Slate/Deadspin Roundtable: Offensive Identities

Another contribution to the Slate/Deadspin NFL Roundtable. In it, I continue my rant against the inanity that is the concept of identity.

We're conditioned to think of a gain of just two or three yards as, well, a gain. But unless those couple yards are for a first down, it's anything but a gain. About 60 percent of all running plays are setbacks, meaning that the offense is less likely to score after the play than it was before. Offenses might as well be throwing downs away. Defenses would be happy to spot an offense a third-and-four every time—the conversion rate on third-and-four (55 percent) is lower than for a series that starts with a first-and-10 (67 percent).

You can read the post at either site: Slate or Deadspin.

  • 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

13 Responses to “Slate/Deadspin Roundtable: Offensive Identities”

  1. Anonymous says:

    Brian have done any work on the idea that you run to setup the pass? Was just curious about the results if you had.

  2. Anonymous says:

    I love it when Brian writes about the advantage of the pass. It just reminds me of the article in the "roundup" two weeks ago. The most pass happy offenses* win 68% of the games, while the "smash mouth" teams win only 30% of the games. Hope no one will find out soon, and that Martz keeps throwing even w/Hanie in the line up, fans screaming for the run :-)

    I believe what i wrote 2 or 3 years ago. You can throw on every down and win**. You just need the guts to withstand the screaming mainstream media.

    * adjusted for score, time etc. as i understood Chase Stuart

    ** unless Brian comes up w/some stats that the run REALLY sets up the pass, meaning play action passes are more effective than for example 3rd- and 8 ShG-Passes...

    Karl, Germany

  3. Anonymous says:

    P.S.:

    AFIR, the Avg. gain on 3rd and 8+ is about 6 (!!) yards, while the Y/PP is around 7. That would suggest that running on 3rd and long is a better idea than passing 90% of the time in such situations (b/c of higher Int-Ratio).

    In reality it just shows that when you pass all the time you really set up the ("long") run. I guess that´s why Faulk, Forte and others had only average seasons before Martz called the offense they played in.

    Faulk w/Martz: 4.8 Y/R
    Faulk w/o Martz: 3.8 Y/R

    Forte w/Martz: 4.7 Y/R
    Forte w/o Martz: 3.8 Y/R

    Karl, Germany

  4. Jonathan says:

    I heard on the radio here, they said "so in Baltimore, they've been talking all week about how they need to re-discover their identity and start running the ball!" I wonder if this talk has spurred our intrepid leader Mr. Burke into ranting against the "identity" narrative. Him being a Ravens fan and all.

  5. Jeff Fogle says:

    Brian,

    Have you ever zoomed back the lens a bit to look at the game theory stuff at the season level? So that...part of the consideration goes to a team keeping its quarterback as healthy for as long as possible (ideally the whole season).

    What's optimal for a particular down or a particular point in a game is influenced in a non-trivial way by a team's need to keep its starting quarterback in the lineup...so that skews the numbers in a way that makes sense for the goal of reaching, say, 10 wins in a 16-game season with your quarterback on the field.

    Passing all the time sounds great in theory...but quarterbacks would get annihilated. Passing more than we see now would yield more yardage and better in game math...until Jay Cutler breaks his thumb.

    Could this explain part of the gulf between what you're suggesting should be done in games, but what's actually done by NFL teams. They're factoring in a "big picture" need to keep their quarterback healthy, while you're looking at down, distance, and per-play math that's not reflecting this big picture consideration?

    Teams this week or last starting a QB who wasn't the original starter because there was an injury: Kansas City, Miami, Arizona, Houston, Chicago, Oakland, and Philadelphia.

    Teams who have seen QB's miss a game because of injury: Indianapolis, St. Louis, Seattle.

    Teams with QB's who have played despite reported injuries: Dallas, Pittsburgh, Detroit.

    I may have left someone out.

    If a team can expect a catastrophic injury for every x number of passes, isn't that a logical influence for run/pass splits? Given that running backs are easier to find than quality quarterbacks?

    Teams aren't just using the running game to keep defenses honest, and to run clock, and the other things you listed. They're also using the running game to keep their QB away from concussions, broken collarbones, separated shoulders, high ankle sprains, broken thumbs, etc...

    So, what we're seeing is the league trying to a run/pass/health balance that works best for a 16-game season, rather than a run/pass balance that works best within any one game in particular.

    Is that possible?

  6. BIP says:

    Jeff,

    If that were the case, shouldn't we expect to see passing rates soar during the playoffs?

  7. Jared Doom says:

    I haven't even read the post yet but THANK YOU. I am so disgusted every week about NFL pundits' fixation on a team's "identity".

  8. Sunrise089 says:

    @Jeff - solution: pay for league minimum RBs, pay for second low-end starter caliber QB like Campbell or Vince Young.

  9. Anonymous says:

    BIP,

    actually it does.

    Besides cold weather, all offensive numbers are up during the playoff history of the NFL.

    Offensive numbers, especially passing rates, dramatically increase in the SB´s itself (thanks to good weather too).

    Karl, Germany

  10. stevekirsch says:

    As a Steelers fan, I hear constant remarks that they need to get back to Jerome Bettis pounding running game, despite the fact that Big Ben's passing game has been what has gotten us 3 SB appearances in the last decade. A team's identity doesn't persist for years on end. It is defined dynamically based on personnel and league tendencies. I think all of this nonsense about teams getting back to their identity of years past is garbage.

    That said, I think there is some merit to teams executing a gameplan/philosophy based on organizational strengths and opposition weakness and adjust as necessary. However, as Brian has pointed out. This isn't what most talking heads refer to.

    Brian, have you (or anyone else) done any research on play sequencing? I'm interested in the interplay between running and passing in regards to how one sets up the other. Does the run really setup the pass? Can the pass really setup the run? To what extent? Following a successful pass, is the following run or pass more likely to be successful/more efficient? I would think that some alternating pattern of 70 runs and passes would be more effective than 35 runs followed by 35 passes.

  11. Brian Burke says:

    Jeff-Very interesting hypothesis. It's hard to imagine that the effect would be large even if true. Optimum balance only means replacing 10 runs per game with passes. That's about .5 extra sacks per game. But maybe. A healthy QB is no good to you in the playoffs...if your team's not in them.

    Jonathan-YES!

    And yes, EPA numbers suggest running is underused on 3rd and longs.

  12. Jeff Fogle says:

    Yup, probably not large...but perhaps enough to explain much of the difference between what looks "optimal" in the numbers and how NFL teams are actually doing things. If it's "only replacing 10 runs with passes," we're in a subtle area to begin with.

    Would point out that many QB injuries come on non-sacks...like the pinky or thumb injuries that come on attempted passes where the hand is hit, the bruised sternums or concussions that come on legal hits as a ball is being released, or the many possible negatives that can happen on scrambles where the QB is tackled downfield on a non-sack or just rips his achille's tendon while running. So, we can maybe estimate the QB runs plus sacks, tweak a little for pinkies banging helments and legal hits on throws, and the dangers seem more scary than .5 extra sacks per game would suggest...particularly when we're talking about 10 more passing plays per game over 16 games. That by itself may cross the threshold from an injury being kinda likely to very likely.

    Agree that keeping your QB healthy and not making the playoffs isn't much of a consolation prize. But, with 12 teams making the playoffs...and several more typically in contention in the last few weeks...the "we have a chance to make the playoffs if our QB stays healthy" mindset could easily take hold over a meaningful percentage of the league for a meaningful percentage of the season.

    And, even non-playoff teams don't want to lose their starting QB to injury just in terms of building for the future, or whatever. St. Louis needs Bradford to move up the learning curve. Carolina wants Cam Newton to keep improving. The number of teams who don't care if their starting QB gets hurt is limited, while the number of teams who do care represents a high percentage of the league.

    I think this is part of the mix, even if it doesn't explain the full gap between what you say is optimal and what the teams are doing.

    This is kinda like pitch counts in baseball. Pulling an elite starter at 115 pitches in a 2-2 game in the 7th inning isn't optimal for that game...but perhaps is optimal for a six month schedule because you don't want him to miss significant action. Pass counts instead of pitch counts. A stathead could easily make pulling the pitcher sound like a bonehead move because it reduces their win potential in that game. Are these managers morons?! Real world managers are thinking about the percentages of getting their teams to 90-95 wins, and they don't want to lose an ace for an extended period. Not a perfect match for the NFL...but there are some similarities in spirit I think. Big picture "seasonal" percentage issues are influencing in-game play calling in a way that doesn't look optimal for that game's win potential but does for keeping the QB healthy as long as possible.

  13. TOP says:

    Couple of things.

    Kind of nit-picky, but using run SR (or something akin to it), and comparing it to NY/A is a little apples and oranges, isn't it? Wouldn't Run SR (where you throw out most of the average skewing breakaway run) to Pass SR be an more apt comparison?

    Also, you're previous work for the Giants-Patriots Superbowl (http://www.advancednflstats.com/2008/02/super-bowl-xlii-and-team-possessions.html) makes me wonder about advocating lots of passing (high variance). More possessions leads to the better team winning more often. It seems like you want to play a ball control game if possible, limit possessions (and scoring) for both teams, and hope they make one more mistake than you (the very highly variable INT, Fumble, or special teams play). The "lose slow" strategy, as opposed to your "get ahead quick" strategy, which may also lead to a slow loss as the superior team asserts itself over the added possessions.

Leave a Reply

Note: Only a member of this blog may post a comment.