It's tough to separate the performance of receivers and the coverage unit from that of the quarterback, when grading players.

I'd imagine this relationship lends to both the importance and instability of coverage grades. https://twitter.com/adrian_stats/status/1311089683565285376

For those unaware, PFF coverage grade is highly associated with good team results. And while it does predict future team results a bit, it's still relatively unstable. This may be partly because coverage players appear to grade worse when playing against better QBs.

Assessing defensive play in a meaningful way is really, really hard. Comprehensive grades are probably still the best resource we have right now for assessing performance in coverage. But there's still room for improvement.