Considering the impact of Wyatt's absence*...
In weeks 1-12 (he was injured Wk 12), we were ranked:
13th in Total Defensive EPA
12th in Defensive Rush EPA
17th in Defensive Rush Success Rate**
And for full context that also includes:
10th in Total Defensive Success Rate
16th in Defensive Passing EPA
7th in Defensive Passing Success Rate
Following that injury, in weeks 13-16 we dropped to:
28th in Total Defensive EPA (-15)
22nd in Defensive Rush EPA (-10)
31st in Defensive Rush Success Rate (-14)
and...
30th in Total Defensive Success Rate (-20)
28th in Defensive Passing EPA (-12)
23rd in Defensive Passing Success Rate (-16)
*Of course, this is a relatively small sample window, and also includes the injury to Parsons for the majority of it.
**For those who are not aware, "Success Rate" is basically measuring how well an offense/defense executes when it comes to "staying on schedule."
It is the percentage of plays that gain A.) 40% of required yardage on 1st down; B.) 60% of required yardage on 2nd down; and C.) 100% of required yardage on 3rd or 4th down - either a conversion or touchdown.
So a "successful play" would gain 4+ yards on 1st and 10, or 6+ yards on 2nd and 10, or 10+ yards or a TD on 3rd/4th and 10. Alternatively, if a team gains 4 yards on 1st and 10, 4 yards (3.6) on 2nd and 6, and picks up the remaining 2+ yards on 3rd and 2, all three would be considered successful plays.
Defensively is of course just the opposite - it's how often you're able to prevent the offense from keeping "on schedule" and gaining the requisite yards as defined above.
To put it simply one might also reasonably conclude that Success Rate better captures "efficiency" or "consistency" but measures such as EPA are better at accounting for "explosive plays" or the like. You have "bend but don't break" teams who might allow the opponent to drive with some consistency, but they keep everything in front of them and don't give up explosive home-run type plays. They will likely have a lower success rate but more favorable (relatively speaking) defensive EPA measure. On the other hand a "high risk, high reward" defense that gives up some big shots every now and then but generally doesn't allow a ton of long drives might have a worse (again, relatively) defensive EPA but better success rate.
And a defense like ours right now - ranking low in both success rate and (pass and rush) EPA against - is basically letting opponents move the ball at will
