There has been a fair amount of discussion on this point in the Brady/Manning XVII thread, and I thought it might be worth a separate thread.
As we all know, the Patriots have (in the past year or so) shown a willingness to almost completely forgo the run game when they feel it is favorable to do so, primarily against defenses with superb offensive lines. The standout examples are the second half of last year's divisional matchup with Baltimore, weeks two and seven against the Bills and Jets respectively, and last Saturday's matchup with the Chiefs. In those (effectively) 3.5 games, the Patriots scored 118 points and accumulated 1,434 yards in 37 drives (excluding three which were entirely kneel downs). 3.19 points per drive would have led the league by a considerable margin in each of the past two seasons, while 38.76 yards per drive would have been 3rd in 2014 and narrowly edged out Arizona (38.62) for first place in 2015. A weighted (14:86) average of the Patriots offenses in the past two seasons would have predicted averages of 2.47 points and 33.16 yards per drive.
Various sources also pointed out that the Patriots in the Brady era are 6-1 when they drop back to pass on 80% or more of their snaps, while the rest of the league is 6-137.
Clearly, the Patriots have succeeded with this approach in the past. But is it sustainable, could it be effective against other defenses, and can other teams succeed with a similar approach?
At the most basic level, this seems like an obvious development. Passing has been the more lucrative approach than rushing in terms of yards for decades, and the gap continues to expand. Other (better) efficiency metrics, such as EPA, WPA, DVOA, and SR show less pronounced but similar gaps. In theory, teams should be passing more on early downs in order to reach the Nash equilibrium.
Of course, reality is not nearly that simple. Personnel differences between teams, on top of the wide range of coverages and plays used by NFL defenses, make it much more difficult to predict how changing playcall distributions would affect efficiency. It is almost universally accepted that some measure of balance is necessary to prevent the defense from keying in on and shutting down one type of play. Even Belichick believes this to some degree.
Now I am not an Xs and Os guy by any means, but my understanding is that there is no coverage which will consistently stop the pass at the expense of the run. Teams can load the box to stop the run game, play press man to disrupt quick passes and screens, blitz to pressure the quarterback, drop extra players in coverage so receivers take longer to get open, or play one of a wide variety of zones that are particularly strong against certain routes, concepts, and locations on the field, but weak against others. All of these are strong against some pass plays, and almost all of them are relatively weak against the run, but they are also vulnerable to different types of pass plays.
It would logically follow that if a team can work a sufficient variety of plays and concepts out of the same handful of formations, they wouldn't need to lean on the run game as a counter to the coverages their opponent uses to stop their passing game. Instead, the run game becomes just another subset of plays, very useful for some situations (short yardage, goal line, and running down the clock) but not much different than the screen game or the triangle stretch concept.
Practice time could be a significant limitation for a team looking to implement such an offense, as they need to be able to execute a wide range of different plays effectively without leaning on certain concepts heavily enough to be predictable. Personnel is arguably even bigger, as nothing is going to work without a quarterback who can make the reads and hit the throws to receivers who can get open. Fatigue could be another consideration: a QB with a sore arm, or an OL worn out from spending most of the game going backward, would also make this approach far less effective.
To the people with a better understanding of concepts and coverages than I, are the Patriots using formations/plays/concepts designed to be unpredictable and give them good options regardless of the defensive coverages? Alternately, are they just relying on the receivers to outperform their defensive backs against unfavorable coverages? Is there anything schematically that defenses could be doing to stop this approach? Finally, does the run game offer any noteworthy benefits not covered here?
As we all know, the Patriots have (in the past year or so) shown a willingness to almost completely forgo the run game when they feel it is favorable to do so, primarily against defenses with superb offensive lines. The standout examples are the second half of last year's divisional matchup with Baltimore, weeks two and seven against the Bills and Jets respectively, and last Saturday's matchup with the Chiefs. In those (effectively) 3.5 games, the Patriots scored 118 points and accumulated 1,434 yards in 37 drives (excluding three which were entirely kneel downs). 3.19 points per drive would have led the league by a considerable margin in each of the past two seasons, while 38.76 yards per drive would have been 3rd in 2014 and narrowly edged out Arizona (38.62) for first place in 2015. A weighted (14:86) average of the Patriots offenses in the past two seasons would have predicted averages of 2.47 points and 33.16 yards per drive.
Various sources also pointed out that the Patriots in the Brady era are 6-1 when they drop back to pass on 80% or more of their snaps, while the rest of the league is 6-137.
Clearly, the Patriots have succeeded with this approach in the past. But is it sustainable, could it be effective against other defenses, and can other teams succeed with a similar approach?
At the most basic level, this seems like an obvious development. Passing has been the more lucrative approach than rushing in terms of yards for decades, and the gap continues to expand. Other (better) efficiency metrics, such as EPA, WPA, DVOA, and SR show less pronounced but similar gaps. In theory, teams should be passing more on early downs in order to reach the Nash equilibrium.
Of course, reality is not nearly that simple. Personnel differences between teams, on top of the wide range of coverages and plays used by NFL defenses, make it much more difficult to predict how changing playcall distributions would affect efficiency. It is almost universally accepted that some measure of balance is necessary to prevent the defense from keying in on and shutting down one type of play. Even Belichick believes this to some degree.
Now I am not an Xs and Os guy by any means, but my understanding is that there is no coverage which will consistently stop the pass at the expense of the run. Teams can load the box to stop the run game, play press man to disrupt quick passes and screens, blitz to pressure the quarterback, drop extra players in coverage so receivers take longer to get open, or play one of a wide variety of zones that are particularly strong against certain routes, concepts, and locations on the field, but weak against others. All of these are strong against some pass plays, and almost all of them are relatively weak against the run, but they are also vulnerable to different types of pass plays.
It would logically follow that if a team can work a sufficient variety of plays and concepts out of the same handful of formations, they wouldn't need to lean on the run game as a counter to the coverages their opponent uses to stop their passing game. Instead, the run game becomes just another subset of plays, very useful for some situations (short yardage, goal line, and running down the clock) but not much different than the screen game or the triangle stretch concept.
Practice time could be a significant limitation for a team looking to implement such an offense, as they need to be able to execute a wide range of different plays effectively without leaning on certain concepts heavily enough to be predictable. Personnel is arguably even bigger, as nothing is going to work without a quarterback who can make the reads and hit the throws to receivers who can get open. Fatigue could be another consideration: a QB with a sore arm, or an OL worn out from spending most of the game going backward, would also make this approach far less effective.
To the people with a better understanding of concepts and coverages than I, are the Patriots using formations/plays/concepts designed to be unpredictable and give them good options regardless of the defensive coverages? Alternately, are they just relying on the receivers to outperform their defensive backs against unfavorable coverages? Is there anything schematically that defenses could be doing to stop this approach? Finally, does the run game offer any noteworthy benefits not covered here?