Jene Bramel
Footballguy
For those of you who aren't into Bramel novellas, a short version with the highlights has been cross-posted to the FBG Blog.
We're going to have some fun with 2007 statistics over the next few weeks and see if we can glean some worthwhile information for buy low, sell high and tiering purposes. I freely admit that the metrics won't be perfect, but I'm a firm believer that there are some interesting tidbits to be learned if you consider the statistical measures in their correct context.
The spreadsheet in Part One of this series (attached below) was inspired by a metric that the folks at Football Outsiders publish in their annual Pro Football Prospectus. It calculates the percentage of the total team tackles made by each team's linebackers last season. PFP percentages include sacks, PD and INT numbers as well and are usually higher (in the 16-20% range in some instances). I'm just considering solo tackles here -- to remove the assisted tackle bias and focus primarily on the benchmark stat for IDP value.
To kick off the discussion...
Limitations:
The calculated percentage is derived from the AVERAGE per game team tackle total. Players who didn't play a full 16 games at their established position (injury, role change, etc.) will be skewed slightly. I adjusted Jon Beason, David Harris and Daryl Smith for their year end positions to eliminate some of that bias.
This is just one season's worth of data. I'll make an effort to generate some additional year's data to find some seasonal norms. It'll be worth noting just how far out of the norm the 15% plus tier may be and what kind of trends might emerge for players in the first three years or between schemes, etc. For now, I was just interested to see how the 2007 players compared.
The NFL.com stats include special teams plays in the total team tackle column. Thus, the raw percentages are actually higher.
Tackles are an UNOFFICIAL stat. Teams award solo and assisted tackles inconsistently.
RANK PLAYER TACKLE PERCENTAGE 1 ZACH THOMAS 15.74% 2 PATRICK WILLIS 15.64% 3 DAVID HARRIS 15.02% 4 JON BEASON 12.72% 5 NICK BARNETT 12.67% 6 DJ WILLIAMS 12.63% 7 MIKE PETERSON 12.02% 8 LONDON FLETCHER 11.95% 9 KIRK MORRISON 11.64% 10 DEMECO RYANS 11.24%Pretty much a who's who of 4-3 MLB and 3-4 ILB talent here. Ryans would have projected to 13.47% and 4th overall on this list if we assume he'd have continued his pre-injury pace over the final four games. I'll argue that it's not a coincidence that, despite having solid WLB next to them, the two teams who used the most Bates-like schemes in 2007 (GB and DEN) both placed their MLB in the top ten. And, again, this metric is proof-positive that a talented 3-4 ILB can make plenty of plays with a favorable combination of skill set, scheme and surrounding cast.Other interesting observations:
**Jacksonville MLB 2007
7 Mike Peterson 12.02%
61 Daryl Smith 7.34%
Acknowledging that Daryl Smith's 2007 MLB sample size was only five games, there's just no way that people should continue to argue that Smith is a better MLB than Peterson. Some of this discrepancy is Peterson making plays as a stud backer and some of it is Durant proving he's an NFL talent. But not all of it. The same statistical differences were evident in 2006. Smith has yet to be an IDP option at MLB, and I continue to believe that he never will be.
**The spreadsheet proves that a good Tampa-2 WLB can take advantage of plenty of opportunity.
15 Ernie Sims 10.63%
16 Freddie Keiaho 10.51%
22 Lance Briggs 10.24%
28 Derrick Brooks 9.78%
The first three on that list all had a higher tackling percentage than their MLB counterparts. Not that a MLB in a Tampa-2 scheme isn't valuable, but those MLBs are going to need a higher raw opportunity than their non-T2 counterparts.
**Tennessee OLB 2007
11 David Thornton 11.20%
53 Keith Bulluck 8.19%
Other than making a fool of myself for arguing that Thornton was a primo sell high candidate last offseason and that Bulluck was a lock to rebound at some point this season, I don't know what else this stat shows. I'm almost hoping to hear that Bulluck needs offseason surgery for some nagging injury. Going to be hard to avoid Bulluck as a LB2 next season, but I'm known to be stubborn about these kinds of things. Which is really part of the purpose for searching out new ways to break down the raw statlines -- we're big on acting rather than reacting and anticipating rather than falling behind the curve. Holding on to a guy like Bulluck to anchor your lineup for too long is a killer.
**The first linebacker that doesn't see every down snaps is Matt Wilhelm at #32.
And he really only sits in certain subpackages and dime situations. We argued the point with another hefty statistical slant in an early season Reading the Defense, but those that would argue that paying attention to who plays and who sits in the nickel defense is a fool's errand can take the non-nickel backers at their own risk. It's also worth noting that Wilhelm still finished with a higher percentage than LILB Stephen Cooper (#50 -- 8.49%), which again proves our 1-gap 3-4 corollary -- WILB are valuable targets in the 1-gap 3-4 schemes.
**35 Lofa Tatupu 9.64%
I've heard the arguments about how the loss of Marcus Tubbs affects the Seahawk run defense. I know Tatupu has never been a huge solo tackle guy. And I understand that Patrick Kerney, Leroy Hill and Julian Peterson are solid pursuit players and Seattle has one of the best run supporting corners in the league, but this number just doesn't make sense to me. I'll update this thread with Tatupu's 2005 and 2006 percentage at some point, but you have to think that Tatupu's 83 solos are an anomaly.
**17 Paul Posluszny 10.47%
Small sample size issues again, but don't forget about this guy in redraft leagues. Mid LB2 floor in 2008.
PCT_TEAM_TACKLE_BY_LB.xls
We're going to have some fun with 2007 statistics over the next few weeks and see if we can glean some worthwhile information for buy low, sell high and tiering purposes. I freely admit that the metrics won't be perfect, but I'm a firm believer that there are some interesting tidbits to be learned if you consider the statistical measures in their correct context.
The spreadsheet in Part One of this series (attached below) was inspired by a metric that the folks at Football Outsiders publish in their annual Pro Football Prospectus. It calculates the percentage of the total team tackles made by each team's linebackers last season. PFP percentages include sacks, PD and INT numbers as well and are usually higher (in the 16-20% range in some instances). I'm just considering solo tackles here -- to remove the assisted tackle bias and focus primarily on the benchmark stat for IDP value.
To kick off the discussion...
Limitations:
The calculated percentage is derived from the AVERAGE per game team tackle total. Players who didn't play a full 16 games at their established position (injury, role change, etc.) will be skewed slightly. I adjusted Jon Beason, David Harris and Daryl Smith for their year end positions to eliminate some of that bias.
This is just one season's worth of data. I'll make an effort to generate some additional year's data to find some seasonal norms. It'll be worth noting just how far out of the norm the 15% plus tier may be and what kind of trends might emerge for players in the first three years or between schemes, etc. For now, I was just interested to see how the 2007 players compared.
The NFL.com stats include special teams plays in the total team tackle column. Thus, the raw percentages are actually higher.
Tackles are an UNOFFICIAL stat. Teams award solo and assisted tackles inconsistently.
RANK PLAYER TACKLE PERCENTAGE 1 ZACH THOMAS 15.74% 2 PATRICK WILLIS 15.64% 3 DAVID HARRIS 15.02% 4 JON BEASON 12.72% 5 NICK BARNETT 12.67% 6 DJ WILLIAMS 12.63% 7 MIKE PETERSON 12.02% 8 LONDON FLETCHER 11.95% 9 KIRK MORRISON 11.64% 10 DEMECO RYANS 11.24%Pretty much a who's who of 4-3 MLB and 3-4 ILB talent here. Ryans would have projected to 13.47% and 4th overall on this list if we assume he'd have continued his pre-injury pace over the final four games. I'll argue that it's not a coincidence that, despite having solid WLB next to them, the two teams who used the most Bates-like schemes in 2007 (GB and DEN) both placed their MLB in the top ten. And, again, this metric is proof-positive that a talented 3-4 ILB can make plenty of plays with a favorable combination of skill set, scheme and surrounding cast.Other interesting observations:
**Jacksonville MLB 2007
7 Mike Peterson 12.02%
61 Daryl Smith 7.34%
Acknowledging that Daryl Smith's 2007 MLB sample size was only five games, there's just no way that people should continue to argue that Smith is a better MLB than Peterson. Some of this discrepancy is Peterson making plays as a stud backer and some of it is Durant proving he's an NFL talent. But not all of it. The same statistical differences were evident in 2006. Smith has yet to be an IDP option at MLB, and I continue to believe that he never will be.
**The spreadsheet proves that a good Tampa-2 WLB can take advantage of plenty of opportunity.
15 Ernie Sims 10.63%
16 Freddie Keiaho 10.51%
22 Lance Briggs 10.24%
28 Derrick Brooks 9.78%
The first three on that list all had a higher tackling percentage than their MLB counterparts. Not that a MLB in a Tampa-2 scheme isn't valuable, but those MLBs are going to need a higher raw opportunity than their non-T2 counterparts.
**Tennessee OLB 2007
11 David Thornton 11.20%
53 Keith Bulluck 8.19%
Other than making a fool of myself for arguing that Thornton was a primo sell high candidate last offseason and that Bulluck was a lock to rebound at some point this season, I don't know what else this stat shows. I'm almost hoping to hear that Bulluck needs offseason surgery for some nagging injury. Going to be hard to avoid Bulluck as a LB2 next season, but I'm known to be stubborn about these kinds of things. Which is really part of the purpose for searching out new ways to break down the raw statlines -- we're big on acting rather than reacting and anticipating rather than falling behind the curve. Holding on to a guy like Bulluck to anchor your lineup for too long is a killer.
**The first linebacker that doesn't see every down snaps is Matt Wilhelm at #32.
And he really only sits in certain subpackages and dime situations. We argued the point with another hefty statistical slant in an early season Reading the Defense, but those that would argue that paying attention to who plays and who sits in the nickel defense is a fool's errand can take the non-nickel backers at their own risk. It's also worth noting that Wilhelm still finished with a higher percentage than LILB Stephen Cooper (#50 -- 8.49%), which again proves our 1-gap 3-4 corollary -- WILB are valuable targets in the 1-gap 3-4 schemes.
**35 Lofa Tatupu 9.64%
I've heard the arguments about how the loss of Marcus Tubbs affects the Seahawk run defense. I know Tatupu has never been a huge solo tackle guy. And I understand that Patrick Kerney, Leroy Hill and Julian Peterson are solid pursuit players and Seattle has one of the best run supporting corners in the league, but this number just doesn't make sense to me. I'll update this thread with Tatupu's 2005 and 2006 percentage at some point, but you have to think that Tatupu's 83 solos are an anomaly.**17 Paul Posluszny 10.47%
Small sample size issues again, but don't forget about this guy in redraft leagues. Mid LB2 floor in 2008.
PCT_TEAM_TACKLE_BY_LB.xls
Last edited by a moderator:
on the record here for now.
