This is what has to happen for Donald Trump to beat Hillary Clinton

So. Donald Trump will be the Republican nominee for President. Can he beat Hillary Clinton, the likely Democratic nominee?

It's not likely, but it could happen. Here's how.

People who study political polling are all agreed: There is no way that Donald Trump can beat Hillary Clinton and become US president this year. It's simply not going to happen.

But Trump keeps winning things that people said he wouldn't win. Time called the death of Donald here. The Guardian called it here. US News called it here. Even Nate Silver, the polls expert, said Trump wouldn't win the nomination.

So can Trump pull off another unexpected result and beat Clinton in November?

Here's the state of the head-to-head polling for Trump and Clinton, from RealClearPolitics:

In recent days Trump appears to have narrowed the gap to just six percentage points. Hillary is comfortably ahead, and always has been. Her support has always stayed above 44%, usually more. Trump can barely crack 44% on his best days.

There were two points, in December 2015 and February 2016, at which Trump apparently pulled within striking distance. In the December poll, the pair were separated by less than a point - well inside the margin of error. Most people are writing off those two polls as outliers. They appear as sudden deviations from the trend, and the following polls showed no diminution of Clinton's overall lead.

One polling company, Rasmussen, published a poll on May 2 showing that Trump might beat Clinton in a scenario where voters who would normally stay at home, don't. Trump won that one 41-39. That poll looks like an outlier, also.

The University of Virginia's Center for Politics has a great map of what that looks like in the electoral college - a landslide for Clinton.

Nonetheless, there are several polls in that RCP series where the difference between them is only three or four points. If these two were any other candidates, jumping that gap would look doable for the Republican.

That's what Ronald Reagan did to win in 1980:

(Some people think that chart is a product of Gallup's historic polling under-representing Reagan's support; polling back then was not as sophisticated as it is today.)

The implication is that this scenario might play out in favour of Trump: The media and the pollsters underestimate Trump's appeal, as usual. Trump now steadily gains ground by tacking toward the political centre. Clinton puts in an uninspiring performance. And Trump wins by a narrow margin in November, the theory goes.

Another scenario in which Trump wins was outlined on Quora by Matthew Gagnon, a Republican political strategist from Maine. Now, before you start freaking out, Gagnon makes it clear that he doesn't like Trump and doesn't think Trump can win. But the rest of his analysis is an interesting speculation.

It goes like this: If you're white and don't have a college degree, and you lost your job or saw your wages cut because of the globalisation of free trade, then you're angry at both the Republicans and the Democrats right now. Both parties cheered the free trade deals that gutted the good manufacturing jobs from the North and Midwest of America, and Trump has a staunch anti-NAFTA, anti-China position.

Gagnon writes:

The number of these voters who have been depressed, ignored, dismissed, and outright ridiculed is tremendous. And there are a lot of them in states like Pennsylvania, Ohio, Michigan, and Wisconsin. A lot.

That means that these states -- particularly Pennsylvania and Michigan -- which have been Republican targets for nearly thirty years to no avail, have a very real possibility of flipping.

How real is that "very real possibility"? Perhaps not as real as Gagnon thinks because this demographic will shrink from 36% of the electorate in 2012 to 33% in 2016, according to FiveThirtyEight, the respected polling and stats website. In other words, even if Gagnon's theory is accurate Trump is dependent on a declining portion of the electorate.

Having said that, FiveThirtyEight has a nifty interactive graphic that lets you adjust levels of support among the various demographic groups so that you can plot various election scenarios. It uses the 2012 election, which Obama won with 51.7% of the vote, as its baseline.

If you're a Trump supporter (or if you buy Gagnon's theory), then try sliding the scale for "non-college educated white people" from 62% Republican support in 2012 to 69% today. You get this:

A Trump victory in the electoral college with just 48.2% of the vote

If you believe that under-educated whites are the key to the election then Trump needs an increase of seven points in Republican support among that group. That's a big leap, but not impossible. That result also assumes all the other demographics vote the same way as they did for Obama in 2012 - which isn't likely.

The appeal of this scenario is that Trump loses the popular vote 48.2% - 50.1% to Clinton but because of the placement of his support he wins in the electoral college. Al Gore lost the 2000 race to George W. Bush in much the same way.

Again, the likely result is that Trump is so offensive to Democrats and moderate Republicans that he drives them to vote for Hillary, or to not vote at all, thus handing Clinton the White House.

But stranger things could happen.