Skip to content

The Patron Saint of Superheroes

Chris Gavaler Explores the Multiverse of Comics, Pop Culture, and Politics

First a reminder: the 2016 polls predicted Clinton would beat Trump by 3%. She instead beat him by only 2%–in the popular vote, which was all anyone cared about back then. Now we know better. It’s the Electoral College, stupid.

That’s good news for Trump, because his chances of winning the popular vote are essentially zero. He’s the first president in polling history never to have a positive approval rating. He’s hovered around 42% for most of his term.

But that’s where Trump’s good news stops.

Of the twelve presidential election forecasters, nine predict Biden will win the Electoral College. None predict Trump. Three are undecided but heavily favor Biden. None favor Trump. Trump’s electoral count ranges from 164 to 204. Biden’s ranges from 268 to 325. You need 270 to win. Here’s the list, starting with best for Biden and ending with best for Trump:


325 Biden,          187 Trump


319 Biden,          163 Trump

Inside Elections:

319 Biden,         187 Trump


318 Biden,          125 Trump

The Economist:

308 Biden,          164 Trump

Cook Political Report:

308 Biden,          187 Trump


307 Biden,          185 Trump


297 Biden,          170 Trump

U.S. News:

278 Biden,          186 Trump


268 Biden,          170 Trump


268 Biden,          203 Trump

Crystal Ball:

268 Biden,          204 Trump

Trump is at best 66 votes short, and Biden is at worst two short. All twelve forecasts place some states in the toss-up range, leaving their electors unknown. The number of toss-ups in each forecast ranges from two to six, with a total of nine different states. Here’s the list, from most to least commonly identified as toss-ups:

North Carolina 12, Arizona 8, Georgia 7, Ohio 7, Florida 4, Wisconsin 4, Iowa 3, Texas 1, Minnesota 1.

North Carolina is unanimously in play, and it’s probably equally safe to ignore outliers Texas and Minnesota, leaving a total of seven serious toss-ups. (Though if you define a toss-up as any state where neither candidate is currently leading by more than five points, you would instead have to leave in Texas and Minnesota, add Nevada and Arkansas, and strike Florida.)

This is not unlike last November when, according to the four earliest predictions, there were eight possible toss-ups: Arizona, Wisconsin, Florida, Pennsylvania, North Carolina, Michigan, New Hampshire, and Iowa. But note which two states are no longer in any of the forecasts’ toss-up categories:

Pennsylvania and Michigan.

Those were two of Trump’s key victories in 2016. Since Trump and Clinton were polling within the margin error, both states were toss-ups then. Now neither are. All twelve forecasts predict Biden will win both.

Worse for Trump, over half of the current forecasts have Ohio and Georgia in play, two states Trump won easily in 2016. Still worse, for Trump to win in 2020 he has to sweep all seven toss-ups. Biden has to win any one of them, giving him seven different paths to victory.

But of the seven, one stands out:


It was the one actual upset in 2016. Clinton was polling just above the margin of error. Instead, Trump took 1,405,284 votes to Clinton’s 1,382,536. That’s a difference of 22,748 from just under 2.8 million total.

And Wisconsin looks just as key in 2020. The three forecasts that place Biden at 268 electors identify Wisconsin as a toss-up. Eight of the nine forecasts that place Biden over 270 predict he will win Wisconsin. There’s only one exception. NPR identifies Wisconsin as a toss-up, but they also predict Biden will win Florida, still putting him over the top. Seven other forecasts give Florida to Biden too, as do the current polls. None give Florida to Trump.

What does this all mean for November? Nothing. These are just predictions. And given the unknowns of voting during the pandemic, predictions may be significantly less predictive this year. Fortunately, Virginians have some excellent voting options:

1) you can wait till November 3rd and cast your vote as you’ve probably always done,

2) you can order an absentee ballot and vote by mail, or

3) starting 45 days before election day on September 18th, you can vote in-person at your town hall or municipal building during regular business hours, plus some Saturdays.

So if you want to prove the predictions right, go vote. If you want to prove the predictions wrong, go vote too. That’s one thing all Americans can agree on: it’s the vote that matters.


%d bloggers like this: