It's May 20, there are 171 days until the presidential election, Donald Trump is the presumptive Republican nominee for president, and everything is normal. This is the premise of Hillary Clinton's general election strategy, which her campaign detailed in a conversation with the Washington Post.

If there's anything my extensive background in watching action movies has taught me, it’s that you never publicly lay out your strategy in advance. The first reason for this is that doing so can take some of the fun out of actually watching everything unfold. The second, more germane reason is that revealing your plan of attack affords your nemesis the chance to figure out how to counter it.

So why did Clinton cast off conventional wisdom passed down by the action-movie gods? It's part of what political analysts and election junkies call a "general election pivot." She’s attempting to signal that, in her opinion, the primary is over, she will be the nominee, and it's time to shift the party’s focus to defeating the Republicans in the fall. It's the equivalent of pointedly turning away from someone you've been arguing passionately with and asking a third party why they haven't washed the dishes.

The reason the Clinton campaign was willing to sacrifice strategic surprise is obvious once you learn the details. Here's the core of her general election, anti-Trump message, as conceptualized by her chief campaign strategist, Joel Benenson:

This isn’t about bluster. It’s about having real plans to get stuff done. When it comes to the economy, Hillary Clinton is the only candidate with plans that have been vetted and will make a difference in people’s lives.

Those real plans? Job creation centered around manufacturing and small businesses, and steps to make the economy a bit more equitable — pretty much exactly what you’d expect from a Democratic front-runner. The reason the Clinton campaign was willing to give away its strategy is because there's absolutely no surprise involved, nothing here to indicate that Clinton considers Trump an unusual candidate or that he presents any unique threats. This is the same kind of election message she could have run, with very few changes, against Bush, Rubio, or Cruz. Her campaign is treating this as a conventional election and pitching itself as a continuation and extension of Obama's policies, especially its economic policy. And in normal circumstances, running in the wake of a decently popular incumbent and an unemployment rate of 5.5 percent, it'd be perfectly reasonable to do just that. Things are going in the right direction; let's keep it that way.

As with the affirmative case for Clinton, her campaign’s case against Trump is also pretty standard: Trump's business record with its multiple bankruptcies makes it difficult for him to be believable as caretaker of the American economy. "Credibility matters,” Benenson says. “I don’t think he’s got credibility to make that argument." This is not that different from the case the Obama campaign made against Mitt Romney in 2012.

The argument that this workaday message will stop Trump is straightforward. It's based upon the idea that presidential elections, barring those that occur during recessions or wars, are not determined by the particulars of the candidates, but instead by demographics and economics, an argument well summarized by Slate’s Jamelle Bouie. The thinking goes like this: Trump is all but doomed because there won't be enough white voters to make up for his incredible unpopularity among blacks and Latinos. Romney clobbered Obama with whites but lost anyway because minorities voted against him. And that was in 2012; Trump's nativist appeal might drive minority turnout even higher this year. The campaign and candidates matter less than the math. And the math is just against Trump.

I know what you're thinking: Wait a second, didn't the pundits tell us that Trump couldn't be the nominee in the first place? Why should we believe them when they screwed up so badly the first time? These are valid questions, and whether you should believe pundits now depends on what you think their earlier screw-up was. If, like Nate Silver, you think that pundits’ primary mistake was not putting enough stock in polls that showed Trump leading the GOP since last summer -- essentially, not putting enough faith in the math -- then you should trust the pundits and conventional wisdom. And you should be pretty confident in Clinton's strategy.

It's not necessarily that simple, though. The problem wasn’t just that pundits didn't put enough faith in the polls. The real reason so many people got their primary predictions wrong is that they were basing them purely on precedent, without a good idea of what made the preceding events happen. Lots of well-informed people thought that Trump's views would make him unacceptable to the establishment, and that candidates who are unacceptable to the establishment can't win presidential nominations in modern elections. The reasons to not treat Trump's polling numbers as predictive were actually valid: Candidates wither after strong polling all the time. Just take a look at the polling from 2008: Rudy Giuliani led the polls for months and months before his lead collapsed.

The reason these predictions ended up being wrong is that candidates unacceptable to the establishment not becoming the nominee is not the consequence of some ironclad law of politics. It's the result of an underlying mechanism: The party has to put into action a plan to stop the unacceptable candidate, and the plan has to work. Neither of these things happened this time, and so predictions that Trump could never grasp the nomination were wrong.

If you don't have a great idea of the underlying causal factors, those that are actually driving the polls, then all you can really do is naïvely apply precedent: What happened before will probably happen again. And in normal circumstances, when the underlying factors aren't too different than they usually are, you'll make the right prediction. But when they are different than they usually are, you can screw up massively.

This matters because the case that, electorally speaking, Trump is a conventionally unpopular candidate with unfavorable winds blowing against him and thus will be beaten more by circumstances than by clever strategy -- essentially, that everything is fine -- isn't just based on polls. The polls show only a narrow lead for Hillary and are probably too far out from the election to be very predictive. It's also based in the idea that, given his unfavorability ratings among minorities, and the fact that his campaign is rooted in nativism, Trump will find it hard to expand his appeal beyond the share of the GOP base that he's already convinced to vote for him. And it's based on the idea that campaigns and candidates "don't matter," at least not as much as demographics and economic conditions.

But just as the idea that the party decides their nominee is not some immutable fact about the world (but rather a result of the party taking some action to decide), the idea that campaigns and candidates don't matter is predicated on those parties choosing reasonably qualified candidates and running reasonably competent, conventional campaigns. In other words, campaigns and candidates don't matter in a world in which the candidates have been rendered acceptable to the establishment by passing through the primary process, and are thus seen as avatars of that party's stances. But that’s not the world of math anymore. That's the world of precedent, and we're in much shakier territory there.

Because Donald Trump is, quite literally, without precedent. He has a profile of someone we've never seen as a presidential nominee before. You don't have to take my word for it — listen to Bouie: "[Donald Trump is] a racist, nativist demagogue with few ties to government, no experience in public office, no service in the armed forces, and little to no knowledge of anything involving governance." As Bouie notes, we've had candidates that embody one or two of these qualities, but never has a candidate embodied all of them at once. We don't know what kind of effect this will have in a modern presidential election, because we've never had a candidate like this before. So from here on in, we are for all intents and purposes flying blind. All of us.

For instance, the Washington Post’s Greg Sargent, citing the work of demographer Ruy Teixeira, says, "All of the things that Trump might say and do to drive up white turnout — particularly working class white turnout — would also likely drive up nonwhite turnout. So there’s no reason to expect a major boost in turnout from one group and not the other." This is not the inexorable result of math. It's an educated guess. And educated guesses, for this cycle, are going to be about the best we can do.

The way Trump has run his campaign is weird, too. It's not just all the offensive stuff only he can say without seeming to suffer any consequences, or his shoestring budget, or the fact that he got most of his publicity for free from the media instead of paying for advertising. Instead of running on the right, then pivoting (there's that word again) to reposition himself in the center for the general election, Trump essentially ran to the middle and is now tacking to the right with the nomination in hand. Trump's list of possible Supreme Court nominees, for instance, is meant to appeal to the kind of conservatives who might be tempted to either vote for Hillary or sit this election out. And it does.

Now, certainly, all of these things that are unconventional about Trump might work against him rather than for him, might narrow his appeal instead of expanding it, and they might lead to an even more thorough Democratic victory in the fall. But I don't know, and the truth is that nobody else knows, either.