How I Learned to Stop Worrying and Love the Needle
The New York Times’ reviled interactive is back for 2020 — sort of
Four years ago, a New York Times election graphic earned a place in media infamy. For much of the night, “the needle” — a quivering indicator of which candidate is more likely to win, updated in real time as data pours in — showed Hillary Clinton as the favorite. And then, suddenly, it didn’t.
As the needle swung in Trump’s direction around 9 p.m., it became much more than a data visualization. For Clinton voters, it became an encapsulation of everything that was going terribly, unfathomably wrong: The fate of the country and democracy distilled to a lone meter that was suddenly pointing to “red alert.” Pundits and media theorists would later dissect its journalistic merits and failings, but its memory lives on most vividly as a jittery icon of liberal trauma.
That hasn’t stopped the Times from trotting the needle back out for subsequent elections, beginning with the 2017 special election for a Senate seat in Alabama, between Democrat Doug Jones and disgraced Republican Roy Moore. The needle has since had its share of triumphs — including that race, in which the needle began gravitating toward Jones more than an hour before the race was called in his favor — and hiccups, as in a 2018 Congressional race where the Times turned off the needle due to a lack of precinct data. These races, however, were all prologue.
Today, the needle will once again tempt the cursors of a frazzled nation amid a tense presidential election, albeit in substantially modified form. Instead of one needle, the Times will now show three: One each for the election returns in Georgia, Florida, and North Carolina.
As Times’ Nate Cohn and Josh Katz explain, the prevalence of voting by mail in this election will muddy the election-night picture, especially in states that don’t break out vote counts by method. The three swing states that will get needles, the reporters noted, do provide that data, and are expected to have their ballots counted relatively early.