by Hollywood Sports
Saturday, Oct 31, 2020Seattle Seahawks fans, as well as the “you already lost if you did not pass on first down” football analytics crowd, have been vocal with their claims regarding how good that team would be if they passed the ball more in past seasons. This sentiment has evolved into the catchphrase “Let Russ Cook.”
Yet when these critics call on the “old-fashioned” Seahawks' offense to run less and pass more (as offensive coordinator Brian Schottenheimer did in the first two games of the 2018 season before head coach Pete Carroll intervened), they typically fail to address Carroll's reasons for wanting to run the ball more. The direct and indirect benefits of running the football are under-appreciated by the football analytics community. Their arguments would strengthen if they better engaged with the rationale of (Super Bowl-winning) head coaches who find subtle advantages in running the football that transcends the yards-per-carry metric.
First, “Russ can't cook” if Russ is on the sidelines with an injury. Carroll is very cognizant of the number of hits made on the quarterback. He has this number tracked by his coaching staff. Carroll speaks of "not taking the sugar" when it comes to the short-term allure of relying on Wilson to make yet another pass. Every drop back risks another quarterback hit.
Have there been studies in the analytic community regarding the correlation between the number of QB hits and injuries? I have not seen any (and I pay attention). These stats are not cited in the "you already lost if you ran on first down" genre of analytics.
I would be surprised if this area was not being studied by internal analytics departments. Is there is a threshold where QB hits correlate with a higher risk of injury (like the 400 carry threshold pointing to RB regression the next season)? That seems to be a fascinating subject to investigate.
Second, Carroll thinks his team has a better chance of winning close games if he can manage the game to put his coaching staff and Wilson in that position. This belief runs counter to the conventional wisdom in the analytics community that winning close games is random.
In general, that conventional wisdom makes sense: close games tend to be decided on a small number of plays (or decisions including by the refs) that would seem to even out over time. However, there is some interesting work being done in basketball suggesting winning close games can be a skill.
Certainly, the eye-test watching Wilson, Aaron Rodgers, Tom Brady, Patrick Mahomes, et al, supports the notion that "you don't want those QBs to have the ball last." Do these QBs behave differently in crunch time? For Wilson, that becomes the time he is allowed to "cook."
What is fascinating about Seattle is that they began a rebuild for the '18-19 season after missing the playoffs. Rather than eating a couple of losing seasons, they made the playoffs. One would think that Carroll would get credit for "reloading" on the fly.
Instead, implicit in the "Let Russ Cook" argument is that this team was closer to winning a Super Bowl these last two seasons than suffering 6-10 records. In college basketball, Carroll's tactics would elevate him to the genius level in the conventional wisdom of that sport.
Now in Year Three of the rebuild, I was of the belief that Seattle would pass more this season — just as they did when they were making Super Bowl runs with Wilson (with a better defense and overall roster). I await the 538 dot com article where credit is taken for Seattle’s shift in tactics.
I'm not a Carroll stan. I have issues with his approach to the offensive line. Rather, I am a stan for answering arguments -- both explicit and implicit. Fortunately, the misguided conventional wisdom on the Seahawks has contributed to point spread value in the last two seasons. So that has been good!
As we now approach the halfway point of the NFL season, what if one of the unintended consequences of "Let Russ Cook" was a decline in play by the Seattle Seahawks defense? More early-down passing shortens Seattle's time of possession on drives. Consider these defensive numbers:
2020 Seattle Defense (after Week Six): 27.0 PPG, 471.2 total YPG. 6.4 Yards-Per-Play allowed. 2020 Offense Average Time of Possession: 28:21.
2019 Seattle Defense: 24.9 PPG, 381.7 total YPG. 6.2 Yards-Per-Play allowed 2019 Offense Average Time of Possession: 31:26.
Last year, the Seahawks had a 30/32 average run-to-pass play ratio during the regular season for a 51.6% pass rate per offensive snap. Now in the Let Russ Cook era, Seattle enters Week Eight of the NFL season with an average run-to-pass play ration of 25/36 fora 59.0% pass rate per snap offensive snap.
Some defensive coaches claim that their players only have about 50 plays in them per game. When defensive players go beyond that point, then their energy level begins to decline. Of course, these defensive coaches can not code R to save their lives.
Was "establishing the run" Carroll's method to elevate a mediocre defense? A closing thought from two-time Super Bowl champ (but failed R coder) Jimmie Johnson:
"How you protect a defense is you eliminate the negative plays, and you increase your time of possession by running the football."
Best of luck for us — Frank.
Yet when these critics call on the “old-fashioned” Seahawks' offense to run less and pass more (as offensive coordinator Brian Schottenheimer did in the first two games of the 2018 season before head coach Pete Carroll intervened), they typically fail to address Carroll's reasons for wanting to run the ball more. The direct and indirect benefits of running the football are under-appreciated by the football analytics community. Their arguments would strengthen if they better engaged with the rationale of (Super Bowl-winning) head coaches who find subtle advantages in running the football that transcends the yards-per-carry metric.
First, “Russ can't cook” if Russ is on the sidelines with an injury. Carroll is very cognizant of the number of hits made on the quarterback. He has this number tracked by his coaching staff. Carroll speaks of "not taking the sugar" when it comes to the short-term allure of relying on Wilson to make yet another pass. Every drop back risks another quarterback hit.
Have there been studies in the analytic community regarding the correlation between the number of QB hits and injuries? I have not seen any (and I pay attention). These stats are not cited in the "you already lost if you ran on first down" genre of analytics.
I would be surprised if this area was not being studied by internal analytics departments. Is there is a threshold where QB hits correlate with a higher risk of injury (like the 400 carry threshold pointing to RB regression the next season)? That seems to be a fascinating subject to investigate.
Second, Carroll thinks his team has a better chance of winning close games if he can manage the game to put his coaching staff and Wilson in that position. This belief runs counter to the conventional wisdom in the analytics community that winning close games is random.
In general, that conventional wisdom makes sense: close games tend to be decided on a small number of plays (or decisions including by the refs) that would seem to even out over time. However, there is some interesting work being done in basketball suggesting winning close games can be a skill.
Certainly, the eye-test watching Wilson, Aaron Rodgers, Tom Brady, Patrick Mahomes, et al, supports the notion that "you don't want those QBs to have the ball last." Do these QBs behave differently in crunch time? For Wilson, that becomes the time he is allowed to "cook."
What is fascinating about Seattle is that they began a rebuild for the '18-19 season after missing the playoffs. Rather than eating a couple of losing seasons, they made the playoffs. One would think that Carroll would get credit for "reloading" on the fly.
Instead, implicit in the "Let Russ Cook" argument is that this team was closer to winning a Super Bowl these last two seasons than suffering 6-10 records. In college basketball, Carroll's tactics would elevate him to the genius level in the conventional wisdom of that sport.
Now in Year Three of the rebuild, I was of the belief that Seattle would pass more this season — just as they did when they were making Super Bowl runs with Wilson (with a better defense and overall roster). I await the 538 dot com article where credit is taken for Seattle’s shift in tactics.
I'm not a Carroll stan. I have issues with his approach to the offensive line. Rather, I am a stan for answering arguments -- both explicit and implicit. Fortunately, the misguided conventional wisdom on the Seahawks has contributed to point spread value in the last two seasons. So that has been good!
As we now approach the halfway point of the NFL season, what if one of the unintended consequences of "Let Russ Cook" was a decline in play by the Seattle Seahawks defense? More early-down passing shortens Seattle's time of possession on drives. Consider these defensive numbers:
2020 Seattle Defense (after Week Six): 27.0 PPG, 471.2 total YPG. 6.4 Yards-Per-Play allowed. 2020 Offense Average Time of Possession: 28:21.
2019 Seattle Defense: 24.9 PPG, 381.7 total YPG. 6.2 Yards-Per-Play allowed 2019 Offense Average Time of Possession: 31:26.
Last year, the Seahawks had a 30/32 average run-to-pass play ratio during the regular season for a 51.6% pass rate per offensive snap. Now in the Let Russ Cook era, Seattle enters Week Eight of the NFL season with an average run-to-pass play ration of 25/36 fora 59.0% pass rate per snap offensive snap.
Some defensive coaches claim that their players only have about 50 plays in them per game. When defensive players go beyond that point, then their energy level begins to decline. Of course, these defensive coaches can not code R to save their lives.
Was "establishing the run" Carroll's method to elevate a mediocre defense? A closing thought from two-time Super Bowl champ (but failed R coder) Jimmie Johnson:
"How you protect a defense is you eliminate the negative plays, and you increase your time of possession by running the football."
Best of luck for us — Frank.
