Much has been made about the Green Bay Packers’ 2019 season exceeding expectations and probability, suggesting that the team is due for a regression to the mean in 2020. It’s true that the Packers won close games frequently last fall, and they rode that success to a 13-3 record and a playoff bye.
But just how much of an outlier were the Packers last year? And what does that mean for the team moving forward? To examine those factors, we’ll take a look at the concept of Pythagorean Record.
The idea of the Pythagorean Record comes from baseball, but it is easily applied to football as well. In baseball, the theory goes that one can predict a team’s record with a high degree of confidence by looking at its run or point differential. If a team’s actual record varies significantly from its Pythagorean Record, that suggests that that team is over-performing and is a candidate to regress towards the mean.
A large sample size like a baseball season, with its 162 games, helps the formula work out fairly well. A tweak to the exponent tries to account for the differences in the two sports’ scoring systems, but the fact that football simply has fewer games means that the typical difference between a football team’s actual and Pythagorean winning percentages is larger than that of a baseball team. In 2019, for example, the standard deviation of the difference between NFL teams’ actual wins and Pythagorean wins was about 1.5, equating to about 0.094 in winning percentage.
What does this mean for the Packers? Simply put, the Packers overachieved in 2019 by the biggest margin of any NFL team since 2012.
The Packers outscored opponents by 63 points in 2019 — a decent enough number, but not one that suggests that they would win 13 games. Indeed, the Packers’ Pythagorean record was 9.7-6.3, and their 13-3 result saw them overachieve by 3.3 wins, or 0.206 in the winning percentage column. That was the biggest difference in either direction by any team last year; the Seattle Seahawks had a +2.8 win differential, while the Los Angeles Chargers underperformed by 2.8 wins as they went just 5-11.
In the last several years, Green Bay was within one win of its Pythagorean record every year between 2012 and 2018. The last time they they exceeded even a one-game difference was in 2011, when their 15-1 record represented a 3.1-win overperformance. Going back a bit further, the 2008 Packers underperformed by 2.9 wins, going 6-10 despite outscoring their competition by 39 points on the season. All told, the +3.3 mark in 2019 was the Packers’ highest variance in either direction in at least 50 years.
To find any another NFL team that had a bigger margin than that, one must look back to the 2012 season, when the Indianapolis Colts posted a truly incredible season. That Colts team, led by first-year head coach Chuck Pagano and rookie quarterback Andrew Luck, finished 11-5, with a record that included a 30-27 victory over the Packers in Indy. However, the Colts were outscored by 30 points, and their Pythagorean record was just 7.2-8.8, leaving them with a 3.8-win overperformance and a Wild Card berth despite being mediocre at best.
A look at the record explains the difference. Three of the Colts’ losses came by 20 points or more, including a 26-point loss to the Jets and a 35-point blowout at New England. Meanwhile, the Colts won six games by four points or fewer and won only two contests all year by more than a touchdown — one by 17 and one by 12.
The next year, the Colts improved the quality of their team, finishing with a 9.4-win Pythagorean projection in 2013. They did regress somewhat, however, finishing with a matching 11-5 record, but all measures of overall performance improved, including a jump in total DVOA from 25th in 2012 to 13th the following year. Seeing a 25th-ranked team in DVOA make the playoffs, when combined with a sub-.500 Pythagorean record, is a sign that something was very weird in Indianapolis that year.
The Packers, meanwhile, were still a good team in 2019, not a legitimately bad one like that Colts squad. Green Bay ranked tenth in overall DVOA last year, about on par with their ninth-place ranking in point differential. This was a team worthy of a playoff spot, even if they overachieved to earn a first-round bye, rather than a squad that had no business in the postseason.
Still, the cliche about the Packers potentially having a worse record but being a better team in 2020 holds merit. It will just depend on how close to the mean they regress.