≡ Menu


Take a look at the Broncos pass defense this year, compared to the Broncos pass defense last year:

Year Tm G Cmp Att Cmp% Yds TD TD% Int Int% Y/A AY/A Y/C Y/G Rate Sk Yds NY/A ANY/A
2016 DEN 14 260 483 53.8 2562 10 2.1 12 2.5 5.8 5.1 10.7 183.0 67.5 40 221 4.9 4.2 7.6 72.16
Year Tm G Cmp Att Cmp% Yds TD TD% Int Int% Y/A AY/A Y/C Y/G Rate Sk Yds NY/A ANY/A
2015 DEN 16 344 573 60.0 3193 19 3.3 14 2.4 6.2 5.7 10.3 199.6 78.8 52 351 5.1 4.7 8.3 59.89

Three years ago, I looked at the Seattle pass defense and calculate how many standard deviations above average the Seahawks were. At the time, I compared them to an average of the other 31 defenses rather than an average of all 32 defenses, including themselves. I don’t know if there’s a right or wrong answer there, but I’m going to use the latter methodology today, which will explain why the numbers are slightly different.

Anyway, Seattle was 2.80 standard deviations above average in ANY/A allowed in 2013. That’s because Seattle’s pass defense allowed 3.19 ANY/A, while the league average was 5.89 ANY/A. That’s a difference of 2.70 ANY/A, and the standard deviation among the 32 pass defenses that year was 0.97. Divide 2.70 by 0.97, and you see that Seattle was 2.80 standard deviations above average.

The 2016 Broncos? They are allowing just 4.25 ANY/A. That is over a full yard “worse” than Seattle, but worse needs to be put in quotes. For starters, the league average is 6.25 ANY/A this year; in addition, the rest of the league is bunched together. The standard deviation for the 32 pass defenses is 0.74 ANY/A. That means the Broncos have a Z-Score of 2.69 standard deviations better than average (here, negative is better).

That puts Denver as the 5th best pass defense, by this metric, since 1970:

RkTeamYearANY/ALg AvgLg StDvZ-Score

With two games left, there’s still a chance that the Broncos rise (or fall) in these rankings. But don’t let the 8-6 record fool you: Denver has a historically great pass defense. In fact, the 8-6 record might be the best evidence of that yet: that’s because Denver is bad at just about everything else. The Broncos are terrible at running and pretty bad at stopping the run; meanwhile, the passing offense is below-average, too. Given this team even an average pass defense, and they would likely be one of the worst teams in the league.

  • You couldn’t go all the way back to 1970, but it would be interesting to see fumbles forced on pass plays too (or at least QB fumbles; WR fumbles seem much more random). Right now, it’s difficult to use public information to parse QB fumbles on pass versus rushing plays more than a decade back.

  • The Mysteries of Rosewood

    Makes me wonder where the 2016 Broncos would place in the Billick index.

  • Tom

    Great post…yeah, with even just an average offense, Denver would be pretty scary right now.

  • Adam

    I’m not entirely sold on using standard deviation instead of RANY/A, but regardless it’s amazing that Denver’s pass D has somehow improved over last year’s dominant unit.

    • I like using SD if we are measuring dominance in a season.

      Say two seasons have identical averages, maybe 6.00 ANY/A.
      Season A sees most teams huddled between 5.00 and 7.00 ANY/A.
      Season B sees a bunch of teams at 3.00 ANY/A and a bunch at 9.00 ANY/A.
      A defense in Season A allows 3.00 ANY/A, which is a full 3.00 better than average.
      A defense in Season B allows 3.00 ANY/A, which is a full 3.00 better than average.
      Both these defenses have the same RANY/A, but I think the team from Season A is clearly the more historically dominant. They could be ~2.45 SD better than average, whereas the other team may only be 1.00 SD better than average.

      This is especially important, to me, when looking at pre-1950 seasons where the talent disparities were enormous.

      • Adam

        I agree that SD is superior when comparing modern seasons to pre-modern seasons, for the reason you stated. But I disagree with SD for comparing seasons within the same era, as I believe it injects more noise than signal. The distribution of league talent was roughly the same in 2013 as it is in 2016, so it seems wrong to penalize the Legion of Boom with a 25% higher SD than the No Fly Zone.

        • I think the point is to highlight that they stood out more, regardless of the actual distance from the league mean.

          • Adam

            If we’re going for “stood out” rather than “best”, I agree with you. It’s like the argument between elevation and prominence in measuring mountains…is it more impressive to be higher or to stand out more from your surroundings?

            • Well, Everest would look at lot more astonishing in the middle of Florida or the Outback.

  • sacramento gold miners

    I think what’s helped Denver’s pass defense rating is the strategic nature of football. Why fool around with throwing the ball against a good pass rush and secondary when your offensive line is winning the physical battle to run the ball? Not only has the overall defense declined, but the offensive issues have threatened to keep the Broncos home for the playoffs. I’m surprised John Elway didn’t pursue QB Matt Hasselback(who later decided to retire), instead of taking the other route. Hasselback played well in relief of Andrew Luck last season.