Exploring the Causes of a Sack Pt. 1

A guest post by David Giller. Born and raised in Swampscott, MA, David attended Vanderbilt University where he was the starting longsnapper for the Commodores. He graduated summa cum laude with a degree in economics/corporate finance. David currently works as a business analyst for a Bain Capital Ventures portfolio company.

I would first like to thank Brian for his suggestion to post my study here in an effort to spark some interesting conversation and obtain some valuable takeaways. My post contains the results of a recent study I put together which focuses on the causes of sacks in NFL games. Although it is fairly detailed, I believe there are still areas of further development, some of which have been explored in an appendix to this initial study and will be coming in the second installment of this post.

As a disclaimer, the number of sacks were provided from an official source; however, the timing of the sacks, count of offensive blockers/defensive rushers was determined from my individual film study.The full piece is attached in a link, however, I have included some highlights below.



* The red portion of the trendline represents one standard deviation from the mean, the black dotted line represents the mean of the data set, and the orange dotted line represents the mean drop back and throw time of an NFL QB.

Glossary
Leveraged (L): Offensive blockers outnumber defensive rushers
Neutral (N): The number of offensive blockers equals the number of defensive rushers
Non-leveraged (NL): Defensive rushers outnumber offensive blockers

Conclusions
• A sack is attributed mainly to a quarterback taking too long to pass the ball and his offensive blockers getting beat physically by the defender.
o The average time from the snap until a sack is initiated (4.3 sec) is far greater than the average drop and pass time of an NFL quarterback (2.7 sec).
o Of the sacks that occurred in the 2011 season, 77% were in situations where an offensive blocker was beaten physically

• The offense was leveraged and offensive tackle was physically outmatched on 37% of sacks. This demonstrates that the offense, specifically the offensive tackle, was simply overpowered, compounding the negative effects of the quarterback taking too much time to pass the ball. This data suggests that perhaps the team should potentially look for ways to aid the offensive tackle in pass protection.
o One potential way to help in pass protection would be to devote more players to blocking. However, given the data, leveraged offenses are obviously no less apt to give up a sack (84% of sacks occurred when the offense was leveraged). Therefore, perhaps dedicating additional blockers to stop a defense is a waste of resources. It might be more beneficial for the offense to have more targets for the quarterback to pass the ball to. This larger number of potential targets would hopefully lower the time the quarterback holds onto the ball and negate poor offensive tackle play by increasing the probability that an offensive player will be open to pass to in a shorter amount of time, thus limiting the team’s sack count.

• Beyond reducing the time the quarterback holds onto the ball, the other option is to look for scenarios where the average time until a sack occurs is significantly higher. Plays that involve play action seem to offer the quarterback a little more time to pass the ball (on average about .3 extra seconds). However, given that there is not a significant difference in the average time until a sack occurs in each scenario observed in this study (the majority of sacks occurred around 4.3 sec), it would be interesting to find a scenario where the quarterback has more time until a sack is initiated.

• Unfortunately, the reason for the quarterback taking too long to pass the ball is unclear as the data does not show whether or not the extra amount of time taken was due to the superior coverage of the defense downfield or the indecisiveness of the quarterback (a subjective opinion).

• In situations where there is a non-traditional blocking scheme, there are almost twice as many sacks when facing a non-traditional rush than a traditional rush, regardless of whether the offense is leveraged, neutral, or non-leveraged (89% difference). It is unclear what causes this discrepancy, but it would be interesting to find out why so the team can plan accordingly.

Link to complete study

  • Spread The Love
  • Digg This Post
  • Tweet This Post
  • Stumble This Post
  • Submit This Post To Delicious
  • Submit This Post To Reddit
  • Submit This Post To Mixx

11 Responses to “Exploring the Causes of a Sack Pt. 1”

  1. Anonymous says:

    Not sure if this was addressed or if the data collected were only plays where sacks happened (as it appears), but you'd think it would be more helpful to look at success rate (or those pessimistic, failure rate) of the metrics listed, instead of just the raw number.

    This would apply to play action, chips, non traditional blockers, and pass rushers.

    Additionally, splitting the data into number of linemen rushing and then additional linebackers, or non-linemen rushers. In which case you could break down 3-4 and 4-3 success along with nickel, dime, or other packages.

    I'm sure you guys already have a ton of suggestions and know what's more possible.

  2. Robert Burns says:

    Wow. You did a lot of work, and I really appreciate the work and effort.

    Analysis of game tapes is difficult and many of the calls are subjective.

    I notice that you earmarked “C” for coverage sacks, and then none of tables showed any coverage sacks. An offensive lineman can’t control a defensive player forever. There is a point in time (say 5 seconds or more) where an offensive player is beaten, but the sack is not his fault. That is a coverage sack.

    There is also the problem of a defensive tackle pushing the guard into the pocket so that the QB can not step up, so the offensive tackle does can’t ride the defensive end behind the QB. Who was beaten… the guard or the tackle, or both??

    There is no analysis of the type of offense. There are so many offenses in use that the type of offense may influence the number of sacks and the sack rate. By type, I mean the offense may be a pistol, a west coast, a shotgun, a two step drop, etc.

    Lastly, there is no data on game situations….down and distance, score, field position, and time remaining in the game. These factors greatly influence both the offensive and defensive play calling, which in turn influence sacks.

  3. David Giller says:

    Hi Robert,

    Thanks for your comments.

    I shied away from coverage sack designations given that the film I was watching sometimes made it difficult to determine if a player truly was covered due to film quality/extent of what was being captured on film. Also, at the end of the day, it was just an extremely subjective category to place things into. So there are most likely several coverage sacks designated as another category. I just felt I would rather have miscategorization as a result of an effort to eliminate subjectivity, as opposed to miscategorization as a result of trying to see something in the data/film that might not actually be there.

    Your suggestion about various offensive styles and systems is very interesting to me. As I mentioned in the "Next Steps" section I think there could definitely be some value in expanding on the play action/non play action analysis to include the length of QB drop and shotgun vs. under center. Ultimately, I would see the value add as having relative benchmarks for when a QB must release the ball (ie. if the team is running a pass play from under center with a 3 step drop and the longest route either in yards/average time to complete the route is X, then the QB will have to release the ball in Y sec to have a Z% chance of avoiding a sack)...essentially add relativity to the internal clock QB's are often praised or ridiculed on.

    To your last point, the next installment of this analysis with be a visualization heavy, deep dive into situational outcomes based on down an distance/game clock so stay tuned.

    Thanks again,
    -Dave

  4. Robert Burns says:

    Hi Dave,

    thanks for the quick reply.

    On coverage sacks, I think there is a difference between no one open (your definition) and the quarter back can't find anyone to throw to (my definition). I find it hard to imagine that a play is designed to have the QB throw to the primary receiver 5 seconds or more into the play.

    I have seen balls throw to people who appear covered and the receiver comes down with the ball. I have also seen balls thrown out of bounds to avoid a sack or an INT. That is a QB function, and that is why they get the big bucks. What I am saying is that some of the blame for a sack rests with the QB.

    Is it possible to break out your table shown on page 4 by team?

    Looking forward to the future parts of your study.

    Bob

  5. Anonymous says:

    great stuff, thanks. very interesting.

    I find the graph a bit puzzling. Are there really ~2 sacks at 3.4 seconds, and ~35 sacks at 3.5 seconds? That seems wrong. And it wouldn't even be a problem with bin size; one would expect a smoother transition.

  6. Anonymous says:

    Interesting analysis.

    Without having read all the study I am curious if busted plays, muffed snaps, etc. account for those sacks 2 seconds and under. I guess sack types need to be qualified in some respects is my point.

  7. Anonymous says:

    Along with the wildly varying results from one time bin to the next, this analysis obviously needs to report the results in terms of the percentage of plays, and not just the total number of sacks, for a given scenario. No offense, but the information is not useful as presented.

    The comments at the end of the paper (improvements/future work) really needed to be completed before releasing it.


    I'd also question the use of a mean when comparing these asymmetric distributions, I'd also like to see the median.

    It would be interesting to see the distribution of the qb release time as well.

  8. Scott Allen says:

    Interesting work. I didn't see it in your detailed analysis, but did you do any comparison by team? Which teams had sacks the fastest, or which area on teams were weakest allowing most sacks? I think my detailing those areas would be interesting as teams we as fans and teams themselves could pin point the exact area that is weakest from a statistical standpoint. Moreover, if that kind of statistic was generated over say 5 years, you could really pinpoint where offensive lines are the weakest.

  9. naturalnoble says:

    I don't think it is appropriate to conclude that leveraged offenses are no less apt to give up a sack. 84% of sacks occur vs a leveraged offence, and that is probably because on >90% of plays in the nfl the offense is leveraged, not because the extra blockers are useless.

  10. Darcy Phillips says:

    Great study. Down and distance, time left in the game, and score all contribute to a quarterback's decision making including when the ball comes out of his hand. These factors also apply to everyone else on the field including defensive lines and linebackers. Watch a defensive lineman's effort throughout a game. It is amazing that coaches do not substitute these players more often!
    I agree the quarterback does play the biggest role in determining sacks. Jay Cutler comes to mind. For years, Cutler did not drop back traditionally the way quarterbacks are taught. Instead he backpedalled (ala Dan Marino) leaving himself unable to step up in the pocket and an easy target for defenses. He contributed to many of his own sacks based on his style of play.

  11. Dave O says:

    Isn't one reason leveraged offenses are no less likely to give up a sack is because teams are more inclined to keep extra blockers in on plays with deep drops and slow-developing routes?

Leave a Reply