Last week, Republican presidential nominee Donald Trump declared himself “Mr. Brexit” in a post on Twitter (a place where all important news is disseminated).

His apparent point is that his candidacy — which, according to the polls, is down anywhere from five to 10 points, depending on the survey, against Democratic presidential nominee Hillary Clinton — will resemble the vote Great Britain held back in June. Across the Atlantic, despite polling suggesting a different outcome, Britons voted to leave the European Union by a narrow four-point margin.

The lesson we were all supposed to take away from Brexit is that no matter what the polls say, your vote does matter and you should cast it wisely.

Here in the United States, that lesson seems to be forgotten.

Trump and his campaign have been a lightning rod for criticism, much of which has been self-imposed. But many in the punditry world have taken their scoffing a step further, which is to suggest the polling gap will be too much for Trump to overcome. He’s done. We should consider third-party options. We should start focusing on down-ballot congressional contests. And we should start looking to 2020.

The media seem to be sending mixed messages, on the one hand defying the notion that your vote matters. If Trump has lost before Labor Day, then what’s the point? On Election Day in 78 days, why wait in line for an hour or two to cast a ballot if this contest was over long ago?

On the other hand, there is this caveat that, a la Brexit, anything could happen. The polls can change. Things can tighten or things could shift in a way on the eve of Election Day that won’t be picked up by the polls, which are a lagging indicator.

In 2012, pundits made a similar suggestion about the polling. After the political conventions wrapped up, then-GOP presidential nominee former Massachusetts Gov. Mitt Romney lagged behind incumbent President Barack Obama.

Conservatives maintained the polls could be wrong. They argued these polling groups were undersampling Republicans and oversampling Democrats, and that after the election results came in they would be proven correct.

As it turns out, the GOP leading pundits were wrong. Obama outperformed the final national general election polling, winning by nearly four percentage points over Romney, whereas Real Clear Politics polling averaged showed Obama with a 0.7 percent lead, which is technically a statistical dead heat. A month earlier, the spread was four points in Obama’s favor.

The Monday morning quarterbacks attributed the Romney loss to any number of things — lack of enthusiasm on the Republican side, loss of the Hispanic vote in key states and Romney stepping on himself with his infamous “47 percent” remarks at a private fundraiser.

Just for the sake of argument, what if the polls are wrong this time. What if on Nov. 9 Americans wake up and it is President-elect Donald J. Trump? What if it is Brexit 2.0?

People already mistrust the media. The public’s perception of the media is already in the gutter with an approval rating of 6 percent, according to a study by the Media Insight Project, which is a partnership of The Associated Press-NORC Center for Public Affairs Research and the American Press Institute.

If the polls get it wrong this presidential cycle, then it may take on an entirely different and uglier public reaction to the media. Critics will now have the ammunition to say the media is attempting to subvert U.S. democracy by suppressing the vote with underhanded methods in the name of some ideological left-of-center political agenda.

Some may even call for congressional hearings, probably raising some important First Amendment questions.

One polling operation saw this trouble brewing. The 81-year old Gallup Poll, whose 2012 final presidential poll was off by five points, giving Romney a one-point edge over Obama, announced last year it was abandoning the presidential horse-race polling business altogether.

It is clear that many in the media think Trump is going to lose in November, which they are basing on the polling. The polls will tighten, but Trump doubters, which include not just Democrats but many Republicans, are already declaring a certain outcome.

Getting a read on this election will require a more nuanced approach. The national polling may be correct, but the national popular vote is not where the presidential elections will be decided. Individual swing states will likely determine the next president, and polling in individual states has proven over time to be less accurate.

There could be a scenario that Clinton wins the national popular vote, taking more liberal states like California and Massachusetts by an enormous margin. Yet Trump could manage to eke out victories in populous winner-take-all Electoral College swing states like Ohio or Pennsylvania.

There are any number of possibilities. Keeping an open mind and not getting overly discouraged or encouraged by polling results is a safe bet. There are too many variables at play for the exact science of polling to be in stone, and in the end only one poll will count.