Republicans should be winning this election. Trump’s new campaign chief shows why they’re losing.

If you combine the best election forecasting models that political scientists have produced, they tell you that Republicans should be favored to win the 2016 presidential election, albeit narrowly.

If you look at the polls, however, you find Donald Trump is losing this election, and it isn’t even close. Our best estimate, as of today, is that Trump is costing Republicans about 5.7 percentage points in the polls — enough to turn a narrow win into a big loss.

Today’s news that Trump is putting the head of an online conservative tabloid in charge of his campaign helps explain the drivers of the gap.

When making their forecasts, political scientists — not to mention pollsters, pundits, and even voters — tend to assume a rough equivalence in competence: The two political parties will both nominate professional politicians, hire the best campaign staff available, run top-notch campaign operations, and generally fight each other to a strategic draw.

And that usually happens. It happened in 2012, and 2008, and 2004, and 2000, and 1996, and 1992, and 1988, and so on. But it’s not happening this year.

Trump is an inexperienced politician running an amateurish campaign that’s stocked with a B (or maybe C or D) team of Republican staff. And that throws all our normal assumptions into chaos.

Take polling. Most polls look at either registered or likely voters. In both cases, what the poll is telling you is this: Given the people who fit the traditional category of “registered” or “likely” voter, this many support Clinton and this many support Trump.

But polls aren’t elections. What happens in an election is that the two campaigns have to mobilize the largest group of registered voters possible. We expect polling to be accurate because we expect the two campaigns to be roughly equal in their ability to mobilize voters — sometimes one campaign is a bit better than the other, but it’s usually pretty close.

The Trump campaign has no ground game to speak of. As Politico showed, many of the campaign’s state directors are completely unknown by key political figures in the states they’re trying to organize. Trump himself doesn’t seem particularly interested in ground game.

“I don’t know that we need to get out the vote,” he told Fox News. “I think people that really want to vote, they’re gonna just get up and vote for Trump.”

So is it really so safe to assume that Trump will mobilize as large a share of his potential voters as Clinton?

Trump’s decision to bring in Steve Bannon, the CEO of Breitbart News, to lead his campaign follows much the same model. Just as we assume that both presidential campaigns will have ground operations helmed by top staffers, we assume both campaigns will be led by the best campaign managers available.

Clinton’s campaign, for instance, is run by John Podesta — chief of staff in Bill Clinton’s White House, founder of one of Washington’s largest and most important think tanks, and a top adviser to President Obama. This isn’t Podesta’s first rodeo.

Trump’s campaign is now being led by a guy who has never run a serious campaign, has no experience running serious campaigns, and has no experience relevant to running a national presidential campaign.

It’s actually worse than that. Bannon is a communications professional who only knows how to communicate with far-right conservatives. Breitbart doesn’t court, and hasn’t attained, a broad audience — Bannon’s instincts might be sound when it comes to what hyper-confrontational, news junkie conservatives want to read, but there’s no reason to believe he has the first idea about how to appeal to swing voters. In that way, he mirrors Trump perfectly — and that seems to be the real reason Trump hired him.

This was a winnable election for the Republican Party, and a number of Republican candidates looked ready to win it. Back in March, Marco Rubio led Clinton by 4 points in the polls, and John Kasich led by 7. Trump trailed Clinton even then, and now it’s looking less and less likely that he’ll make up the gap.

But his standing in the polls may actually underestimate his weakness in this election: So many of our assumptions about how presidential politics works are based on the assumption that both sides are pretty good at running presidential campaigns. Trump is blowing that assumption apart, and I’m not sure any of us knows quite how to update our predictions to account for it.

What’s clear, though, is that Republicans took a race they could have won and nominated someone likely to lose it, perhaps quite badly. Trump is an abnormal candidate running an abnormal campaign, and the result is that the Republican nominee is underperforming the fundamentals of this election, and may have yet further to fall.