Donald Trump’s greatest fear is probably public humiliation ― and there’s a decent chance he’s headed for a historic one.

The most unpopular presidential candidate in modern history is George McGovern, whose 1972 race against Richard Nixon ended with him raking in a puny 37.4 percent of the vote. There’s at least one surface similarity to today’s politics: Though it didn’t happen in anywhere near as public a fashion as the GOP flight from Trump, the Democratic establishment abandoned McGovern, the darling of lefty activists.

The political fallout was immense. McGovern haunts Democrats to this day. Hillary Clinton worked on his behalf in Texas that year and often uses his annihilation as a warning to progressives not to push too far, lest you wind up with a Nixon. The party transformed itself in subsequent years, culminating with the centrist presidency of Bill Clinton.

Just as the left was thoroughly buried in 1972, the country now has the opportunity to bury not just Donald Trump, but the racist, misogynistic politics he stands for. A complete repudiation in full view of a global audience would send a clear signal that Donald Trump is not America.

The Huffington Post crunched some numbers, and as of now, there’s a 23 percent chance Trump’s vote share on Election Day will fall below McGovern’s.

Trump is lagging pretty far behind Hillary Clinton in the presidential race. According to the HuffPost Pollster aggregates, he trails by 7.7 points in a two-way race with Clinton and by 6.5 points when third-party candidates are included.

In most polls, Trump is stuck between 38 percent and 42 percent support. In polls that include third-party candidates, his probability of dipping below the mark is 23 percent. His numbers are higher in two-way polls; if those are closer to accurate, he has only a 4 percent chance of achieving the low mark.

There is, it turns out, a role in this biggest-loser crusade for people who don’t want to vote for either Trump or Clinton. Any vote that is cast for someone or something other than Trump puts the tiniest dent in his vote percentage, since it increases the total number of votes. So your Constitution Party vote, or your write-in vote for “WEED,” actually has a purpose this year.

To calculate all this, we took all the polls that were completely conducted in the last 50 days before the Nov. 8 election ― in other words, polls that concluded on or after Sept. 19. According to an analysis of elections from 1952 to 2008 by political scientists Robert Erikson and Christopher Wlezien, polls that are conducted 50 days before the election predict about 75 percent of the variance in the actual vote. Polls’ predictive power increases steadily from there to about 95 percent on Election Day.

The Sept. 19 cutoff gives us 87 polls that include third-party candidates, and 65 two-way Trump vs. Clinton polls (as of Thursday, Oct. 27). Trump’s support ranges from 37 percent to 49 percent in the head-to-head polls. When third-party candidates are included, his support ranges from 36 percent to 44 percent.

Of course, the polls get more indicative of the outcome as they’re conducted closer to the election, so we weighted Trump’s vote shares by the number of days before the election. The weights were calculated using the approximate variance explained by the polls that ended that number of days before the election in Erikson and Wlezien’s analysis. Polls conducted on day 50 (Sept. 19) would receive a weight of 0.75, and for every day toward Election Day, that weight increases by 0.004.

This weight makes the most recent polls more important, and increases the uncertainty of the polls according to how many days before the election they were conducted. And that’s key. We know there’s uncertainty in the polls, but the further out from the election we are, the more uncertainty there is for predicting the final outcome. The average of Trump’s support weighted by this date factor is our baseline for where Trump is now.

The weighted average of the three-way polls is 39.2 percent support for Trump. The weighted standard deviation ― which tells us how much the poll estimates vary on average using the same weight ― is 2.4 percentage points. To figure out how likely it is that Trump gets 37.4 percent or less of the vote, then, is a simple statistical calculation.

We assume that errors can go in either direction, above or below the current polling: Trump could gain among third-party or undecided voters; he could lose votes among those who said they’d support him (say another scandal erupts); or the polls could be off in either direction. So we use a normal distribution that assumes a 50 percent chance of Trump’s vote share being above the weighted poll average and a 50 percent chance of going below that weighted average.

The outcome we want to know about ― 37.4 ― is 1.8 percentage points less than the weighted average of 39.2 percent, which converts to a 23 percent chance of Trump getting 37.4 percent or less of the national popular vote.

Doing the same calculation with the two-way polls nets a much smaller 4 percent chance that Trump will come in at or below 37.4 percent. That’s because the weighted average of Trump’s support in these polls is higher ― 42.4 percent with a weighted standard deviation of 2.7 percent.

Research shows that actual electoral outcomes usually fall somewhere between the results of polls that include third-party candidates and those that do not. Going by that rule of thumb, Trump’s overall chances of getting down to or below 37.4 percent of the national vote are probably between 10 and 15 percent.

This is, of course, an approximation. But it illustrates that Trump has a nontrivial chance of becoming the biggest loser in modern presidential history.