In the table below I report two numbers: the worst defenses difference in performance compared to the best defense through that season's week #6 (Overall) and the difference between the worst defenses' performance and the second to worst team's defensive performance through that season's week #6 (Next Worst).
Season | Team | Overall | Next Worst | ||
2011 | Kansas | -152.632 | -3.753 | ||
2012 | Massachusetts | -213.481 | -2.472 | ||
2013 | New Mexico State | -265.320 | -19.773 | ||
2014 | UNLV | -181.704 | -1.941 | ||
2015 | UTEP | -229.388 | -1.178 | ||
2016 | Charlotte | -254.605 | -22.784 | ||
2017 | Connecticut | -265.607 | -12.381 | ||
2018 | Connecticut | -269.982 | -24.846 |
As you can see above through week #6 of the season, Connecticut has the worst Overall defense as compared to the best defense for each season since 2011 (slightly ahead of their prior seasons performance) and also the worst defense as compared to the next worst defense for each season since 2011. Admittedly, one of the problems of the analysis above, is that each season is different, and it might not be accurate to compare one season with another season.
Is there a way to overcome this problem? I propose the following: let's use the defensive performances through week #6 for the last eight years (time period that I have weekly data) and apply the Complex Invasion College Football Production Model over the entire time period. The impact that each action on the field will be weighted the same for all defenses over those eight years and as such, I can compare any teams defense during this time period against any other defense during this time period since I am using the same measure for all the teams.
So, I gathered the data, ran the regression and calculated each team's defensive production. I ranked the defenses from worst to best, and I found that the 2018 Connecticut Huskies have the worst defense through week #6 since 2011. In fact, its not even close. This year's Huskies are over 16% worst than the next worst defense during the 2011-2018 time period through six weeks. In other words, find the second to worst defense, Connecticut is over 16% worse than that through six weeks.