On October 18, Dr. Sam Wang, Princeton professor of neuroscience and semi-celebrity pollster (he correctly predicted the presidential vote outcome of 49 out of 50 states in 2012), tweeted:

Wang’s prediction was in line with the numbers released by other pollsters at that time. In mid-October, for example, the New York Times estimated the probability of a Hillary Clinton presidency at between 92 and 94 percent. Nate Silver, the statistics genius behind FiveThirtyEight, put the Democratic candidate’s chances at 85 percent, a number so conservative he felt compelled to write an article explaining “Why Our Model Is More Bullish Than Others On Trump.” Other outlets with similarly lopsided predictions included the Huffington Post, the Cook Political Report, and PredictWise.

Even after James Comey’s October 28 revelation that the FBI had discovered new and possibly incriminating evidence related to Hillary Clinton’s use of a private email server during her tenure as secretary of state, few pollsters changed their tune. The Times lowered Clinton’s chances by single digits to 85 percent; on CNN’s Smerconish, Wang stood by his earlier predictions. “Honestly, this race is the most stable, statistically speaking, since Eisenhower beat Stevenson in 1952,” Wang said. “I think that things are looking the same, and we should stay calm.” (For his part, Nate Silver upped Donald Trump’s chances to 35 percent — an estimate that provoked the Huffington Post’s Ryan Grim to accuse Silver of “putting his thumb on the scales” by “changing the results of polls to fit where he thinks the polls truly are, rather than simply entering the poll numbers into his model and crunching them.” Grim’s own HuffPost Pollster doggedly kept its own estimate of Trump’s chances at a mere 2 percent.)

I believed them all. Here is an actual text exchange I had with a friend the day before Election Day. (Fittingly, I’m the blue one.)

I spent the hours between 8:30 p.m. and midnight on Election Day in a friend’s room in Pyne with about ten other people. When I arrived, Clinton was ahead in Ohio, Michigan, and North Carolina, and Texas had only just turned pink after a brief but thrilling period when Democratic-leaning San Antonio and Houston tipped my home state blue. My friends and I drank cheap wine out of red Solo cups and munched on chips and salsa as we watched CNN’s John King and his color-coded map.

Things started to go south fast. Trump pulled ahead not only in Ohio, which I had expected, but also in North Carolina, Florida, Virginia, and New Hampshire.

“If she loses Ohio, North Carolina, and Florida, she’s FUCKED,” someone moaned.

I reassured him with the knowledge I had gleaned from the websites I pore over like Scripture. “Trump has to win all of those to have a shot. If she takes just one, she’ll win. And she can win without any of them.”

Florida started to look bad. “But they barely have any votes counted from the Miami area,” I reasoned.

Trump steadily widened his advantage in North Carolina. “Just wait until they get the results in from Asheville and Charlotte,” I said.

Suddenly Wisconsin and Michigan turned pink. I shut up.

My best friend, frantically checking her phone for updates from other news outlets, announced that the New York Times was predicting a Trump victory.

That was when my disappointment at the realization that Hillary would not pull off a historic landslide started to morph into dread that she might not win at all.

After another agonizing half hour of staring at John King’s map of the upper Midwest, I was too jittery to sit still anymore. I left Pyne and walked the half mile to the grad college, down shadowy residential streets lined with beautiful homes owned by rich white people. I wondered who they had voted for.

At the grad college, I sat down on a bench overlooking the golf course. It was cold and I didn’t have a jacket. I prayed the Our Father and the Hail Mary for the first time in a long time, because what the hell, it couldn’t hurt. (While I was at it, I also prayed for my uncle, sick with leukemia, and the nice man who works in Wilcox dining hall and whose brother died over fall break.) I didn’t check my phone. When my teeth started to chatter, I went back to my friend’s room in Pyne.

The wine was gone. Nobody was talking. I lasted until midnight before I went to bed. I’ve always preferred dealing with loss in solitude.

* * * * *

At six o’clock, I was woken up by the sound of hysterical crying coming from the room next door. “I CAN’T TAKE IT!” a girl’s voice screamed. I caught the words “Trump” and “president,” but I couldn’t really understand much through the sobs. This was my version of the moment faced by millions of Americans on the morning after Election Day: the realization that there was no waking up from this nightmare.

All day, campus felt dead. Part of it was undoubtedly the weather, which was as bleak and rainy as I thought befitted the occasion. But there was none of the laughter or levity that usually permeate the place where five thousand 18- to 22-year-olds live. People pulled their hoods up and avoided eye contact as they hurried to class.

I oscillated between rage and despondency as I drifted from meetings to class to the library. I didn’t check the news or read my text messages. At one point I fantasized about running into a professor who had publicly stated his intention of leaving his ballot blank. I knew then and I know now that he was in no way responsible for the result of the election, but still, I imagined confronting him and demanding if he was happy with the way things had turned out.

In a journalism seminar, inevitable discussion of the election results was predictably dominated by my white male classmates, who mansplained the mechanism and consequences of Trump’s improbable victory while I seethed.

My sister, who voted for a third-party candidate, called me while I was in a meeting. I picked up and told her that I couldn’t talk at the moment. “I’ll call you back,” I said. I didn’t call her back.

* * * * *

In the days since the election, a lot has been written about why the polls got it so wrong. Apparently, pollsters overestimated minority voter turnout and underestimated the uncertainty inherent in predictive polling. To be honest, I care a lot less about the source of polling error than I do about the way Trump’s Cabinet is taking shape. But I did experience some schadenfreude when Dr. Wang reappeared on Smerconish a few days after the election and ate a cricket.