In this election cycle, the broad consensus was that Trump was an amusing epiphenomenon with little staying power and few prospects. By contrast, since the declaration of her candidacy, there was a pervasive assumption that Hillary Clinton would coast to victory. This cycle seemed to turn conventional wisdom on its head at every turn, but nonetheless, as Americans went to cast their ballots even the most rigorous and gutsy of poll analysts predicted less than a 30 percent chance for a Clinton loss (and even this was viewed by many as being far too generous to Trump.)

To be sure, prediction is a perilous game. But one of the biggest problems in the way we rely on predictions in our public discourse is that pundits are rarely held to account for their reliability. And even in those rare instances where someone actually issues a mea culpa for grievous errors, little seems to be learned in terms of how to approach subsequent developments.

For instance, “no one” saw it as possible that Trump would win the Republican nomination—even down to the final moments. When he did win, there was a lot of hand-wringing about humility and lessons learned—but then almost immediately the same narrative emerged again: Trump stood no chance at winning the general election. And not just he, the Republican Party was finished, perhaps for generations, as a result of his candidacy. Things look very different today, with the Democrats standing on the brink of irrelevance, while the signature accomplishments of their most charismatic and transformational leader in generations seem set to be erased.

The centerpiece of the Democrats’ (over)confidence was their supposed “electoral firewall”—a safeguard they were so sure of that they scarcely even bothered to reach out to these constituents who were supposed to serve as their last line of defense against Trump (ensuring that even in the unlikely event that they lost the popular vote, they would win the Electoral College.) How’d that turn out? Clinton is shaping up to lose the Electoral College by a larger margin than those who were defeated in 1996, 2000 or 2004.

“How could this happen?” was the refrain on November 9. Here’s how:

Yes, Clinton had a ridiculous advantage in fundraising and organization. She had a consistent lead in the polls. In a normal cycle, all of this would seem to presage a victory. But we were not in a normal cycle, and this should have been obvious by how the primaries played out: on the Republican side, despite having no campaign organization to speak of, no government experience, painful ignorance about most political issues, and no sense of basic decency or decorum, Trump overcame what many viewed as the fullest and strongest bench of GOP candidates in recent history—to include energetic young voices, women and minorities, well-liked governors, etc. Across the aisle, Clinton found herself unable to secure the nomination despite her unprecedented qualifications, her well-established relationships and “ground-game,” her overt support from the party establishment, and the reality that her chief rival was a Jewish socialist with no money, operation or even name recognition to speak of at the outset of the campaign. Red flags, red flags, red flags.

By recognizing we were not in a “normal” cycle, analysts could have realized that signals which are reliable in other elections might be misleading here. This should have significantly undermined their confidence in conventional metrics, and spurred them to seek out indicators which might track the idiosyncratic dynamics of this particular cycle more reliably.

Oblivious to the Obvious

Eight months prior to the election I was able to not only predict its outcome, but identify most of the specific states that Trump would win. Six months prior to the election, I was able to predict—running sharply against the popular grain—that Trump would exceed Mitt Romney’s share among Hispanics, blacks and Asians, and that his supposed “white people problem” was overstated. On and on (and on…and on) in the months that followed.

With each warning, I was met with incredulity and often scorn by arrogant, condescending Clintonites. But what I feel now, in the wake of Trump’s election, is not schadenfreude—but righteous anger. Because this outcome was both totally foreseeable and avoidable. The data I relied upon was 100 percent available to the public. The conclusions I drew were not conjured through esoteric feats or great leaps of faith: they were painfully obvious if one was willing to entertain—even for a minute—the humble idea that the logic of typical elections may not fully hold in this instance.

Most Clintonites outright refused to even consider this possibility. They were confident in her structural advantage, and the polls were telling them what they wanted to hear, so chose not to dig deeper (and even when as the polls increasingly confounded their expectations, they simply chose to write them off).

But here’s the thing about polls: in order to be predictive, they cannot be taken as sufficient indicators unto themselves. They must be qualified by the weaknesses of the method (and the reliability of the surveyors); the findings must be compared against analogous and contemporaneous surveys from others; analysts must consider longitudinal trends in addition to time slices; they must cross-check how the public views issues not just candidates. And most importantly—all of this must be interpreted in terms of broader ongoing socio-cultural trends.

What I could see in this landscape is that a plurality of the public wanted radical change, even to the point of nihilism. They hated the establishment, including their own parties and leaders. They had no confidence that the kinds of solutions which have been peddled to them in election after election would ever be implemented—or would sufficiently improve their lives if they were. They found the status quo to be insufferable. They had no appetite for more war abroad, for new trade deals, or other forms of expanded globalization. They were fearful of threats from without and within. This trend has been occurring throughout Western democracies, and the result has been consistent: the elevation of autocrats and demagogues.

In this context, it was ludicrous and arrogant for the DNC to put forward a corrupt, hawkish, neoliberal technocrat as their candidate of choice. It was suicidal for the party to actually put their thumb on the scales for Clinton (via the super-delegates) prior to most votes being cast: while this maneuver convinced many that no one else could plausibly win the nomination, it demoralized a large contingent in the process. And to make matters worse, when these measures proved insufficient to quell Sanders’ revolt, they began to outright cheat in the service of Clinton’s candidacy, while working on ways to push her opponent out of the race. It boggles the mind that the DNC so zealously rallied behind a candidate whom a plurality of the public—across the political spectrum—distrusted and even despised (with lots of baggage and new scandals on the horizon to boot.) It’s stunning that Clinton couldn’t see that her campaign would be ill-fated and have the good grace to not run (or to withdraw), as some of her advisors recommended.

But let’s set aside Clinton’s nomination to focus on her messaging: In a cycle where many Americans would have been just as content to burn down Washington D.C. and the Financial District of Manhattan, Clinton warned that Trump is a threat to prevailing civic norms and institutions. She had the bad judgment to make this the centerpiece of her campaign—over and above any policy platform—apparently oblivious to the reality that this kind of narrative would make Trump seem more appealing to many (even as it reinforced Clinton’s image as a defender of the status quo.) Worse, this ill-constructed narrative was often delivered (by surrogates, and at times, the candidate herself) with elitist condescension: those disinclined to vote for Clinton were derided as either ignorant, stupid, racist or misogynistic. How this would actually endear Clinton to a skeptical public is difficult to comprehend.

And this is just her outreach to Republicans and Independents! As for the rest, it was just taken for granted that blacks, Latinos and Bernie voters would turn out in large numbers, and in support of her candidacy, strictly in virtue of her opponent (i.e. requiring little substantive concessions or advocacy on her part for their priorities.)

How anyone thought this was a winning formula, I cannot understand. This entire election was like watching a slow-moving car crash, but being unable to convince the driver or one’s fellow travelers that a collision was imminent. Now here we are, among the ruins. Hopefully, people at least learned something from all this. But I’m not holding my breath. Kanye 2020.

Musa al-Gharbi is a Paul F. Lazarsfeld Fellow in Sociology at Columbia University.  Readers can connect to his research and social media via his website, fiatsophia.org.