Project description

“What if it’s NOT a map?”

With this question we kicked off our first brainstorm on what would become our 2018 Election Night results app, which provided real-time U.S. House election results using the AP API.

While the common approach of using a map to show election results may still be the best for choice for some outlets like the New York Times or Washington Post, we wanted something that would differentiate us. Unlike those larger outlets that are the prime source for millions of viewers for most general elections, we wanted to capitalize on our niche, given our central focus on Congress and a long traditions of deep coverage on House races.

The result of that process was a visual table depicting both results and live analysis as the night unfolded. Day-after election stories often look at where they key gains or losses were: suburbs vs. rural, gender, race, and comparison to past results. By baking that data into our design ahead of time, we put the analytical capabilities into the users’ hands as the results started coming in.

Users start with a set of dots, divided horizontally by party incumbent and vertically by race ratings, as determined by Inside Politics. The dots became colored by winning party, first a light shade as soon as a party had the upper hand and then a darker shade when races were called. Safe Democrat seats predictably were all one by a Democrat, but on the ‘safe’ Republican side, there was one glaring outlier that jumped out when Kendra Horn, a Democrat beat the incumbent, Rep. Steve Russell.

With a click on the sort buttons, users can switch the basis for sorting, showing how the results break down by urban vs. rural districts or compared to Donald Trumps 2016 presidential election performance.

Filter buttons narrow the focus to districts that changed from one party’s control to the other and where woman won.

What makes this project innovative?

Our apporach allowed users to analyze the election returns and they unfolded. Both the live anaysis and letting viewers craft the narrative to find the characteristics that made the difference in the 2018 election.

What was the impact of your project? How did you measure it?

Despite being the most ambitious and complicated app we’ve done for an election, the elction night performance was nearly flawless. The effectiveness of our approach was brought home to us in the days following the election, as we returned again and again to our own work in our stories and further analysis.

Source and methodology

We complied data in advance from AP API, Census and CQ, fact checking each aspect and relying on the AP API preview tests to make sure our app would interpret the real results correctly.

Technologies Used

Our data backend relied on Python and Flask to retrieve and compile data from the AP API, Census and CQ and prepare json enpoints for the front end pages. The dot chart was constructed in the react javascript library. We used AWS cloudfront to serve the json files without fear of getting overhelmed by the Election Night surge in viewers.

Project members

Aditya Jain, Ryan Kelly, Sean McMinn, Sara Wise and Randy Leonard



Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.