Here are a few headlines from the last few years:

Now, here’s what Donald Trump had to say in an interview with the Associated Press on Tuesday:

In the interview, Trump outlined a general election campaign that banks heavily on his personal appeal and trademark rallies while spurning the kind of sophisticated data operation that was a centerpiece of Barack Obama's winning White House runs.

"I've always felt it was overrated," Trump said. "Obama got the votes much more so than his data processing machine. And I think the same is true with me."

In the same interview, Trump said he wouldn’t release his tax returns. As I write elsewhere, that would be dangerous step against disclosure and transparency for the entire republic. His pooh-poohing of data poses a danger, too, but mostly to the Republican Party.

For years, the two parties have been in an arms race over how to deal with the new capabilities opened up by technology and data—things like finding potential voters, advertising to them, and getting them to vote, though the whole suite of things is often shorthanded as “microtargeting,” which is just part of it. And for most of that time, Democrats have been the acknowledged leader.

Republicans were aware of the missile gap headed into the 2012 election. They even knew that President Obama’s team was developing something called “Narwhal” to centralize its data. So Mitt Romney’s campaign developed a countersystem called “Orca,” for the narwhal’s leading predator. It turned out to be a disaster. The system crapped out on Election Day, leaving the Romney campaign frantically without information it needed and contributing to his loss to Obama.

In its post-2012 autopsy, the Republican Party pointed to the need to address that as a major goal:

Identify a team of strategists and funders to build a data analytics institute that can capture and distill best practices for communication to and targeting of specific voters. Using the GOP's data, the data analytics institute would work to develop a specific set of tests for 2013 and 2014—tests on voter registration, persuasion, GOTV, and voter mobilization—that will then be adopted into future programs to ensure that our voter contact and targeting dollars are spent on proven performance. These tests should be the first order of business of the analytics team and should incorporate pollsters, data managers, and messaging professionals at the table developing a variety of approaches that would be subject to measurement.

The GOP won big in 2014, but that win was generally attributed to a different factors: The widely varying demographics between midterm and presidential elections, for example, or Republicans’ massive success in redistricting after the 2010 Census. Headed into the 2016 cycle, the GOP was still viewed as trailing behind Democrats. Now comes Trump, saying he won’t bother with data. (To be fair, Trump aides have suggested in the past that they really are building an info effort.)

Trump is right that he his campaign has succeeded largely on his personal appeal. He’s probably wrong about Obama, or at least incomplete: A good candidate isn’t enough. Just ask Donald Trump, who’s repeatedly complained that Mitt Romney lost what should have been a winnable election. One x-factor in that race was Obama’s more sophisticated campaign.

Nor is it clear that what worked for Trump in the primary election will work in a general. During the GOP primary, he was competing for a group of voters who were more or less committed to the Republican Party. During the general election, he starts at something of a demographic disadvantage, or at least with a political landscape that’s highly polarized and has few swing voters. In such an environment, winning those swing voters becomes less important, while driving turnout among one’s owns supporters grows in import. Trump’s experience during caucuses—especially in Iowa—raises some worries. Without much of a ground game in caucuses, Trump repeatedly got beaten by Senator Ted Cruz, though he was still able to win the nomination thanks to primary wins.

Here’s the problem for GOP: If Trump is wrong, it will produce a Democratic president, and could deprive the party of presidential coattails needed to perform in down-ballot races, especially in the Senate, which Democrats hope to capture. Even if Trump is right, it seems fairly clear that he’s a one-of-a-kind candidate, and the party can’t rely on recruiting someone like him next time there’s an opening. Without the muscle and money that a presidential campaign can provide, whatever the RNC is doing won’t turn out as well as it might have.

As Republican wrestle with the Trump nomination and what it means for the party, some leaders have suggested that the party concentrate on House and Senate races, and try to distance itself from its presidential nominee. What if Trump goes through with his no-data pledge? Will the RNC turn away from the Trump campaign and let him sink or swim on the force of his charisma? Or will the party bite the bullet and try to fill in the gaps?