On election night, knowing who is ahead isn’t enough.
In states like Virginia, Georgia and elsewhere, larger cities have traditionally been among the last to report their votes. Because these cities lean Democratic, this pattern means that returns for Democratic candidates will appear weaker than they actually are for a few hours after the polls close. Similarly, in states with slow-counting Republican areas, Republican candidates will seem further behind than they are.
The worst election analysts use these changes to tell comeback stories, asserting that a candidate was trailing early in the night but made up ground.
More sophisticated analysts interpret leads through the lens of outstanding votes. “There’s a lot of votes left to be counted in heavily Democratic Cuyahoga County,” Jeff Greenfield said on CNN in 2004. “Remember, some of the votes outstanding are down here in Marion County where Obama is winning,” John King said on the same network in 2008.
Like many sites, we tracked the leads reported by The Associated Press.
But a team at The Times’s Upshot site wanted to apply more rigor to the anecdotes about the areas with outstanding votes. So they created a statistical model that adjusted those reported leads based on from where the votes were coming to give readers an estimate of who was really ahead — and by how much — in races’ early returns.
The result was a sophisticated way to watch the election unfold. Instead of guessing how races would go based on incomplete numbers, our readers had a much clearer picture.
Take Virginia, for example. At just before 8 p.m. and with less than a quarter of the precincts reporting, Ed Gillespie, the Republican candidate, appeared to hold a nine percentage-point lead. But The Upshot model showed that when adjusted based on the types of areas that had been reported, the lead evaporated, putting Mark Warner, the Democratic candidate, just ahead in a neck-and-neck race. Hours later, Mr. Warner was declared the winner with less than a one percentage point margin of victory, just as the model estimated early in the night.
This was just one of the features of The Times’s election-night coverage that provided readers a comprehensive and clearly designed picture of the results through mobile phones and tablets as well as desktop computers.
We provided pages of clearly organized results. The first question many readers had was “Who is winning in my state or district?” The map at the top of the page quickly gave them the answer. But as good as the map was at answering that question, it was not meant to explain which candidate was doing better than expected.
For that, readers could turn to the Big Board below the map, which organized the results by the expected winner in each race. As the results came in, red squares creeping on to the left side of the table meant Republicans were doing better than expected. Blue squares creeping on to the right meant a better night for Democrats. Though just a simple table, it allowed readers to scan across dozens of races while easily focusing on the tightest.
We also published a series of maps showing precinct-level results. The data for these highly detailed maps did not come as part of an Associated Press feed. To get it, we sent scores of emails, tracking down maps that showed the boundaries of each precinct from dozens of state and local officials. Then we had to set up a way to “scrape” the numbers from many different websites.
But it was worth it. These maps showed readers what happened in individual neighborhoods, revealing a political topography that doesn’t come across at the county level.
Detailed results pages for each state went beyond the commodity tables found on many news sites. As races were called over the course of the night, a team of a dozen reporters wrote summaries of the most notable in each state, bringing additional context to a page that would otherwise be a stack of numbers.
A live blog brought together the best of The Times’s election-night content: updates from across the country from The Times’s reporters; analytic charts and tables from our graphics department; analysis of exit polls; and a clear display of the results in the most watched races.