Fumble Rates by Play Type

A lot of analysis of running vs. passing takes into account the added risk of interceptions. Almost all sources of passing stats will include team or player interceptions. But fumbles are more tricky. Stat sites will tell you team fumbles, but they usually won't tell you how many were due to rushes or due to passes.

Unlike interceptions, fumbles can happen on either type of play. But is the risk of a fumble even between runs and passes, or are fumbles more likely to occur on one or the other type of play? Further, what about fumbles lost? Are fumbles more likely to be lost on runs or passes? And how does the sack-fumble factor in?

Fortunately, with play-by-play data we can tell whether the play was a run or a pass, or if the fumble was lost. And for pass plays, we can also tell whether it was a sack that caused the fumble or whether the receiver dropped the ball after a completion.

Here are the numbers based on all non-preseason games from 2000 through week 10 of 2009. I'm just going to give you a bullet list of the relevant percentages:

  • Fumbles occur on 1.67% of all run and pass plays. About half of all fumbles are lost, for an overall rate of 0.83% per play.
  • Fumbles occur on 1.16% of run plays. 55% of run play fumbles are lost, for an overall rate of 0.65% per run.
  • Fumbles occur on 2.04% of pass plays. 47% of pass play fumbles are lost, for an overall rate of 0.97% per pass play.
  • Fumbles occur on 18.0% of all sacks.
  • Fumbles occur on 0.96% of all pass completions.
  • Of all pass play fumbles, 56% occur due to sacks.
  • Sack-fumbles are lost 47% of the time.
  • 8.5% of sacks cause a turnover.
  • Fumbles after completions are also lost 47% of the time.
Some of this is just trivia, but I think the most pertinent numbers here are the relative likelihood of fumbles lost. They are about half again more likely on pass plays than on run plays (0.97% vs 0.65%).

And in case anyone is wondering, yes, my recent series of articles on run-pass balance do factor in fumbles. Expected Points Added (EPA) includes anything that can happen on a play--turnovers, sacks, penalties--anything that moves the ball or changes possession.

Hat tip to the Numbers Guy, Carl Bialik for asking the question.

  • 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

12 Responses to “Fumble Rates by Play Type”

  1. Anonymous says:

    Raises some questions...
    Did you filter kneel downs out of the run plays ?
    Are run fumbles related to down or yards to go ?

  2. Brian Burke says:

    Yes, QB kneel downs are removed. And good question about run fumbles and down-distance. Don't know, but I'll investigate.

  3. Eiad says:

    Would it be possible to do fumble statistics on special teams plays?

  4. billsfan says:

    First, a bookkeeping question: how are you counting aborted snaps? I'd guess as QB runs.

    Football Outsiders' research suggests that, since fumble recovery rates (for a single team) vary highly from season to season, fumble recovery isn't a "skill" (they call it "fumble luck," which makes sense, given the unpredictability of a bouncing oblong spheroid). It's very interesting that the fumble loss rates on run (55%) and pass (47%) plays are close, but not that close, over a decade. Do the single-season numbers vary drastically from the decade averages, i.e. are there seasons in which the fumble loss rate is higher on pass plays? Any thoughts as to *why* there is a seemingly significant difference in the loss rates?

    It would be interesting to classify fumbles further by position, to see if there were a higher fumble rate on QB runs, for example, or what kind of receiver fumbles most often.

  5. Anonymous says:

    I wonder if more fumbles are "recovered" on run plays because the ball goes out of bounds? Typically the QB will be sacked in the middle of the field so that will almost never happen on sack-fumbles.

  6. Ed Anthony says:

    Typically on a run play there are an equal number of players from each team around the ball. By this I mean that on a run the offense has 5 or 6 players in the area of the ball. The defense has a similar number of players although it may be 4 or 5. On a fumble that gives each team a 50/50 chance of recovering the ball. On a pass play there may be 2 or 3 defenders in the area of the ball and usually only 1 player on offense - the receiver. Hence I would expect that the defense has a greater chance of recovering the fumble. I appreciate that as a pass play develops the numbers may change but seldom is a pass completed in a situation where we have more receivers than defenders.

    This may not explain completely why pass fumbles are recovered with less frequency by the offense but I think it shows a trend.

  7. James says:

    I agree with Ed as to passing fumble recovery rates. That's the first thing I thought of: runningbacks usually fumble the ball by the offensive linemen, receivers fumble the ball near the secondary.

  8. Chase says:

    Very interesting stuff.

  9. notque says:

    Great work. I run a simulated football game called http://deeproute.com and will be using these numbers for rates.


  10. Anonymous says:

    Im suprised the sack-fumble lost percentage is so low. I'd think the defense would have the better shot at it

  11. JJ says:

    I suspect the reason the sack-fumble lost percentage isn't higher, Anonymous, is because the 'D' is busy sacking while the offensive lineman who allowed the sack is standing free when the ball comes out. I say that only half in jest.

  12. Anonymous says:

    What percentage of snaps when the quarterback is under center are fumbled?

Leave a Reply

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