Presidential horse race polls measure attendance at baseball games.

Let me explain. I was talking to a friend and sharing what I believe the polls are really showing. They are not, by and large, showing changes in allegiance (especially in this election; while Clinton probably has some room to grow among some small groups of traditionally Republican voters, Trump most likely has a pretty firm ceiling). Also, a quick google will let you know that most experts agree. So I’m not about blow things open with a new algorithm, I simply found what I think is a pretty cool metaphor.

At this stage, polls measure enthusiasm, which translates also into the likelihood of voting. When Trump dropped in the polls after the debate, it wasn’t because Trump voters switched sides in great numbers. It was because some Trump supporters, feeling down in the dumps, either said they were undecided, because Trump made it kind of embarrassing to be his supporter, or else, when asked, they downgraded their chances of voting, causing them to be caught by likely voter filters.

I compare this to a baseball fan base.

If you’re a Dodgers fan and the Dodgers are on a losing streak, you don’t suddenly become an Red Sox fan. No, you remain a Dodgers fan. But what does happen is that you don’t attend as many games, so the stadium in Chavez Ravine becomes more empty.

When the Dodgers start doing very well, you start attending games. If they start leading their division and ESPN tells us that they are World Series favorites, then people who rarely attend games or had stopped attending games start showing up again.

This is what polls are measuring. People aren’t switching between the Red Sox and the Dodgers (or Clinton and Trump). They are simply attending more or fewer games. Clinton is up, so her stadium is filled with fans. Trump is down, so attendance at his stadium is down.

So, does that mean that these polls aren’t measuring anything useful?

Absolutely not! Because enthusiasm and hope and involvement (read: stadium attendance) are key measures of likelihood to vote.

Just as you are less likely to attend your favorite team’s game if you believe they are going to lose, so are you less likely to vote if you see your candidate doing poorly (or making nonsensical word salads during a debate). Season ticket holders will always show up, but it’s the other, more occasional fans who are the difference between 47% and 51% of the vote.