NFL Quarterback Ratings (DQR) - 2016

[deljzc]

As many know, I revised and edited the old quarterback rating system. I wanted something better than the old system and not quite as arbitrary as ESPN's QBR (and maybe a bit easier to understand).

I call it DQR (for del's quarterback rating).

Like the traditional quarterback rating we all know, it comes from available box score information and measures the following:

1. Completion %
2. Yards per Pass Play (includes all drop backs and scramble/running yards)
3. Points per Possession for the Offense
4. Turnovers (INT's and Fumbles Lost)
5. Rewards Big Plays (20+ yards) and Penalizes Sacks (in a ratio type formula)

For reference the average traditional quarterback rating this season was 87.6. The average ESPN DQR rating this season is 61.2.


My league DQR is 82.4 and is based on the following with the averages for this 2016 season as indicated below:

Average Completion % - 63.0% = 17.3 rating points
Attempts + Scrambles + Sacks = 682
Yards Passing + Scrambling = 4248
Yards/Pass Play = 6.23 = 18.6 rating points
Points per Possession = 1.97 = 15.4 rating points
Turnovers = 16.5 = 22.9 rating points
Average 20+ plays = 50/team
Average Sacks = 35/team
Big Play - Sacks/Pass Play = 2.2% = 8.3 rating points


I understand most rating points are gained through turnovers and that is a growing trend (it's even more of an impact with the old, traditional rating system). Teams and quarterbacks just continue to turn the ball over less. Even compared to 2010 (when I started working on this formula), teams turned the ball over on 13.3% of possessions. Now it is down to 11.3%. In raw numbers there were 860 turnovers in the 2010 NFL season while this season there were only 700. That's almost a 20% decrease in turnover just over the last 6 seasons.

In terms of quality quarterback play with regard to "protecting the football", there is no doubt the league and QB's as a whole have DRASTICALLY improved this part of their job over the last 10-15 seasons.

That said, here are the DQR scores of every team this season.

 

[slashsteel]

don't see em

 

[deljzc]

1. Atlanta (M. Ryan) = 126.7 (new record)

2. New England = 117.5 (Brady is 123.6)

3. Washington (K. Cousins) = 109.4

4. New Orleans (D. Brees) = 108.3

5. Dallas (D. Prescott) = 107.9

6. Green Bay (A. Rogers) = 102.0

7. Pittsburgh = 98.7 (B. Roethlisberger = 102.6)

8. Oakland (D. Carr) = 98.4

9. Detroit (M. Stafford) = 95.1

10. Tennessee (M. Mariota) = 94.6

11. Indianapolis (A. Luck) = 94.2

12. Minnesota (S. Bradford) = 93.1

13. Cincinnati (A. Dalton) = 92.0

14. Seattle (R. Wilson) = 90.9

15. Kansas City (A. Smith) = 90.2

16. Miami = 88.3 (R. Tannehill = 84.1)

17. Buffalo (T. Taylor) = 87.1

LEAGUE AVERAGE = 82.4

18. Arizona (C. Palmer) = 82.3

19. San Diego (P. Rivers) = 82.1

20. Baltimore (J. Flacco) = 78.4

21. Chicago (Barkley/Hoyer/Cutler) = 78.4

22. New York Giants (E. Manning) = 78.0

23. Philadelphia (C. Wentz) = 76.2

24. Denver = 75.8 (T. Siemian = 78.1)

25. Tampa Bay (J. Winston) = 75.4

26. Carolina (C. Newton) = 72.0

27. San Francisco = 70.9 (C. Kaepernick = 74.2)

28. Jacksonville (B. Bortles) = 70.6

29. Houston = 65.4 (B. Osweiler = 63.5)

30. Cleveland (Various) = 63.1

31. New York Jets = 62.3 (R. Fitzpatrick = 69.2)

32. Los Angeles Rams = 56.8 (C. Keenum = 70.5)


Remember, this is not a definitive ranking of quarterbacks, just a measurement of their work in the 2016 season. Potential, past success, injuries around them, etc. are just part of the ebb/flow of statistics for any professional athlete.

 

[Chiba]

So you're telling me that Flacco is not elite, Osweiler sucks, anyone that suits up for Cleveland sucks and Dalton is just a little bit above average?

 

[Troglodyte]

1. Completion %
2. Yards per Pass Play (includes all drop backs and scramble/running yards)
3. Points per Possession for the Offense
4. Turnovers (INT's and Fumbles Lost)
5. Rewards Big Plays (20+ yards) and Penalizes Sacks (in a ratio type formula)
.

Some thoughts and questions:

#3. So if a RB takes a handoff and goes the distance on a drives first play, that improves the QBs rating?

There's some redundancy in your metrics: 5 and 2, 4 and 3, and 2 and 1 are dependent on the each other so you're measuring some of the same thing twice.