Jan 13, 2013; Atlanta, GA, USA; Seattle Seahawks cornerback Richard Sherman (25) runs to the sideline with strong safety Kam Chancellor (31) and cornerback Brandon Browner (39) before the first quarter of the NFC divisional playoff game at the Georgia Dome. Mandatory Credit: Daniel Shirey-USA TODAY Sports

Closing In On NFL Defensive Evaluation Tool


I have finally found something mathematically significant in my now week long study of NFL defenses. It’s not what I was hoping for, but ultimately it might be the best measure for the quality of a defense that we have. I’m going to continue to work with the data to develop a statistical model that fits all the data that I’ve presented over the last week, but I’m now very close to the ultimate defensive analysis tool I was after.

First a recap:

The NFL uses yard allowed to rate defense. We all know this is extremely flawed. Yards allowed doesn’t correlate highly to points allowed, and thus it doesn’t correlate to wins. It’s just a poor measure. I also showed that yards allowed was highly dependent on the # of possessions, which is something the defense has no control over.

The other measurement current used is points allowed. This is especially true for fans of teams like Seattle; a team that was #1 in that category, despite being less than steller in other statical measures. The problem is that this is also dependent on the number of possessions, and I also have shown that it is dependent on the average starting field position for the opponents. Again, these are things that the defense can’t control.

This problem led me to break down the results of each drive by it’s conclusion: touchdowns, punts, etc. The result there was too many variables and nothing to regress that data to. While there was plenty of anecdotal results that were very interesting, there wasn’t anything significant mathematically that could be found at this point.

Recap over. Now on to new results.

This time,  looked at something simple: yards per drive. It’s is one of the only factors that the defense has full control over. The better the defense, the quicker they’ll get off the field, right? Sure, there’s the “bend but don’t break” philosophy  but while that might be functional, it’s is hardly dominant.

The problem here is that this evaluation of the defensive play doesn’t correlate highly to anything significant, like points given up, or wins. There’s just too many other factors that effect those things we’d want to correlate it to.

To begin to account for that, I took the average starting field position for each defense, and added that to the yards given up to give me the average yard yard where drives ended.

That number, actually correlated quite nicely to points given up. It’s still not perfect, and there’s still one more variable I need to account for (getting turnovers appears to be a big factor), but I think that I may have stumbled on to something that is very mathematically significant.

Unfortunately, it’s very late here, so that’s going to have to wait until tomorrow. I will say that once I account for turnovers, i believe I’ll end up showing that yards per drive is an extremely accurate indicator of how well a defense played.

Until then, I’ll just leave you with today’s data:


Rk Team yards/dr Rk Team ave drive end
1 Denver Broncos 25.01 1 Denver Broncos 49.24
2 Pittsburgh Steelers 25.36 2 San Francisco 49ers 47.87
3 Chicago Bears 26.30 3 Chicago Bears 47.75
4 Arizona Cardinals 26.37 4 Cincinnati Bengals 47.04
5 Houston Texans 26.39 5 Arizona Cardinals 45.97
6 San Francisco 49ers 27.23 6 Seattle Seahawks 45.93
7 Cincinnati Bengals 27.50 7 Houston Texans 45.63
8 New York Jets 27.67 8 Pittsburgh Steelers 44.77
9 Green Bay Packers 29.28 9 Green Bay Packers 44.76
10 Detroit Lions 29.50 10 San Diego Chargers 44.19
11 Baltimore Ravens 29.55 11 New York Jets 43.94
12 San Diego Chargers 29.68 12 Baltimore Ravens 43.43
13 Seattle Seahawks 29.87 13 New England Patriots 43.13
14 Cleveland Browns 30.01 14 Cleveland Browns 42.85
15 Minnesota Vikings 30.94 15 Minnesota Vikings 41.80
16 St. Louis Rams 31.15 16 Atlanta Falcons 41.67
17 Philadelphia Eagles 31.20 17 Tampa Bay Buccaneers 41.08
18 New England Patriots 31.27 18 Detroit Lions 40.91
19 Carolina Panthers 31.35 19 Indianapolis Colts 40.41
20 Oakland Raiders 31.51 20 Miami Dolphins 40.23
21 Miami Dolphins 32.07 21 Carolina Panthers 40.00
22 Buffalo Bills 32.44 22 Dallas Cowboys 39.94
23 Tampa Bay Buccaneers 32.50 23 St. Louis Rams 39.26
24 Tennessee Titans 32.60 24 Jacksonville Jaguars 39.18
25 Kansas City Chiefs 32.97 25 Kansas City Chiefs 39.05
26 Atlanta Falcons 33.23 26 Washington Redskins 38.23
27 Jacksonville Jaguars 33.27 27 Oakland Raiders 37.62
28 Dallas Cowboys 33.85 28 Tennessee Titans 37.36
29 Washington Redskins 34.14 29 Buffalo Bills 37.35
30 Indianapolis Colts 34.61 30 Philadelphia Eagles 36.94
31 New York Giants 36.08 31 New York Giants 36.08
32 New Orleans Saints 37.66 32 New Orleans Saints 33.02

Dick's Sporting Goods presents "Hell Week":

Tags: NFL Seattle Seahawks

  • Hawkman54

    The first column is really mind boggling when you consider the games from last year . Now the second column is more what I would have thought it would be , although I am surprised SF is that high, they had a mediocre secondary. I would have figured Seattle at 5 ( oh well) But am really surprised to see Green Bay that high . Watching their games you certainly wouldn’t have expected anywhere near that ranking .

    • 12thMan_Rising

      I’m right there with you on that. As I get into this, I’m learning a lot about how I, and fans as a whole, view certain teams. I didn’t expect the Seahawks to be the #1 defense overall, but for a measure to suggest they are #13 is quite unexpected. And for the Lions and Packers to be above them, that is just messing with my entire reality.