Ads Top

My Election

Four years ago I took a few minutes every day to check and reassure myself that the worst was not going to happen in November. I looked at the latest poll numbers. I read FiveThirtyEight. The map looked good. The numbers looked good. I felt confident as Election Day grew closer that all would be well.

And then the worst happened.

I vowed I would never again expend that kind of energy and give that much attention to polls and predictions about election outcomes. Polls, after all, are not news. Polls aren't truth. Polls aren't facts. Polls aren't even predictive of anything. They are just someone's snapshot of something that doesn't matter. "The only poll that counts," the saying goes, "is the one on election day."

I remember Nate Cohn and Nate Silver and others doing some Monday-morning quarterbacking, insisting that the polls weren't actually wrong. Even as recently as three days ago, Grace Panetta of Business Insider wrote
Since the 2016 election, there's been a consistent but a misleading narrative that the polls were disastrously wrong. In reality, the polls were a lot more accurate than some commentators make them out to be.
And this past April, Scott Rassmussen wrote, "The problem was not the polls, but the analysis." He goes on to explain that only in Wisconsin did the polls truly miss. In Pennsylvania and Michigan, the polls were well within the margin of error.

I was one of the fools who bought into the analysis. So when I woke up the next morning, I canceled my subscription to the New York Times. I deleted my bookmarks for and for FiveThirtyEight. I decided I was done listening to pollsters and pundits. I remember thinking at the time that this was a bigger calamity for the United States than 9/11. And certainly the last three-and-a-half years have borne that out.

But now I find myself once again reassuring myself with this map from (which I check every day...):

I've resubscribed to the New York Times. I'm reading analysts and pundits who are reassuring me that this time the poll numbers look very different from 2016. This time, the "strongly Dem" plus "likely Dem" numbers total 302 electoral votes, well over the 270 needed for a Biden victory. (On this date four years ago, those numbers totaled just 235 for Hillary.) There's little chance, the pundits say, that these numbers could turn around by November.

I'm reading and reassuring myself. But I don't want to let myself get too confident this time. I wasn't prepared for the worst in 2016. I don't want to fall into that trap again. 

1 comment:

Powered by Blogger.