Do presidential debates ever have a dramatic impact on presidential elections — or do they only matter in razor-close contests? Before we get to that question, we now know the criteria on which the Commission on Presidential Debates will grant inclusion. Any candidate who clears 15% in the average of five media-sponsored polls will have a podium in the debate — and that might actually matter:

The nonpartisan, nonprofit Commission on Presidential Debates (CPD) today announced the polls to be used in applying the Candidate Selection Criteria for the 2016 general election presidential and vice presidential debates. The polls are chosen with the professional advice of Dr. Frank Newport, Editor in Chief of Gallup.

They are selected based on:

• The reliable frequency of polling and sample size used by the polling organization
• The soundness of the survey methodology employed by the polling organization
• The longevity and reputation of the polling organization

The polls to be used by the CPD in 2016 are:

• ABC-Washington Post
• CBS-New York Times
• CNN-Opinion Research Corporation
• Fox News
• NBC-Wall Street Journal

It’s worth noting a couple of other prerequisites, though. Besides the 15% support level in the aggregate and the usual constitutional requirements, a potential participant must have ballot access in enough states to have a shot at 270 Electoral College votes. That will almost certainly leave out Evan McMullin, even if he could get to 15%. It does, however, keep Gary Johnson and Jill Stein in the running; the Libertarians have ballot access in all 50 states and DC, and the Greens have enough states to compete for 341 Electoral College votes, even without pending court action and petitions.

For right now, though, only Hillary Clinton and Donald Trump qualify for the debates. The last iterations of these five poll series give Johnson and Stein less than the threshold support level. In fact, two of the five haven’t even had polling listed at Real Clear Politics for a four-way race:

• NBC/WSJ: Johnson 10, Stein 5
• WaPo/ABC: Johnson 8, Stein 4
• CNN/ORC: Johnson 9, Stein 3
• CBS/NYT: N/A
• Fox: N/A

It seems passingly curious that the Commission on Presidential Debates would have two polling series included that haven’t produced any data on these races. CBS has partnered with YouGov in state-level four-way polling, and YouGov has partnered with the Economist on national four-way polling, but that’s all. The CPD could have chosen both Reuters and Bloomberg, both of whom have had significant polling on the four-way race over the last three months, but neither have Johnson or Stein anywhere near 15% anyway.

Will the debate matter? Historian Gary May at the Daily Beast thinks it will settle the whole election:

Hillary Clinton faces a unique challenge when she faces Donald Trump. He is sui generis, one of kind, unlike anyone else who has ever sought the presidency. He has no programs she can attack, only proclamations—a wall will be built, law and order will return, new jobs will appear, terrorism will be defeated—all will be achieved, as if by magic, after he takes the presidential oath. So far, that strategy has succeeded, at least in some Republican circles.

Will attacking such fiats as empty promises offered by an unqualified, even dangerous, opponent create sympathy for Trump as Carter’s attacks did for Reagan in 1980? Only if Trump borrows a page from the Reagan playbook and shrugs them off with a smile. But is he likely to do that? His behavior during the Republican debates suggests that he won’t because he is incapable of dealing quietly with anyone who strikes at the Trump brand. Vanity is his Achilles’ heel. Insulting his opponents is his style—“Little Marco,” “Lyin’ Ted,” “Listless Jeb,” “Crooked Hillary.” …

It’s unlikely that Trump will be able to avoid a face-off with Clinton, but he may be able to reduce the number to the one evening—Wednesday, Oct. 19, which does not conflict with football. There is even precedent for holding only one debate—1980, when Carter faced Reagan. However many do occur, it’s almost a certainty that the person who wins that debate will become the next president of the United States.

May lists a number of classic debate showdowns as evidence for this thesis, but at least a couple were relatively close elections in the popular vote, if not in the Electoral College. Ford just barely lost to Carter by 1.7 million votes, so the Poland gaffe arguably mattered — even if it seems much more likely that Watergate and Ford’s pardon of Richard Nixon was the real deciding factor. For that matter, Nixon barely lost to John F. Kennedy, trailing only by 118,000 votes, so again the debate might have made the difference. The rest of the elections May cites were won by 7 million votes or more, which makes May’s argument that the debates cinched the elections less convincing.

Assuming the polling doesn’t change much between now and the debates, it doesn’t seem likely to produce a dramatic inflection point in this election, either — with possibly one exception. If Gary Johnson can qualify for the stage, he might benefit from the weaknesses of the other two candidates. He’s competent on stage, has some natural charisma, and has more executive experience in government than either of the major-party candidates. If he can look good and sound rational while the other two maintain the performance level that has made them the most unfavorable major-party candidates in modern presidential-election history, the debate might vault Johnson into prominence as a None of the Above candidate and challenge the usually binary nature of the contest.

It will be a stretch for Johnson to get to 15% in any poll, let alone an average, of course. But if he does, then the debate could very well become the game changer May envisions.