For a long time I’ve conceptualized Trump voters as falling into two camps: the enthusiastic and the reluctant. No doubt there are some who fall in between, too, but I think that mainly there are those two categories.

The first group consists of people who supported Trump in the primaries, either as first or second choice. They believe he would be a good president, or at least that he was the very best or one of the very best of the lot of GOP candidates who originally threw their hats into the ring.

The second group is composed of people who support Trump now only because he’s running against a person they consider worse, Hillary Clinton. They don’t always agree on why he would be better than she, but they agree that he probably would. Some of them even detest him otherwise and think that he would make a bad president, but are still willing to vote for him as the alternative to Hillary.

Members of the first group sometimes appeal to members of the second group, urging them to make sure they vote for Trump. Their arguments can vary widely. Sometimes the argument is that Trump will be a good president and will do a number of good things for the country. Sometimes it’s that Trump will do one or two good things, usually involving SCOTUS justice choices and/or immigration policy. And sometimes it’s that even though we don’t know what Trump might do, we know that Hillary would be awful and there’s at least a chance that Trump would be better.

That latter type of reasoning has probably become the most common argument we see these days, at least as reflected in blog comments and blog posts, in newspaper and magazine columns and TV commentary, plus social media. As Trump’s poll numbers rise and fall and then rise and fall and rise, and state polls become increasingly common, it occurs to me that the latter argument rests almost entirely on the race being at least somewhat close.

For example, let’s say a reluctant Trump voter lives in California. If Clinton is ahead by 12% in the polls there (I chose that number because it was the figure in a recent poll, but the point is that no one really disputes that she’s way ahead there), why would a reluctant Trump supporter be motivated to compromise what he/she might see as his/her principles and integrity by voting for a man he/she detests, if that man is seen as having absolutely no chance of winning in that state?

It behooves Trump to stay close in swing states, because a whole lot of his supporters everywhere are in the category of “reluctant supporters.” Often very reluctant supporters. They will only hold their noses and vote for him in that voting booth if they see a good chance of defeating her. Because elections are decided by winning state electoral votes rather than winning the national popular vote, voting decisions will be made by most voters who know much of anything about elections (and I would guess that the category “reluctant Trump voter” is probably made up of people who are aware of the way the Electoral College works) by taking into account this system.

That’s one way in which polls matter. There are margins of error in polls, to be sure, and there are just plain errors, too. But overall, polls that show Trump far behind in a state are probably going to further discourage the Trump vote in that state. And polls that show him to be close will almost certainly encourage even a reluctant Trump vote.

[Neo-neocon is a writer with degrees in law and family therapy, who blogs at neo-neocon.]