My latest column for The Week is all about how even folks who knew better — like me — still missed what happened Tuesday:

Before 2009, I worked on Wall Street, where I had a front-row seat for the financial crisis. I watched as a business I’d helped build — and that we thought we had approached with real concern for investor well-being — collapsed in the face of the financial equivalent of a 100-year storm.

It was what former risk analyst Nassim Taleb famously termed a black swan event, a highly improbable disaster that revealed everyone’s faulty assumptions. But I can assure you: Everybody I worked with knew, on some level, that such a storm was possible, and more likely than anybody acknowledged. We knew the ways in which credit quality was deteriorating. We knew that the ratings agencies allowed themselves to be arbitraged. We knew that the financial entities that insured the bulk of various banks’ portfolios were thinly capitalized, and that their assets were highly correlated with one another. We’d joke about extreme risks out on the tail of the distribution, risks that couldn’t really be quantified but that didn’t correspond to anything we’d actually observed, saying, “well, if that happens, we’re all dead anyway.”

We knew, but we didn’t want to know. And so we did what we knew how to do — as well and as conscientiously as we knew how to do — and battened down the hatches when we saw the storm brewing. And then watched our business get swamped when the storm hit anyway.

The same is true of the possibility of Donald Trump becoming president. For the political class, the possibility was inadmissible because it meant that all their knowledge was worthless: Anything could happen. For the journalistic class, the possibility was inadmissible because it would mean that their efforts to inform and influence were worthless: They were less trusted than Donald Trump of all people.

Some knew even less than that, but their ignorance was also deliberately chosen. The hedge fund managers in the film The Big Short made a killing betting on the collapse of the mortgage derivative market. How did they decide to place that bet? They read the offering documents. And they went and visited some of the properties that were being mortgaged, and talked to the owners and the lenders. That was all it took. With just a little bit of research, they learned what reams of historical data couldn’t tell them — that the market was built on sand.

How many of the pollsters and aggregators and political journalists attempted to measure, in advance, the likely voting propensity of the people who put Donald Trump over the top? Plenty of articles referenced the potential importance of non-college-educated white voters in the Midwest. Who seriously tried to answer the question of whether the various polls’ assumptions about that propensity were right?

All of that ignorance, meanwhile, fed the growth of the very risk that ultimately undid the system as a whole. That’s the difference between a black swan in zoology and a black swan in finance. Literal black swans exist or don’t regardless of whether we look for them. But if you undervalue the risk in the tail of the distribution, you create an incentive to pile up risk there, which drives the probability of that extreme event up and up. And If you don’t try to value it at all, then you are surely undervaluing it. And if you don’t collect the information that might have told you that the risk out there was increasing, then you’d never know to value it. Similarly, if you don’t ever try to turn qualitative pieces about potential Trump voters in western Pennsylvania into quantitative analysis, how will you know the likelihood that the polls will be wrong?

And what about people who just knew in their gut that something was up? How did they fare? Well, I was one of them.

I’ve spent some time looking back over my commentary on this election cycle. I started with my August 2015 column, “Why not Donald Trump?” that first explored why Trump was different from past flash-in-the-pan outsider GOP contenders. After Trump’s primary victory, I explained how the GOP would adapt itself to Trump’s leadership by adapting him to their policy priorities. Clearly, I knew GOP voters would mostly come home.

Turning my attention to Hillary Clinton, I wrote a series of columns on how she needed to redefine herself for the general election. More pointed was my warning to Clinton of the risks in focusing on adding unhappy Republicans to her coalition, and that she urgently needed to pitch more of her message at Trump’s key constituency of non-college-educated whites if only so she could understand how they were receiving Trump’s pitch.

Reading my own stuff, it’s clear I knew something like what happened could very plausibly happen, even if I wouldn’t have said it was more likely than not.

I still missed it. I didn’t want to believe what on some level I knew.

Speaking of things I knew: a lot of women friends of mine are especially anguished that the revelations about Donald Trump’s appalling behavior towards women didn’t flat-out disqualify him in the minds of voters. Unfortunately, I knew that would be the case as well. From one of my “advice to Hillary” columns back in May:

Let me make a suggestion. Have Huma put up a picture of Marcia Clark on the inside of the door to your Brooklyn office, to serve as a constant reminder of how to lose a sure thing by misreading your audience. Clark, as you no doubt recall, was the lead prosecutor in the O.J. Simpson murder trial. She thought she had a slam-dunk case and a jury eager to hear it, having stacked it with women who she figured would sympathize with the victim. She failed to account for the possibility that, as African-American women, they might have split sympathies — and that the more she painted Simpson as a cold-blooded killer, and the more she harped on the innocence of his white ex-wife, the more she was pushing their sympathies in the wrong direction, toward standing up for one of their men against a white woman’s defamation.

The 2016 election could present you with a similar problem — even without the explicit racial polarities. Say you focus your energy on attacking Trump and his supporters for being misogynists. You’ll have plenty of fuel for such an attack — but how will the women whose husbands are interested in Trump react? Are they going to let you get between them and their husbands? Or are they going to rally to their defense, and against this insulting, elitist outsider?

To get inside that defense, you can’t rely on female solidarity, or on women’s issues. Any voter for whom that kind of pitch has a strong appeal is already actively supporting you in the primary, and will certainly be with you in the general election. The women you need to reach are precisely those who are less-amenable to this kind of appeal. They are women who would consider voting Republican — who may have voted Republican in previous elections, whatever reservations or frustrations they might have had with that party. . . .

[I]t isn’t about the issues, or about experience. None of that matters if people believe that Trump is a straight-talking independent man who will put America first, while you are a cosmopolitan insider eager to do the bidding of special interests so as to win and retain power. You need to turn that around, and get people to believe that you’re a flawed human being who went into the business of politics in order to accomplish something, while your opponent is a fraud and a charlatan who has accomplished almost none of what he claims, and will do nothing of what he promises.

To make that case, you need to make an emotional connection, which means a personal one. A revelation of common experience that enables them to trust your judgment. That’s what the reintroduction is all about.

That reintroduction never happened. Instead, her campaign did exactly what I had warned wouldn’t work. And somehow, knowing it wouldn’t, I still convinced myself it had.