Jene Bramel
Footballguy
Thanks to the questions and issues raised by Rovers and DonFue in other threads, I’ve been taking another look at the 3-4 defensive schemes around the league, and in a more detailed and systematic fashion than I confess I did at the beginning of 2006 when I first joined the staff.
If you don’t care about scheme issues or don’t think predicting which ILB in the 3-4 is the better IDP option is valuable information, stop reading now.
The upshot of the rest of this post for you is that there is no reason to believe that Patrick Willis cannot become an absolute stud in San Francisco, despite what some would have you believe about all 3-4 ILBs.
For the rest of you who have closely followed my ramblings on the 3-4, it appears, after beginning to look through Gamebook data (correct and incorrect), playbooks, coaches’ quotes and discussion, and some old discussions I’ve had with the Guru himself (Norton), that I have been working under some faulty assumptions and conclusions. Some were correct anyway, some were not.
I’ve still got some research to do and am hoping to present a cleaner, clearer and hopefully useful article about the evolution of the 3-4 and its IDP implications later this summer. But, since there seem to be some clear trends that are different than what I’ve presented in the past, I felt I should post the short version here for those who have interest and begin to set the record straight.
So…the short version. And yes, as you’ll see this summer, this is the short version.
There are two major 3-4 defensive philosophies. Sparing you the historical and X and O details, one branch of the tree originated from Bum Phillips and was modernized by his son Wade, the other branch got its professional start with a couple of Patriot coaches who influenced Bill Parcells, Bill Belichick, and, indirectly, the Steeler coaches of the early 1990s.
Below is the raw data from 2002-present that I’ve collected from the FBG/Drinen database. There’ll be more data from before 2002 to be presented in the coming detailed summer article.
* denotes more than one tackle per game difference over 16 game seasonWade Phillips and assistants (incl Mike Nolan, Ted Cottrell, Greg Manusky, Brian Stewart):________________________LILB_______________RILB_____________2006 Chargers 57 *1002005 Chargers 66 *1142005 Niners 66 *992004 Chargers 73 *1052004 Ravens 56 *1012003 Falcons 103 *1262003 Ravens 68 *1212002 Falcons 75 *1112002 Ravens *125 104Bill Parcells:________________________LILB_______________RILB______________2006 Cowboys 66 642005 Cowboys *73 55 Dom Capers/**** Lebeau:________________________LILB_______________RILB______________2006 Steelers *85 622005 Steelers 87 762005 Texans 83 822004 Steelers 68 532004 Texans 80 *982003 Steelers 94 812003 Texans 101 1072002 Steelers *55 372002 Texans 100 96Bill Belichick hybrid and assistants:________________________LILB_______________RILB______________2006 Browns *100 762006 Jets *99 672005 Browns 89 772004 Patriots 75 642003 Patriots 82 67Conclusions (so far):1. There seems to be absolutely NO QUESTION that an ILB can produce big tackle numbers, regardless of how good the defense or player in question is. While it appears that only the Phillips’ scheme will produce a stud ILB, I’d caution you to wait for the pre-2002 dataset to be completed. I’ll guarantee that you’ll like the stats for some of Parcells’ ILB from the mid-1990s.
2. The Phillips scheme favors the RILB (weak side ILB). Without getting into the scheme stuff, that makes sense when looking at a playbook from a team that runs the system. Assuming there aren’t other issues, i.e. the RILB sits on passing downs, this should be reproducible moving forward. I hope to further confirm this when I complete the pre-2002 database.
3. The traditional non-Phillips scheme (Parcells/Steeler) doesn’t appear to favor either the WILB or the SILB. I don’t trust this data set completely, however. Parcells has been quoted as saying the WILB is the guy he expects to make tackles in his 3-4 scheme so I suspect this group will get muddled when I include data from the Jets and Giants pre-2002. My guess today is that this group will prove to be team and player dependent given a quick look at playbook and coaching notes.
4. The Belichick hybrid data may favor the LILB/SILB. If the gamebooks are to be trusted, both Andra Davis and Tedy Bruschi have seen significant drops in solo tackle stats when moved from the LILB to the RILB in recent seasons. And I just can’t grasp any other way that Jonathan Vilma gets out-tackled by a pair like Eric Barton and Brad Kassell. I suspect that the LILB is the MLB whenever these teams run a 4-3 and that’s an important reason for the difference in most cases since Belichick comes from the same branch Parcells does and uses the same defensive line scheme. Still, given the fact that the data set for the Patriots over the past two seasons is very unreliable due to the frequent 4-3 in 2006 and the multiple injuries in 2005 and 2006 to the ILBs, I don’t know that I’m willing to come to a firm conclusion on this group yet.
5. The Gamebook data, as we know, are not trustworthy. Unfortunately, while I apparently had recognized that Donnie Edwards was the RILB in the past when researching old posts, I didn’t recognize that the 2006 Charger gamebooks listed Edwards as the RILB, whereas the 05 and 04 suggested he played to the strong side. Given the coaching comments with regard to the Edwards/Godfrey – Wilhelm/Cooper deal recently, it seems clear that Edwards was the Mike LB/WILB. More unfortunately, a lot of my “take the LILB” thoughts derived from Edwards’ stats and a reasonable conclusion that the strong side player would get more opportunity as teams run to that side more often. Lesson learned. Although I don’t know that it’ll end up being a killer mistake, apologies to those who rostered Stephen Cooper over Matt Wilhelm last year on that advice. If it’s any consolation, I’m heavily invested in the wrong stock too and I think the detailed research I’m doing now will put us on the correct track moving forward.
6. While I panned DonFue’s comment suggesting that you draft the better player as a default as not well reasoned enough, it’s clearly not a bad way to go after a closer look at the non-Phillips playbook. My early thought is that his recommendation to take the better player is legitimate if there are no coaching comments or nickel considerations in the non-Phillips teams. I’d still like to see more data on the Belichick teams, though, as that certainly wouldn’t have worked in the Vilma situation.
--------------------------------------------------------
Again, the PRIMARY IMPLICATION for 2007:
Patrick Willis has every chance to be a stud right now. The data set for the Phillips 3-4 over the past five seasons, playbook considerations and comments from the relevant coaches all point strongly to his having plenty of opportunity this season.
Finally, I’d prefer that your projectiles be tomatoes rather than stones.
If you don’t care about scheme issues or don’t think predicting which ILB in the 3-4 is the better IDP option is valuable information, stop reading now.
The upshot of the rest of this post for you is that there is no reason to believe that Patrick Willis cannot become an absolute stud in San Francisco, despite what some would have you believe about all 3-4 ILBs.For the rest of you who have closely followed my ramblings on the 3-4, it appears, after beginning to look through Gamebook data (correct and incorrect), playbooks, coaches’ quotes and discussion, and some old discussions I’ve had with the Guru himself (Norton), that I have been working under some faulty assumptions and conclusions. Some were correct anyway, some were not.
I’ve still got some research to do and am hoping to present a cleaner, clearer and hopefully useful article about the evolution of the 3-4 and its IDP implications later this summer. But, since there seem to be some clear trends that are different than what I’ve presented in the past, I felt I should post the short version here for those who have interest and begin to set the record straight.
So…the short version. And yes, as you’ll see this summer, this is the short version.
There are two major 3-4 defensive philosophies. Sparing you the historical and X and O details, one branch of the tree originated from Bum Phillips and was modernized by his son Wade, the other branch got its professional start with a couple of Patriot coaches who influenced Bill Parcells, Bill Belichick, and, indirectly, the Steeler coaches of the early 1990s.
Below is the raw data from 2002-present that I’ve collected from the FBG/Drinen database. There’ll be more data from before 2002 to be presented in the coming detailed summer article.
* denotes more than one tackle per game difference over 16 game seasonWade Phillips and assistants (incl Mike Nolan, Ted Cottrell, Greg Manusky, Brian Stewart):________________________LILB_______________RILB_____________2006 Chargers 57 *1002005 Chargers 66 *1142005 Niners 66 *992004 Chargers 73 *1052004 Ravens 56 *1012003 Falcons 103 *1262003 Ravens 68 *1212002 Falcons 75 *1112002 Ravens *125 104Bill Parcells:________________________LILB_______________RILB______________2006 Cowboys 66 642005 Cowboys *73 55 Dom Capers/**** Lebeau:________________________LILB_______________RILB______________2006 Steelers *85 622005 Steelers 87 762005 Texans 83 822004 Steelers 68 532004 Texans 80 *982003 Steelers 94 812003 Texans 101 1072002 Steelers *55 372002 Texans 100 96Bill Belichick hybrid and assistants:________________________LILB_______________RILB______________2006 Browns *100 762006 Jets *99 672005 Browns 89 772004 Patriots 75 642003 Patriots 82 67Conclusions (so far):1. There seems to be absolutely NO QUESTION that an ILB can produce big tackle numbers, regardless of how good the defense or player in question is. While it appears that only the Phillips’ scheme will produce a stud ILB, I’d caution you to wait for the pre-2002 dataset to be completed. I’ll guarantee that you’ll like the stats for some of Parcells’ ILB from the mid-1990s.
2. The Phillips scheme favors the RILB (weak side ILB). Without getting into the scheme stuff, that makes sense when looking at a playbook from a team that runs the system. Assuming there aren’t other issues, i.e. the RILB sits on passing downs, this should be reproducible moving forward. I hope to further confirm this when I complete the pre-2002 database.
3. The traditional non-Phillips scheme (Parcells/Steeler) doesn’t appear to favor either the WILB or the SILB. I don’t trust this data set completely, however. Parcells has been quoted as saying the WILB is the guy he expects to make tackles in his 3-4 scheme so I suspect this group will get muddled when I include data from the Jets and Giants pre-2002. My guess today is that this group will prove to be team and player dependent given a quick look at playbook and coaching notes.
4. The Belichick hybrid data may favor the LILB/SILB. If the gamebooks are to be trusted, both Andra Davis and Tedy Bruschi have seen significant drops in solo tackle stats when moved from the LILB to the RILB in recent seasons. And I just can’t grasp any other way that Jonathan Vilma gets out-tackled by a pair like Eric Barton and Brad Kassell. I suspect that the LILB is the MLB whenever these teams run a 4-3 and that’s an important reason for the difference in most cases since Belichick comes from the same branch Parcells does and uses the same defensive line scheme. Still, given the fact that the data set for the Patriots over the past two seasons is very unreliable due to the frequent 4-3 in 2006 and the multiple injuries in 2005 and 2006 to the ILBs, I don’t know that I’m willing to come to a firm conclusion on this group yet.
5. The Gamebook data, as we know, are not trustworthy. Unfortunately, while I apparently had recognized that Donnie Edwards was the RILB in the past when researching old posts, I didn’t recognize that the 2006 Charger gamebooks listed Edwards as the RILB, whereas the 05 and 04 suggested he played to the strong side. Given the coaching comments with regard to the Edwards/Godfrey – Wilhelm/Cooper deal recently, it seems clear that Edwards was the Mike LB/WILB. More unfortunately, a lot of my “take the LILB” thoughts derived from Edwards’ stats and a reasonable conclusion that the strong side player would get more opportunity as teams run to that side more often. Lesson learned. Although I don’t know that it’ll end up being a killer mistake, apologies to those who rostered Stephen Cooper over Matt Wilhelm last year on that advice. If it’s any consolation, I’m heavily invested in the wrong stock too and I think the detailed research I’m doing now will put us on the correct track moving forward.
6. While I panned DonFue’s comment suggesting that you draft the better player as a default as not well reasoned enough, it’s clearly not a bad way to go after a closer look at the non-Phillips playbook. My early thought is that his recommendation to take the better player is legitimate if there are no coaching comments or nickel considerations in the non-Phillips teams. I’d still like to see more data on the Belichick teams, though, as that certainly wouldn’t have worked in the Vilma situation.
--------------------------------------------------------
Again, the PRIMARY IMPLICATION for 2007:
Patrick Willis has every chance to be a stud right now. The data set for the Phillips 3-4 over the past five seasons, playbook considerations and comments from the relevant coaches all point strongly to his having plenty of opportunity this season.
Finally, I’d prefer that your projectiles be tomatoes rather than stones.
Last edited by a moderator:
Back in the late 80's in the Parcell/Belichick era, , Carl Banks put up nice numbers, I think he was the top IDP one year in our league, can't recall if he played inside or out now. Mills played ILB in New Orleans and scored well as did Pat Swilling & Ricky Jackson (OLB). Think Mora was the coach, and Wade Phillips was the DC.
