There’s no real joy in being alive and connected to the internet right now, but projecting the rage, fear, and confusion of these uncertain times onto a bespectacled fox named Fivey Fox does feel pretty good. Fivey is the mascot for FiveThirtyEight’s election coverage and, besides possessing a bit too much excitement for something that’s meant to help people understand the fate of American democracy, he’s part of a strategy to make the most stressful presidential election in our lifetimes a bit easier to understand.
Four years ago, when Trump stunned the world and won the presidency, the election forecasting models that had become a go-to resource for understanding the blizzard of polling data were widely perceived to have missed. Aside from the 28.6% chance of winning that FiveThirtyEight gave Trump, other outlets put his odds at less than one-in-five; The New York Times’s Upshot gave him a 15% chance, while The Huffington Post put his odds at 1.7%. In the aftermath of the election, models became the target of confusion and anger as The New York Times, Politico, The Conversation, and other publications rushed to try and answer the Big Question: “How did everyone get it so wrong?”
This year, FiveThirtyEight visual journalist Jasmine Mithani and the rest of the 20-person team who built the publication’s 2020 Election Forecast aren’t taking any chances. Despite Mithani and many others who worked on the forecast joining FiveThirtyEight after 2016, the new model incorporates some lessons from that year: instead of the addictive, vaguely heartbeat monitor-esque line chart of 2016, this year’s main graphic depicts election outcomes with a “ball swarm” that represents 100 possible electoral outcomes from their model.
Now, as we struggle not to doomscroll our way through Election Day, I caught up with Mithani to talk about her work on the Election Forecast and the origins of Fivey Fox.
This interview has been edited for length and clarity.
What did you do before you got involved with FiveThirtyEight?
I’ve worked at a lot of different kinds of jobs, but the main thing that I’ve been interested in is explaining complex systems and making sense of the world around us, whether that’s societal forces, politics, or science. I’ve worked as a video game designer making games to help people learn relationship skills. Most of my journalism work comes from volunteering at a local alt weekly paper, [South Side Weekly]. I’m on the board there now.
Directly before [FiveThirtyEight], I was a software consultant at a civic tech consultancy here in Chicago. It wasn’t so much doing the educating ourselves, but enabling other people to tell stories. I really wanted to move from the backend, making tools for other people, into telling stories. That’s when I applied for FiveThirtyEight. There aren’t a whole lot of opportunities in newsrooms that have the kind of resources to do visualization-intensive work that are open to people who don’t live in New York or Washington, DC.
Can you walk me through your role on the Election Forecast?
The main things I did were design and development on the 2020 forecast. A lot of our design decisions are backed by research on how people perceive numbers; we also spent a lot of time making sure that anyone can use the forecast. Stakeholders were involved at the beginning for the initial concepts, [including] the decision to have the forecast as one page as people scroll. It’s a story. You have a guide: Fivey Fox.
There was a design phase for the project early in the year, deciding what kind of charts and data we wanted to show people, and also thinking about what kind of audience we wanted. During election years, we get a lot of traffic and a lot of readers who don’t normally check our site. We want to satisfy our core readership, but also make sure that anybody can open the page and understand what’s happening in politics in the country.
The design we ultimately went with was having a character on the side, because there was information that we wanted to add that would have to be shoehorned into the page otherwise. Our art director Emily Scherer found the artist, Joey Ellis. He actually proposed a ton of mascots, but Fivey Fox won.
What were some of the ones that didn’t make the cut?
One was like a little caricature of Nate Silver. There was one that was an animated chad, like a hanging chad.
Is there a divide on staff between people who like and don’t like Fivey?
Oh, I think everyone likes it universally. Or if somebody hates it, they haven’t dared to tell us. [Ed. note: On a recent episode of FiveThirtyEight’s podcast, senior writer Clare Malone announced that she does not like Fivey Fox.] On our Slack, we have a Fivey memes channel where we keep track of all the ways people are remaking Fivey. It has been very polarizing — we’ve gotten a lot of user feedback on Fivey.
What are some of the biggest complaints that people have had about Fivey?
I think the biggest complaint about Fivey is that some people find it infantilizing, but that wasn’t our intention. I recently saw an email that was shared from a reader who said they used the forecast to teach their daughter about politics and [to explain] what the electoral college is. That’s something that we really leaned into this year, being a teaching tool. I joke that a lot of people have concentrated negative energy on Fivey, which means they’re not being negative to us. They have a new effigy to burn in Fivey Fox.
With the new interactive swing state map where you can pick the winner, it almost feels like sports betting. I was wondering if you’ve gotten any comments about forecasts gamifying the election.
That’s one thing that came down to the design of the project. We wanted people to know it’s interactive, but we wanted it to be [about] exploring and learning, not making light of it. That is something we are very conscious of all the time.
I’ve found myself using it late at night. Instead of “doomscrolling,” I’m doom forecasting. Have you heard about anyone using it to cheer themselves up a bit?
It’s certainly a way to see how people can get an outcome they want. It’s interesting because the forecast eases anxiety for a lot of people; that’s sort of where the value of it is. Our political process is pretty opaque — it’s confusing and so much of it varies by region. The forecast tries to break that down and explain it, and that is very useful. On the other hand, there are people who are refreshing, like, 40 times a day.
One thing I really care about is audience interaction. We are a really small newsroom [of around 40-50 full-time staff members], so we don’t have a whole lot of bandwidth to interact with users, but I try to keep abreast of [places] like Reddit. I definitely see people use it to fuel their anxiety, but also, it is a soothing tool. In a process that they can’t control, at least they can understand it.
Is the staff happy with how the forecasting model is being used this year?
I don’t think I could speak to that. But I can say that, like most newsrooms, we have prepared for every scenario. We have a publication plan for every instance we could think of [in terms of] election outcomes. We are all aware that unlikely events do happen, and we’re prepared to serve our readers with analysis, no matter what the outcome of the election is.