On Wednesday afternoon, the hosts of Fox News's "The Five" had a spirited debate about the nature of reality.

The subject, as is always the subject these days, was the 2016 presidential race, a contest in which Donald Trump currently trails Hillary Clinton nationally by about 6 points. He also trails by wide margins in most of the swing states, a more important indicator of where the race is headed.

This has led to a resurgence in the cottage industry of poll-denying, which died out after the effort to "unskew" the polls in 2012. If you don't remember that debacle, it was the concerted effort by supporters of Mitt Romney to argue that polling which showed President Obama likely to win reelection was flawed because of who was being asked. As it turns out, those polls — using statistically validated methods of ensuring accuracy — were right.

Trump supporters (including his chief counsel, Michael "Says Who" Cohen) have used various arguments to dismiss Trump's current position in the polls. One of the most popular arguments is that Trump must be doing well because so many people come to his events. And that's the argument that Fox News's Eric Bolling presented Tuesday.

"Honestly," Bolling said, "we have to stop with these polls. They're insane with these polls. Look what's going on. You look at a Trump rally and there's 12, 15 — 10,000 people."

Dana Perino, former White House press secretary for President George W. Bush, interjected. "You can't ... Eric ..."

"And then you look at Hillary Clinton," Bolling continued, "and you have, I don't know, 1,500? 2,000? It's a poll 82 days out."

"That's a real disservice," Perino countered.

"To whom?" Bolling replied.

"It's a real disservice to his supporters, to lie to them that those polls don't matter," Perino said. "You cannot take 12,000 people at a rally that are your definite supporters, they are going to show up to campaign, and say the polls are wrong."

Greg Gutfeld — hardly a fan of Clinton's — agreed. "One person sitting at home still cancels out somebody at a rally."

But Bolling was undeterred. "Here's why polls really shouldn't matter or shouldn't ever matter. You pick up the phone and you say, 'Who are you going to vote for?' That person on the other end of the phone says, 'Well, I'm going to vote for Hillary Clinton.' They're not out there voting."

There are a lot of problems with Bolling's argument, such as it is, but the most obvious is that last one: "They're not out there voting." Well, no, not right now they're not. But polling works because pollsters have figured out how to best predict who will go vote. There are questions that screen out people who make it obvious that they're not intending to head to the polls, and the consistent accuracy of polling demonstrates that these screens are effective.

But Bolling's broader point is also ridiculous. What polling does is approximate levels of support in huge populations. We can't reasonably survey 200 million Americans in a short window of time, so we come up with statistical methods to take the population's temperature. The best analogy for how this works is a blood test at the doctor's office: The doctor doesn't need to remove all of your blood to figure out what's wrong with you.

Bolling's correct that Trump's rallies are bigger (in part because his emphasis is on having big rallies). But, let's assume a rally of 10,000 people means that Trump gets 10,000 votes. (No sure thing, but bear with us.) If Trump has three rallies of that size every day between now and Nov. 8, that's 2.5 million people coming out and demonstrating their support. If that's every Trump voter, he'd have lost to Barack Obama in 2012 by 63.4 million votes.

Trump's core base of support loves him! They turn out to rallies. But they're clearly a minority of the American electorate. They weren't even a majority of the Republican primaries! More votes were cast against Trump than for him. Crowd size only tells us that Trump has an enthusiastic kernel of support. So did Bernie Sanders, who, you may recall, is not Trump's general election opponent.

Perino quickly pointed out to Bolling that his trust in polls was relative. "But when the polls are good, you say the opposite!" she pointed out, then noting the problem in 2012.

"That's exactly what we said in 2012," she said. "The people that supported Romney were told that the polls are wrong, Romney's going to win. And they were so mad and disappointed."

"Yeah," Gutfeld added. "They stopped watching because they said we lied to them. And we deserved it."

Even Trump seems to understand that his position in the polls is not what he'd like it to be. On Wednesday morning he offered this rather cryptic tweet:

It's apparently a reference to the fact that polling on the British referendum to leave the European Union predicted that voters would choose to stay. Instead, voters chose to leave by 4 percentage points — a six-point swing from the final poll numbers.

A six-point swing would do Trump a world of good right about now. But if crowd size is the only reliable metric to evaluate that, Trump will need one big rally of 7.6 million people — 6 percent of the 2012 electorate — to make Bolling feel comfortable.