Pass Rush Productivity (or why I may have been wrong about Chauncey Davis)
Piggybacking off Dave's post about drafting a pass rusher (good thing I was already working on this, huh?), I've got some pretty cool numbers to share with you guys. Pro Football Focus recently released their lists of the most productive pass rushers in 2009. They include not only sacks, but also QB Hits and QB Pressures, and - here's the big one - divide that by the number of snaps the player was on the field for pass rush situations to come up with a number they are calling Pass Rushing Productivity (PRP) Rating.
For too long sacks have been the measuring stick in judging the productivity of a pass rusher. Going a bit deeper, Pro Football Focus’ Khaled Elsayed has decided to look at the hits and pressures a player gets, alongside their total number of snaps rushing the passer, to work out just who are the most productive pass rushers.
As for the subtitle? Recently, while backing up Dave's position that Chauncey Davis could be a possible candidate to be traded (of flat-out dumped) to save money against the Salary Cap, I said (among other things):
Yeah, Chauncey was pretty bad this year.
Upon closer inspection, it might not have been as bad as the Falcons throwing $5 Million down the toilet last year (which was pretty close to my thinking on Chauncey). But I'll get to him in detail a bit later. First, follow the jump, and we'll go over the methodology for (and limitations of) the PRP Rating.
First the formula, from PFF:
Total Sacks + (Total Hits x 0.75) + (Total Pressures x 0.75) = Pass Rushing Points
Pass Rushing Points/ Number of Snaps Rushing Passer x 100 = Pass Rushing Productivity Rating
The (x100) is just so it's a decent number (somewhere usually between 1 and 15 or so), instead of 0.095 or something similar. The big question is whether QB Hits and QB Pressures are really three-fourths as good as a Sack. I tend to agree with them that disrupting the QB is nearly as important as getting the actual sack. What do you think? Is 75% ok? Half?
The other issue is the numbers themselves, especially the subjetivity of QB Pressures. For this PRP Rating to be useful for us, we have to assume that their game charters are very accurate on their totals for QB Pressures (it's not an offical NFL stat in the Gamebooks) and that their snap count is also very accurate.
You should really read the article at PFF, but for comparison's sake, here's their Overall Top 10 (well, 11 with a tie), minimum 200 Pass Rush snaps:
- Dwight Freeney IND 13.7 (4-3 DE)
- Lamarr Woodley PIT 12.4 (3-4 OLB)
- DeMarcus Ware DAL 12.0 (3-4 OLB)
- Robert Mathis IND 11.2 (4-3 DE)
- Elvis Dummervil DEN 10.9 (3-4 OLB)
- James Harrison PIT 10.8 (3-4 OLB)
- John Abraham ATL 10.5 (4-3 DE)
- Aaron Kampmann GB 10.2 (3-4 OLB)
- Leonard Little STL 10.2 (4-3 DE)
- Ray Edwards MIN 10.1 (4-3 DE)
- Tamba Hali KC 10.1 (3-4 OLB)
Now that we have that out of the way...
Defensive Ends
| PRP Rating | Name | Pass Rush | Sacks | Hits | Pressures |
| 10.51 | John Abraham | 421 | 6 | 12 | 39 |
| 7.45 | Chauncey Davis | 245 | 1 | 2 | 21 |
| 7.01 | Kroy Biermann | 339 | 5 | 7 | 18 |
| 4.82 | Jamaal Anderson | 285 | 1 | 3 | 14 |
| 2.69 | Lawrence Sidbury | 65 | 1 | 1 | 0 |
Abe's Sack number was down, but he was still getting Pressure on the QB. Does that mean he lost a half a step? Maybe, but he still did his part as a pass rusher. Chauncey only had the 1 sack, but got more QB Pressures than Biermann in fewer snaps. Both were about about the middle of the pack for 4-3 DE's. What can we say about JA98 at this point? He has zero ability to rush the passer. To be fair to him, I'm going to list him with the DT's too, since he played so many snaps there in the second half of the season. Sid had the one sack late against Philly, but other than that really did nothing much to speak of (remember it was only 65 snaps, though). I don't think the team should not select a DE in the first round just because we think he might develop into something.
By the way, I mentioned earlier about Cover-2 teams needing two bigtime pass rushing DE's, and here's some of the guys for the teams I mentioned, for comparison:
Colts - Dwight Freeney (13.04) and Robert Mathis (10.81); both in the Overall Top 10 (I'm not sure why my numbers come out slightly different, though).
Vikings - Ray Edwards (11.37) and Jared Allen (9.67); Allen just missed the Top 10.
Panthers - Julius Peppers (9.47), Charles Johnson (8.33, in 264 snaps), and Tyler Brayton (7.41 in 381 snaps).
Defensive Tackles
| PRP Rating | Name | Pass Rush | Sacks | Hits | Pressures |
| 6.25 | Peria Jerry | 36 | 0 | 2 | 1 |
| 5.81 | Jonathan Babineaux | 508 | 7 | 5 | 25 |
| 4.82 | Jamaal Anderson | 285 | 1 | 3 | 14 |
| 4.19 | Thomas Johnson | 215 | 3 | 1 | 7 |
| 3.95 | Vance Walker | 114 | 0 | 0 | 6 |
| 3.30 | Trey Lewis | 91 | 0 | 2 | 2 |
Obviously, you gotta take Jerry's number with a grain of salt, since he only played in (slightly less than) two games. In the PFF article, they note that Babineaux finished #5 in DT's - very impressive. Couple that with his Stop Rate numbers vs the Run, and see if you can explain to me why he wasn't in the Pro Bowl (oh yeah, the weed). JA98 looks better here, but then probably half his snaps were at DE. Johnson and Walker both have decent numbers, maybe slightly above average for DT's. Lewis is lagging behind, and had pretty much fallen back out of the rotation by the second half of the season.
Linebackers
| PRP Rating | Name | Pass Rush | Sacks | Hits | Pressures |
| 17.55 | Stephen Nicholas | 47 | 3 | 2 | 5 |
| 17.31 | Coy Wire | 13 | 0 | 2 | 1 |
| 8.15 | Curtis Lofton | 92 | 0 | 3 | 7 |
| 5.05 | Mike Peterson | 109 | 1 | 1 | 5 |
These numbers are starting to get pretty volatile because of the small sample size (fewer snaps), but one thing is very clear to me - they've gotta dump the plan of blitzing Peterson. He was near the absolute bottom of the league as a pass rusher according to these numbers. 109 snaps is a pretty good sample, and he only got one Sack, one Hit, and five pressures. All that accomplished was to have one fewer guy in coverage. Hopefully in 2010, Nicholas will be the one blitzing in Peterson's place, and we'll see if his great numbers are real. Curtis Lofton had zero Sacks, but had a decent amount of Pressures. He was maybe slightly below average for a MLB, for example: Patrick Willis (9.19), Gary Brackett (8.60), Antonio Pierce (8.12), and Jon Beason (7.61).
Defensive Backs
| PRP Rating | Name | Pass Rush | Sacks | Hits | Pressures |
| 26.79 | Brian Williams | 14 | 0 | 1 | 4 |
| 17.31 | Thomas DeCoud | 26 | 3 | 1 | 1 |
| 11.54 | Erik Coleman | 39 | 0 | 1 | 5 |
| 11.21 | Chevis Jackson | 29 | 1 | 0 | 3 |
Finally, the numbers say something nice about Brian Williams. Of course, that was only on 14 snaps, so... DeCoud was slightly above average, and Coleman and Jackson were slightly below average for DB's. But really it's hard to tell anything about the D-Backs - there were only about 9 guys in the whole league that rushed the passer even 50 times. All the other DB's for the Falcons had either 1 or 0 pass rushes.
Chauncey Davis
Well, according to these numbers, he was the Falcons' 2nd most productive DE rushing the passer on a per-play basis. And in case you didn't notice him in the link about Babs' Stop Rate in the run game, I'll link it again. He was the #7 Defensive Lineman in Stop Rate last season (minimum 16 run tackles), according to Football Outsiders. He still didn't make that many plays, but he was pretty good when he did make a play. He did get the 4th most snaps among the team's defensive linemen, though (behind Babs, Abe, JA98, and ahead of Biermann in total snaps), so he really should be making more plays. The bottom line is that while he was probably overpaid last year ($2M Salary, $3M Signing Bonus for his contract extension), it wasn't like the team set fire to $5 Million at the 50 yard line.
As always, go to Football Outsiders and Pro Football Focus for stats and information beyond what the box scores and official NFL leaders can tell you.
This FanPost was written by one of The Falcoholic's talented readers. It does not necessarily reflect the views of The Falcoholic.
15 comments
|
5 recs |
Do you like this story?
Comments
Those are intersting numbers.
In addition to what they say about Chauncey, they say we need to climb off Abe’s back. Hmm.
How many more days, Lord, must I walk through the wilderness?
GO FALCONS!!!
Best freaking breakdown ever!
Thanks for the jibs. So we need to replace Peterson maybe first and formost with #1 pick maybe kindle then pick up dumervil in FA. Draft Mardy Gilyard in 3rd pick up a CB in FA.
"at least your roof only leaks when it rains."
by Ball Hawk on Feb 12, 2010 7:12 PM EST via mobile reply actions
Peterson
I wouldn’t dump him at all. I thought he did a decent job when not rushing the QB, it’s just a question of how you utilize him.
Agreed
I think they probably need to reduce his number of snaps (he had the second most on “D” last year) by getting him off the field in certain situations. Like you said, he was effective in other situations. Making sure a 34 year old LB gets a breather every so often is probably wise, anyway.
"Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital." - Aaron Levenstein
Great post!
Great job. Thanks for all your work.
Gjmoody
by Gjmoody on Feb 12, 2010 9:03 PM EST via mobile reply actions
I always said you have all the numbers possible! :)
Great job!
I would argue with the methodology though – I disagree with the hits and pressures carrying a 3/4 value of a sack. While I agree that hits and pressures may disrupt the normal flow of passing game, it really depends on the opposing QB. Some QB simply perform better when they face blitz, and unless you sack them they will hurt you.
Case at point: Abe. It might be only my impression, but too often I saw him nearly get to the QB (and that is reflected in these stats putting him in the top 10) only to see the QB make great play. Clearly, in this situation (when Abe rushed the passer, did not get to him and the QB completed a long pass) hits and pressures should carry a lower value if any at all. On the contrary, there were some QB’s who’d simply freak out when facing pressure and fire it away. But even in this case I’d assign a 50% value to a hit or a pressure.
It’s easy to critisize, and difficult to propose a solution. An ideal (and probably impossible) solution would be to assign some value to the hits and pressures if the play did not result in, say, any gain, or at least a big gain.
Atlanta Falcons fan in Moscow, Russia
Yeah, that's probably impossible
Well, I’d have to go back through every play and re-assign them based on whether the result was a “big gain” or not. And that’s not gonna happen… All I did to get these numbers was to dump all the numbers from PFF into Excel.
If you think about it league-wide, it might not be fair to do it that way, either. Because if Abe gets a “pressure”, but the secondary still allows a big play, how do you compare that to if (for instance) Robert Geathers (Cincinnati) gets a “pressure”, but Leon Hall and Jonathan Joseph have their guys covered? It depends on the coverage and, like you said, the QB the team is facing. It’s definitely an interesting thought, though.
I might try to re-run the numbers based on only 50%, just to see what happens.
"Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital." - Aaron Levenstein
re-running it again at 50%
sounds like a good idea.
the reason why I brought up this issue was simply because the image of Abe imprinted in my mind over the season is the following – QB drops back to pass, has enough time to drink a cup of tea, then finally fires downfiled and completes his pass while Abe gets to him after the QB gets rid of the ball. And you’re left wondering what would have happened if Abe were just a half-second faster.
But I guess what I suggested would be a measure of the effectiveness of not just the DEs but rather of the whole defense (DE’s + DB’s).
Atlanta Falcons fan in Moscow, Russia
Download my spreadsheet
There is a tab for 75% and one for 50%. Honestly, there wasn’t a lot of movement (just look at the Rank). Abe was still near the top, Chauncey moved down about 10 spots (behind Biermann), Jamaal Anderson moved down about 15 spots, and Sidbury moved up a bit (since he had an actual Sack, while Vance Walker and Trey Lewis only had Hits/Pressures).
"Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital." - Aaron Levenstein
now I can see
based on these numbers he was better than many other elite pass rushers, including Peppers whom many want to see in our team…
Atlanta Falcons fan in Moscow, Russia
try 66.66% too
2/3 value isn’t quite as bad a hit.
Great article and REC’d
Interesting Fact
Khaled, the guy who wrote the article at PFF, is a Falcons fan himself!
go hard or go home
Excellent article
That confirms what I thought I saw all year
Honest Abe is still elite – ranked higher than Jared Allen – which is what I said last year, only to get laughed at
He’s always open. He catches a lot of balls. He’s un-guardable, no matter how old he is
Thanks
Yeah, I think he can be really good for probably 2 more years. I’m expecting a big bounce-back year next season (not 16 sacks of course, but maybe 10 or so).
"Statistics are like bikinis. What they reveal is suggestive, but what they conceal is vital." - Aaron Levenstein

by 


















