Which Teams Should Abandon the Run?

Yeah, yeah, yeah. It's a passing league. We got it. And still, according to the numbers, teams aren't passing enough. In the cases of some teams, it's painfully obvious that they should be passing more and running less. As a Ravens fan, I watched another game where nearly every run was simply a wasted down. Most of their paltry positive rushing yards seem to come from trash draw plays on long distances to gain, intended to mitigate very poor field position prior to a punt. It's like they're playing with two or three downs when everyone else gets four.

I wonder if, at some point, when an offense is so much better at passing than running, should it abandon the run almost altogether. On top of the general imbalance in the league, some teams are just throwing away downs when calling conventional run plays. Of course, running and passing generally play off of each other in a game-theory sense. To be successful, passing needs the threat of running, and vice versa. But sometimes, the cost of running is so high for some offenses, that it would be worth the trade-off to forfeit the unpredictability and just pass nearly every down.

It sounds crazy, but take a look at the Expected Points Added per play so far this season (through the 1pm games on Sunday 10/13). The right-most column is the pass-run split. The bigger that number, the greater the imbalance. Pay particular attention to the teams highlighted in red:


RankOffenseRun EPA/PPass EPA/PPass-Run Diff
1NO-0.180.350.53
2DEN0.030.480.45
3SD0.000.330.33
4BLT-0.250.080.33
5PIT-0.230.070.30
6ATL-0.050.230.28
7NE-0.150.080.23
8NYG-0.190.030.22
9DET-0.030.170.20
10SEA0.000.190.19
11MIA-0.110.080.19
12WAS-0.060.120.18
13GB0.050.220.17
14SL-0.100.060.16
15TEN-0.040.110.15
16DAL0.070.170.10
17CIN-0.030.070.10
18IND0.130.220.09
19SF0.000.070.07
20KC-0.050.010.06
21OAK-0.050.010.06
22ARZ-0.11-0.050.06
23CAR0.030.080.05
24TB-0.07-0.050.02
25MIN0.070.090.02
26NYJ-0.06-0.050.01
27CHI0.060.05-0.01
28PHI0.160.14-0.02
29CLV0.02-0.01-0.03
30BUF-0.02-0.06-0.04
31HST-0.01-0.09-0.08
32JAX-0.12-0.24-0.12

The teams highlighted in red are a combination of very poor at running and at least decent at passing. Despite the fact that the season is young and the numbers are bound to regress, I'd bet these teams would be better off passing nearly all the time. Of course, circumstances sometimes dictate running, such as milking the clock or short yardage situations.

Some teams with big splits aren't so bad at running, like DEN and SD. It's just that they're so much better passing the ball. So I wouldn't want to see them abandon the run.

Ultimately, while offenses are generally better off mixing up play types to keep defenses honest, I wonder if some offenses pay too high a price. The trade-off of being more predictable but throwing away fewer downs might be favorable.

  • 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

14 Responses to “Which Teams Should Abandon the Run?”

  1. Anonymous says:

    Yes, I was incredibly frustrated by the Ravens; attempts to run.

    My theory is that Baltimore fans love to point the Ravens' high winning percentage when Ray Rice gets over 20 carries, and say that the running causes the winning.

    The coaches then comply in an effort to keep their Jobs. The result is that the Ravens fall behind, meaning that they can't do those clock burning runs caused by winning, and don't get 20 runs.

    Cycle repeats.

  2. Paul Thomas says:

    I realize this is a quick hitter and not a detailed statistical piece, but there needs to be some serious regression on these numbers. Less than half a season's sample size is not much from which to draw conclusions as daring as "team X should stop using one of the two main play types in modern football."

    Put differently, I'd consider this tactic if the numbers were PREDICTABLY terrible, i.e. I came into the season thinking the run was a weak option, or my two best run-blocking linemen suffered season ending injuries, or something. But if I expect to have a decent run game and the numbers are way below what I looked for, I wouldn't adopt such a drastic tactic.

  3. George Chen says:

    To address the problem of game situations inflating the EPA/P of passes and deflating the EPA/P of runs (teams up late in games are both more likely to run and less likely to score while teams down by a lot are both more likely to pass and more likely to produce points on their drives due to prevent defenses), perhaps WPA/P could be used? I have no sense as to whether there will be sample size issues with using WPA/P only on this season however.

    Also, how are runs and passes defined? I'm assuming sacks are passes, but what about qb runs? Looking at the available play-by-play data provided by ANS, I guess scrambles could be isolated from designed qb runs, but please correct me if I'm wrong.

    I think the game theoretic implications of passing vs. running are going to be very difficult to isolate, but there should be little impact in the most extreme situations. I doubt teams facing the Saints, Packers, or Broncos are overly concerned with whether they exhibit any run-pass balance.

  4. Anonymous says:

    I read it not as absolutes but a push toward better play calling. If I recalled from a past article, teams should be passing more then half of the time (about 57%) until the defense starts to modify their schema to make passing less profitable. In this case instead of going into the game with the mindset of running about half of the time, maybe moving toward 10-20% of the time (Saints fan here.) Once the defense starts backing into passing defenses and making passes less profitable, the run will open up and should be tried. (Use the pass to setup the run.)

  5. Anonymous says:

    Balancing the run/pass ratio to maximize EPA/P may in theory be the optimal solution, but it would be very challenging to implement during a game. EPA just isn’t obvious enough after each play to keep a mental note of and adjust play calling accordingly.

    A more practical way for coaches to improve their run/pass balance would be to use success rate. It is much more intuitive for a coach to assess if a given play was a success or not. If passing has been more successful than running, call a pass play and vice versa.

  6. willkoky says:

    What about giving the defense some rest? They always used to show stats like, "you win 75% of the time if you get a 100 yard rusher". Which people assumed was a bias because if you were ahead, you ran. But maybe that wasn't the full story? Maybe you won more often running because it helped you play better defense? I wonder if you could study the winning percentage of teams that ran effectively through the first 3 quarters vs teams the passed effectively through the first three quarters. Being a Packer fan, watching their almost undefeated season was hard because their defense was on the field so much and always looked tired. Also, bad teams might want to run simply to shorten a game they have little chance of winning if lots of plays are run by both teams. A bad running team may be adding to its chance of winning by that method as well, even though their EPA goes down.

  7. Elliot says:

    I think the main problem with run criticism is that it doesn't even attempt to estimate the value of the reduction in variance offered by running the ball. I don't know offhand how to do it, necessarily; obviously you want to reduce variance when ahead late, but I suspect it has value otherwise also, and that it's the main reason, conscious or unconscious, that teams run so much.

  8. Brian Burke says:

    Elliot--I agree. I *know* it has value beyond burning clock when ahead. But is that value enough to justify the essential forfeit of downs by futile run attempts? I'm just asking the question (with the hunch that for some teams with extremely poor rushing, the value doesn't justify running).

    willkoky-Regarding the defensive rest point...I agree that it's not a good thing for the defense to be fatigued. But keep in mind the difference between the game clock and actual chronological duration. Running 3 times and punting doesn't rest the defense any longer than passing 3 times and punting. The way to rest a defense is to convert 1st downs in the best way possible, which means (for some teams) not throwing away downs with runs for a yard to two.

  9. Robert Burns says:

    I think you pinpointed the flaw in your analysis when you said "Of course, running and passing generally play off of each other in a game-theory sense. To be successful, passing needs the threat of running, and vice versa."

    I follow the Giants, one of your teams in red. They run a vertical passing game based on play action. Without a credible running game and reasonable down and distances, the Giant offense fails. They face a lot of third and tens, which are hard to convert. If and when their running game earns any respect, then their over all offense will greatly improve.

    Why does a team have a good (or bad) running game? If the problem lies with the offensive line, then how good is the pass blocking?

    It is note worthy that a team does not play to put up good stats, they play to win. 2 of the six red teams are 5 and 1.

  10. Brian Burke says:

    "Without a *credible* running game..." (emphasis mine). The Giants don't have a credible running game, so why bother?

    You know how to avoid lots of 3rd and 10s? Stop running into brick walls on 1st and 2nd down. Convert by passing on 1st or 2nd down.

    And thank you, Coach Edwards, on the lecture about teams playing to win.

  11. Alan says:
    This comment has been removed by the author.
  12. Alan says:

    I think we can't even begin to guess what happens when the game-theory element is set aside. It's all speculation until some team tries it (and even then, it would take defenses and their coordinators time to adjust); but my personal hunch as a fairly expert poker player is that you just can't be that one-dimensional against good competition.

    I would go so far as to say that even if a team averaged 0 yards per game rushing, I suspect it would be a game-theoretical mistake to stop attempting to run altogether, as the losses in their passing game would likely be greater than what they gain by avoiding brick-wall runs. But who knows.

  13. Anonymous says:

    Wow that run EPA for Philly has to be one of the highest ever right?

  14. Anonymous says:

    "To be successful, passing needs the threat of running, and vice versa."
    running needs the threat of passing because blockers must overcome the inertia of defenders in the run game while inertia can be harnessed by a passing attack; and passing doesn't need the threat of running because the number of people in coverage is self limiting by the need to get pressure on a qb.

Leave a Reply