Vikings Tower over Giants, but How Much Can We Learn from Week 1?
In their first game of the 2024 NFL season, the Minnesota Vikings dominated the New York Giants on both sides of the ball en route to a 28-6 win at MetLife Stadium.
Sam Darnold looked about as good as he ever has, posting his highest single-game passer rating since 2019. Aaron Jones rushed for an average of 6.7 yards per carry and one touchdown (last season, the Vikings’ first rushing touchdown didn’t come until Week 8). The Brian Flores defense looked like its peak 2023 form, forcing two turnovers and scoring more points (7) than the Giants offense (6).
On the other hand, part of the Vikings’ Week 1 success is undoubtedly a consequence of their opponent. The Giants were expected to be cellar-dwellers this season, and they lived up to those expectations in Week 1, committing huge mistakes on both sides of the ball.
QB Daniel Jones consistently struggled to hit open receivers, and aside from a potent pass rush from the interior defensive line, the Giants’ defense was incapable of keeping up with the Vikings’ balanced attack.
With this in mind, how much stock should we put in the result of a single game?
It is sometimes remarked that Week 1 is a bit of a “beta test” for NFL teams. It is the first time anyone has played meaningful football in more than six months, and recently acquired players and coaches need some time to adjust to working with each other at live speed.
New seasons also bring rule changes and different points of emphasis from the officials, which can lead to a surplus of penalties that likely won’t occur later in the season (see, for instance, the Ravens’ numerous illegal formation penalties against the Chiefs). In short, there are plenty of reasons why one shouldn’t put too much stock in a Week 1 result—whether good or bad.
The data bears this out. In the plot below, I’ve computed the correlation between the winning percentage from the first x number of weeks of the season to the winning percentage for the remainder of the season from 2002 through 2023.
Unsurprisingly, the bigger the sample size of games, the more confident we can be about a team’s true ability. It is difficult to predict how good a team will be when we’re only a few games into the season, and it’s difficult to predict how a team will do in the remaining games when there are only a few games left.
The numbers suggest that only about 4% of a team’s winning percentage in its remaining 16 games can be explained by whether or not it won its first game; in other words, not very much.
But it isn’t just the small sample size: Week 1 seems to be a rather unreliable predictor even compared to other individual weeks of the NFL season, as shown in the following plot:
This plot depicts the correlation between a team’s result in a given Week x and their winning percentage in all other weeks of the season. The data are noisy, and it’s clear that no single week of the season is a great predictor of what happens the rest of the season.
Nonetheless, there is some evidence that the games in the middle of the season are a slightly better predictor of season-wide success than the games at the beginning (when teams are still getting back into the swing of things) and the games at the end (when teams may be dealing with injuries).
Week 1 was a great start for the Vikings, who look like a much more dangerous team now than they did a few days ago. But there are still 16 weeks to go in the NFL season, and the Vikings will next face three consecutive teams who made it to the divisional round in last year’s playoffs. Those tests should clarify whether the 2024 Vikings are one-hit wonders or Super Bowl contenders.
Unsung Viking Got a New Contract