The age old adage is “defenses win championships.”
However, sometimes the game changes and our adages may need to change with it. In regards to the NFL, the sport has become very offense-friendly in recent years given the rules now in place to protect quarterbacks and wide receivers. Accordingly, it may be worthwhile to reconsider if maybe the value of defenses isn’t what it once was.
Below is a chart noting the relative (best in the league to worst in the league) rank of each team that played in a Super Bowl since 2012. Notably, these ranks are based on expected points added (EPA) per play, which you can read more about the calculation of here. Broadly speaking, EPA per play estimates the degree to which an offense or defense increases their teams likelihood of winning on their average play. In terms of interpreation of this chart, a team that is further right was ranked closer to the best in the league on offense while a team further to the top was ranked closer to the best in the league on defense.
where each team that has made the super bowl since 2012 ranks on offense and defense— Tej Seth (@tejfbanalytics) February 8, 2023
only one below-average offense has made the super bowl in the past decade pic.twitter.com/yZedTljeB4
As shown here, the teams that typically make the Super Bowl are good on both offense and defense as we would expect; this is shown by the clustering of teams in the top-right corner. However, it does seem a team that is exceptional on offense and okay-to-bad on defense is more likely to make the Super Bowl than a team that is exceptional on defense and okay-to-bad on offense; this is shown by the greater degree of clustering in the bottom right hand corner than the top left hand corner.
For the New York Jets, this hammers home the importance of improving the offense this season if they hope to make the Super Bowl next season. If not, we can only hope that they can be the outlier datapoint that the 2015 Broncos are even if that does seem to be an unlikely path to success.
But what do you think? Does anything else about this chart jump out to you?