Current polls suggest that Vice President Biden will win the presidential election walking away. Perhaps that explains the chorus of Democrats and their media myrmidons denouncing President Trump’s refusal to leave the White House as if it were already January. Or perhaps those denunciations are about driving down a Trump vote they are more afraid of than they let on.

In any event, it’s not out of the question that the polls that say Biden will win really mean he will lose. They focus on the wrong kind of vote and the wrong kind of voter. They divide the country by political affiliation incorrectly. And they fail to account for Trump voters who keep their preferences to themselves, as well as the number of Trump voters willing to crawl over broken glass to the voting booth.

Many polls estimate the popular vote, which is constitutionally irrelevant. Mr. Biden will walk away with New York and California, but millions of those votes will have no effect whatsoever on the presidential election. What counts is not Mr. Biden’s 7.4% lead in Real Clear Politics’ average of recent polls, but his lead in the “battleground” states, whose electors could go to either party.

Real Clear Politics awards Mr. Biden an average lead of 5.5% in the top battleground states. That’s still solid. Roughly half of the polls that feed into that average, though, reckon not “likely” voters but “registered” voters. Likely-voter polls are more reliable. When one looks at likely-voter polls in the many battleground states in which Mr. Biden leads, that lead drops to between 1.5% to 5%, with one outlier at 6.4%, from between to 2.3% to 8%.

These numbers are enough to win the presidency, but wait — there’s more. First, for reasons that remain obscure to me beyond the sad possibility that many pollsters are Democrats who have persuaded themselves of models that favor Democrats, many polls simply do not represent the political dispositions of American voters.

Just a few days ago, the Washington Times noted that the RCP average includes two polls that assume about 24% of voters are Republicans. They are not the only offenders in a country where, in 2016, about 31% of the voters were Republicans. It is, moreover, a stable number.

That difference alone could, in theory, turn the election. Now suppose these pollsters know something we don’t, and we leave these numbers unadjusted. We must still account, secondly, for the “shy” Trump voters that liberal outlets insist do not exist. It isn’t merely common sense that advises caution. A recent Monmouth University poll suggests that the same voters who award Biden a double-digit lead in Pennsylvania actually think it marginally more likely that Trump will win.

Why? Because, it turns out, nearly 60% of voters think there are secret voters — quite possibly lots of them — who support Trump but won’t tell anyone. Yet few think there are secret Biden voters and almost no one thinks there are lots of them.

Finally, the “enthusiasm gap.” According to Economist/YouGov polling as of mid-June, Mr. Trump’s supporters were twice as enthusiastic as Mr. Biden’s. I would have said 68% enthusiasm for Mr. Trump is comfortably more than twice the 31% enthusiasm for Mr. Biden.

A tendency to play down good news for Trump is to be seen in the same organization’s insistence that enthusiasm doesn’t mean votes. That’s true. If the real difference in the battleground states is largely a matter of a few percent, though, then that massive enthusiasm gap need only translate into a small voter gap for Mr. Trump to win. Anything more, and Mr. Trump would win by a landslide.

It’s not my purpose here to make a prediction. Nor should Mr. Trump’s supporters get cocky. Neither, though, is this a lead pipe cinch for the former vice president. Mr. Biden and his elusive vice-president-to-be certainly shouldn’t order the new curtains just yet.

_________

Mr. Rosenberg, a writer and editor living in New York, is an occasional contributor to the Sun.