitted.’”

Boston College professor David Hopkins pre-butted that argument, “…Dem turnout being way up from 2012 doesn’t mean anything. MI passed a law in the interim that allowed voters to automatically get sent mail ballots for all elections if they wanted; many voted this year who wouldn’t have bothered before.”

Hopkins concluded, “Any interpretation of the MI results…that isn’t ‘big wins for both Trump and Biden’ is straining too hard to make drama out of these numbers.”

The numbers aren’t the only shortcoming in this narrative.

A CNN anchor began an interview with this assertion, “More than 100,000 Democrats just voted uncommitted in the Michigan primary this week as a way to protest President Biden’s support for Israel in this war.”

The truth is, neither the anchor, nor the interviewee, nor anyone else has the slightest idea how many Michiganders cast uncommitted votes to protest Biden’s support for Israel.

I have no doubt some people did. But how many? Anyone who says they know is speculating.

In the past, lots of Democrats vote uncommitted without a campaign and without an issue.

In this case, Armenian American leaders asked Michigan voters angry about administration policy toward Azerbaijan to vote uncommitted, and Democrats who think the president’s too old, or hasn’t done enough to forgive college loans, or enough to reduce climate change, or who don’t like his immigration or crime policies have reason to vote uncommitted.

Without careful polling, no one can begin to guess how many of this year’s uncommitted voters cast that ballot based on the president’s support for Israel against Hamas.

Moreover, the story tellers assumed, those who voted “uncommitted” will not vote for Biden in November absent major policy shifts — and said so repeatedly.

Again, there’s not an iota of evidence indicating how many would act on that threat.

And such projections of future behavior are notoriously inaccurate, even if we had a poll — and we don’t even have that. All we have are the hopes and wishes of campaign organizers reported as hard realities by journalists looking for a story. No one even bothered considering how many votes the president would lose by shifting his position in the dramatic way uncommitted organizers are demanding.

Nationally, we know the pro-Israel community is far larger than the anti-Israel segment. So, the president could well lose more votes on the issue than he gains in Michigan and elsewhere.

Moreover, such a move could make a deeply principled president appear to be a crassly political flip-flopper, turning a position issue into a character issue. Unless you know how many votes the president would lose by changing positions, you can’t assess the impact of caving into the demands of the 13 percent.

You won’t get these counter arguments and questions by interviewing only those on one side of the issue.

You won’t get all sides of the story by stationing your correspondents in Dearborn, the heart of Michigan’s Arab American community, and in Ann Arbor, a university town that is the farthest left locale in the state.

When your reporters are there but not in any Jewish areas, not in Oakland County, which produced the largest vote in the state, or anywhere else, you cannot get the full picture.

And when reporters lead off by reading tiny precinct results from those locations, one gets the sense that “uncommitted” is doing vastly better than it is, biasing the rest of the evening’s discussion.

When the subject is data, stick with the experts and turn the storytellers down low.