Two-hundred and seventy. That’s the number of Electoral College votes one needs to become the most powerful person on earth. And according to most of the so-called experts, it won’t be Donald Trump. While I agree that he is the underdog, a Trump victory is not impossible. Here’s how it could happen.

If this were a typical presidential election, the logical first step would be to pick the most likely combination of states to put the Trump campaign over the 270 threshold. Then you would spend all of your time, energy and resources targeting the battleground states. It has worked this way for decades. Pundits, academics and TV analysts are busy re-creating this process using predictive modeling and campaign-honed wisdom.

The reality, though, is that this election is unlike any other in recent history—and Mr. Trump is not your quintessential Republican nominee. If he followed the traditions, norms and standard practices of the past, he undoubtedly would lose.

But there is a path to 270. For Mr. Trump to win, he needs to incite a national movement and a significant shift in voting behavior. Think of the 1970s movie “Network,” with people shouting out their windows: “I’m mad as hell and I’m not going to take it anymore.” This time, instead of shouting they are voting for Mr. Trump. He can worry less about specific states to target and spend more time on a national branding and message campaign centered on motivating specific audiences, not states. Mr. Trump did precisely that to win the Republican primaries.

There has been much criticism of Mr. Trump as a messenger. But what he seems to understand remarkably well is that he has become the megaphone for a large group of voters who feel they have been disenfranchised from the political process, and abandoned on the economic battlefield by both parties. Mr. Trump also knows that “acting presidential” is equivalent to no longer being authentic and believable to his supporters. You’re never going to create an anti-Washington wave of the magnitude he needs if you sound too much like a focus-group-tested, teleprompter-guided candidate.

The other advantage of this type of national campaign is that it costs less than a highly targeted state-by-state campaign. Mr. Trump doesn’t have the pieces in place or the money to run a massive turnout operation, and it is too late to change that significantly. Instead, he needs to use an emotional appeal and create a national dialogue with voters by leveraging the seemingly unlimited supply of time and space the news media is so happy to give him.

Note too that Hillary Clinton’s early money advantage isn’t as important as some political analysts claim. Television ads matter less in a presidential race because candidates get so much free airtime.

To get to 270 electoral votes, Mr. Trump must first win every state Mitt Romney did. I realize that mathematically it isn’t necessary, but symbolically it is. If Mr. Trump were to lose some of the Romney states, it would signal that critical Republican voting groups had deserted him, most likely moderate Republican women or evangelicals.

There are a few Romney states that are attractive for the Clinton campaign to go after, North Carolina in particular. Mr. Trump also seems to be struggling in Arizona, likely because of its large Hispanic and Mormon populations. This is a huge problem, because he has little margin for error.

But let’s assume Mr. Trump carries all of the Romney states. The tempting approach would then be to look at the states where Mr. Romney lost by the smallest margins.

But it would be wiser to look at states with the largest number of middle-income working families, and a sizable population of the children of Reagan Democrats. The largest blocs of those voters can be found in the Rust Belt. This means looking first at Ohio, Pennsylvania and Wisconsin (before turning to Virginia, Colorado and New Hampshire).

The big trio will be remarkably difficult to win. Pennsylvania hasn’t gone for a Republican presidential candidate since 1988, and Wisconsin since 1984. But these states are chock-full of swing voters, many of whom are in sync with Mr. Trump on trade, immigration and national security. I have sat through many focus groups with these voters, and they are starving for a restoration of the American dream. They don’t believe that’s going to happen with a traditional candidate.

The best news for Mr. Trump: With the right nominee, these three states aren’t afraid to vote Republican. Each has a Republican senator, and two of the three have Republican governors. Mr. Trump pretty much has to win two of the three. I think Ohio and Pennsylvania are his best chances, as recent polls confirm.

By putting Pennsylvania and Ohio in the Trump column, combined with the Romney states, Mr. Trump would sit at 244 electoral votes. At this point there would be two ways to get to 270: Win a collection of states such as Virginia, Colorado and New Hampshire, or win Florida. Neither John McCain in 2008 nor Mr. Romney won any of these states. But George W. Bush won all of them—twice.

I think the victory-in-Florida scenario is the more likely. Mr. Trump will benefit from a close relationship with the state’s governor, Rick Scott, and, ironically, by having Marco Rubio back in the U.S. Senate race, increasing Republican enthusiasm and turnout.

More than 30,000 new Republicans registered to vote in Florida this year and many observers believe that Mr. Trump was the driving factor. In recent days, Mrs. Clinton’s numbers in Florida have eroded significantly, highlighting the state’s volatility.

All of this is a heavy lift for the Trump campaign. But who dreamed that Donald Trump could defeat 16 highly skilled and credible adversaries to become the Republican nominee?