Times Insider explains who we are and what we do and delivers behind-the-scenes insights into how our journalism comes together.
There are some drawbacks to being a Minnesota Vikings fan, said Kevin Quealy, editor of The Upshot, The New York Times’s data-driven site about politics, economics and everyday life. His team has never won the Super Bowl, and just trying to calculate their playoff odds can become a headache. Even sportscasters appear to be run ragged by the probabilities.
“They always say these weird things like, ‘the Vikings will make the playoffs if,’ and then a string of convoluted things like, ‘if the Bears lose or the Packers win and the Titans tie,’” Mr. Quealy said.
From that frustration, The Upshot’s N.F.L. Playoff Simulator was born. Mr. Quealy wanted to create a tool that would help him, and by extension The Times’s readership, put a number on their hopeful thinking. Or at the very least, a tool that would let him have fun understanding the Vikings’ chances.
In 2014, Mr. Quealy, then a graphics editor at The Times, brought the idea to his colleagues, including Mike Bostock, a programming specialist who led data visualization projects. Mr. Bostock knew little about sports, but he was excited by the prospect of using computation to help football fans forecast the future of their favorite team.
“You have this arcane set of rules determining which teams are eligible, and from this results a complex system of behavior,” Mr. Bostock said.
The first N.F.L. playoff simulator was published online that fall. Building the tool required adding the N.F.L.’s elaborate tiebreaking procedures into the code that powered the simulator, as well as the team schedules and game scores.
“I’m a computer scientist by training, and in the same way statisticians are drawn to analyzing sports because they are readily quantified, I liked how amenable these rules were to simulation in code,” said Mr. Bostock, who left The Times in 2015 and now works at Observable, a web platform he co-founded that makes and shares data visualizations.
With the N.F.L. data, the tool was able to simulate a possible outcome for the league by repeatedly flipping a virtual coin to choose what team wins each game — and sometimes allowing ties, Mr. Bostock explained. By repeating that randomized process thousands or millions of times for every game, the tool could estimate playoff probabilities for every team.
The original simulator did not factor in team strength, however, so every outcome was equally likely, though N.F.L. teams are rarely perfectly evenly matched. Since 2015, the simulator has used Sagarin ratings, a method for measuring the strength of sports teams, when generating game outcomes. This season, it is using its own system partly based on Elo ratings, a different evaluation set.
“It’s a way of finding an objective or empirical measure for stating ‘Team A is stronger than Team B, therefore both teams do not have a 50 percent chance of winning,’” Mr. Quealy said.
For the 10th year in a row, the N.F.L. simulator is letting readers calculate a team’s chances of making the playoffs. Only 14 out of the league’s 32 teams will do so.
Additionally, the simulator can estimate the odds of a team clinching a first-round bye, earning a wild-card berth or winning its division. And, if a team is performing poorly, it will show what outcomes need to happen for that team to finish the season in last place — and get the No. 1 overall pick in the draft.
This year, Las Vegas betting odds are being used in the code for the first time, which helps the simulator better account for differences in team strength, which can change quickly.
But the N.F.L. simulator isn’t meant to be a prediction tool; Vegas oddsmakers are better suited to that task. Instead, Mr. Quealy said, “It’s something to help you think about your team’s playoff chances, conditional on some outcomes that you chose.”
The simulator, which closes at the end of the regular season, and the code that powers it, have changed very little over the years. The same can be said of its fan base.
“It’s one of the few things where we’ll get emails from readers wondering, How come this isn’t published yet?” Mr. Quealy said. “It really has stood the test of time.”
And there might be hope for Mr. Quealy and the Vikings yet. After defeating the Atlanta Falcons on Sunday to improve to five wins and four losses, Minnesota has a 61 percent chance of making the playoffs, most likely with a wild-card berth.