The caveat before we get started with this exercise is to point out Mark Craig’s very good piece last week about how the strength of schedule of NFL teams going into a season is not a good indicator of how that team will fare — or even how hard their schedule will really be.
The NFL is a league of change, with half of all playoff teams being displaced routinely from year to year. And because there are precious few games, the difference between 9-7 and 7-9 is pretty tiny. One bounce here, a missed kick there, and the fortunes of a team and their seeming outlook going into the next season can totally change. The difference, say, between 91-71 and 71-91 in baseball is not that subtle.
But we can acknowledge that there are a lot of variables from year to year and inherent flaws in trying to predict based solely on the past while also asking this: Are we doing the actual tabulation of strength of schedule incorrectly?
In an interesting 538 piece, the answer is invariably, “Yes.” Long story short, using “Pythagorean wins and losses,” which is a fancy word for what a team’s record was expected to be based on point differential, we get a more accurate picture of how good each team was and how good we should expect them to be the following year.
It makes sense, right? Not all wins and losses are created equally. A team that wins a lot of close games one year might be due for a regression, while a team that loses a lot of close ones might be due for a jump up in wins the next year. So if we look at a team’s 2018 opponents using their expected wins/losses in 2017 rather than their actual wins and losses, we get a more accurate picture of the outlook.
And here’s where the narrative changes for the Vikings. Based on just wins and losses of their 2017 opponents, the Vikings are deemed to be tied for the eighth-toughest schedule in the NFL. That sounds pretty daunting, right? But 538 ran the numbers and adjusted all of the SOS for Pythagorean winning percentage instead of actual winning percentage. Using that more accurate measure, the Vikings are tied with the Eagles for the 18th-toughest schedule — actually on the easier side of average.
The Vikings have a lot of teams on their schedule deemed to have been won/loss overachievers in 2017 based on their point differentials, including Buffalo, Arizona, Philadelphia, New England and Miami.
Before you start planning a Super Bowl route, 538 acknowledges the obvious: “There’s no model that can account for player age, coaching changes, free agency, the draft or player injuries before they happen.”
Even using this new projection system has its flaws. The Packers, for instance, were expected to win just 6.3 games last season, but they were also 4-1 before Aaron Rodgers was injured. With Rodgers presumably back and healthy, this year’s Packers are not a six-win team — but they are treated as such when considering the Vikings’ strength of schedule, appearing on it twice.
But the adjusted method is at least a better predictor than just using wins and losses. And it should make us look differently at the Vikings’ 2018 slate.