Putting away groceries, I see that the "best by" date on my daughter's yogurt is farther off than November 3, Election Day. It's coming! Who's gonna win? I grabbed the above pictogram from the FiveThirtyEight blog, where Nate Silver &c construct a model made of aggregated polls and other data, then run it thousands of times to see what happens. Biden's chance of winning, according to their model, now stands at 85 percent—that's how much of the area below the graph is on his side of 270 electoral votes. When I first started checking back in midsummer, FiveThirtyEight pegged the probability of a Biden victory at about 70 percent, so the chance of Trump winning has been halved over the past three months. Turns out that hosting a coronavirus superspreader event on the White House grounds isn't just a public health debacle—it's political malpractice, too. If FiveThirtyEight is right, the chance you have to endure four more years of this shitshow is slightly less than the chance of losing a game of Russian Roulette.

Typing the word "shitshow" in this context reminds me that I recently overheard my daughter telling a friend, "My mom swears a lot more than my dad—in fact, I've only heard my dad swear when the Twins are losing or Trump is talking on the news."

Going to pull out a little of my statistics vocab and observe that, in the FiveThirtyEight model, there is quite a distance between the median result and the mode. The former would refer to the point where a vertical line divides the area below the graph into two regions of equal area. Just eyeballing it, this appears to be at a point where Biden wins about 330 electoral votes: if I'm gauging that correctly, it would mean he has a roughly fifty-fifty chance of winning more than 330 electoral votes. Viewed from the other direction, Trump has a roughly fifty-fifty chance of winning more than 210 electoral votes (and a 15 percent chance of winning at least 270).

But the single most likely outcome—the highest point on the graph—occurs where Biden wins around 425 electoral votes, which would be an electoral beheading: it would leave Trump with fewer than 120 electoral votes and would almost certainly mean the Democrats also pick up more than enough seats to win a Senate majority. Why is there such a distance between the "average" result and the single most likely one? I think it must be that Biden is now a clear favorite in most of the states that everyone thought would be the "main battleground": he's ahead in Wisconsin, Michigan, Pennsylvania, Arizona, and, more precariously, in Florida and North Carolina. Winning these states, or maybe all but one of them, is what puts him in the 330 electoral vote area. The states in which the race is closest—where the model indicates no clear favorite—are those representing the next line of Republican defense, ones they didn't think they'd have to defend: Iowa, Ohio, Georgia, Texas. Except for Iowa, these are huge states, and, according to FiveThirtyEight anyway, they are on the proverbial knife's edge. The likelihood that they move together, one way or the other, is what accounts for the trough between the mean result and the single most likely one.

Lots of people are concerned about the prospect of Trump claiming victory on election night and then raising objections to the methodical counting of mailed ballots. In many states, the processing of mailed ballots cannot by law begin till Election Day—I mean, the envelopes cannot be opened, the ballots smoothed and inspected and prepped for the machine, until November 3. If Trump led on Election Night in several key states and then moved to end the tabulating of mailed ballots, the country could be plunged into a crisis. Luckily, the concern might be exaggerated. It's not that Trump isn't that bad a guy—he is, and has clearly signaled that he plans to challenge the result if mailed ballots put Biden over the top in the days following November 3. Florida and Texas, however, are two states that allow mailed ballots to be processed and counted as they arrive. This means that by around 9 o'clock on November 3 we will likely know what happened in these states. If Biden wins either one, Trump has no plausible path to victory, and it's hard to see how any extra-curricular shenanigans could obscure that fact. If Biden wins both of them, the route is on.

I guess what I'm saying is that

(a) Trump might not be close enough to execute his plan; and

(b) vote.

## Comments