Skill v. Luck: why FSU should abandon the fumble drill

As football fans, we want to believe that everything is within the control of our team.  Ideally, the final result would be a perfect net of the competitor's efforts.  Unfortunately, not everything in football is skill.  Like in life, football incorporates a certain amount of random events out of the control of either team.

In looking to improve your team then, it becomes very important to identify what events are within the dominion and control of your team, and what events are not.  If a team had unlimited practice time, that might not matter, but any good coach will tell you that practice time is his most precious resource.  In addition to plays and scrimmaging, teams practice various skills during their practice time, and one of the things they practice is recovering fumbles.  But this is a problem, because recovering fumbles isn't a repeatable skill. 

If you're the kid who ran down and recovered that fumble on the wet field in 1982, you probably disagree, but I challenge you to find any evidence that recovering fumbles is a repeatable skill.  And if you're a football coach, chances are you spend some time every week doing the fumble recovery drill.  Good news, you can eliminate that drill and work on something that will improve your team's chances of winning. 

The experts at FootballOutsiders.com have done some serious research on this subject and well, I'll just let them tell it...

Stripping the ball is a skill. Holding onto the ball is a skill. Pouncing on the ball as it is bouncing all over the place is not a skill. There is no correlation whatsoever between the percentage of fumbles recovered by a team in one year and the percentage they recover in the next year...

Fumble recovery is a major reason why the general public overestimates or underestimates certain teams. Fumbles are huge, turning-point plays that dramatically impact wins and losses in the past, while fumble recovery percentage says absolutely nothing about a team's chances of winning games in the future...

...[T]hey have no value whatsoever for predicting future performance or gauging the true ability of a team

I hate this.  Football is a game of such limited sample set (as opposed to baseball with its 600 at bats and 162 games, not to mention their 7 game playoff series).  I hate the finality of football.  Too often the best teams do not get to play for the championship, and for what, a few bounces not going their way?  Football would be much more equitable if the ball was round and not oblong.  The bounces would be truer. 

But football is played with an oblong ball, and that's unlikely to change any time soon.  But fumbles can have a dramatic effect on a team's season!  So with that in mind, let's look at which teams recieved those lucky bounces, and which team's did not; remembering that nothing suggests they will get those bounces in the coming season.

Offensive Fumbles

Let's start with a look at fumbles by the offense.  As per usual, these numbers come from conference play. 

Name Fumbles Lost Fumble Lost % ACC Record
North Carolina State 15 2 13% 4-4
Duke 11 4 36% 1-7
Miami (Florida) 8 3 38% 4-4
Wake Forest 12 5 42% 4-4
Clemson 16 8 50% 4-4
Florida State 15 8 53% 5-3
Virginia 12 7 58% 3-5
North Carolina 15 9 60% 4-4
Boston College 23 14 61% 5-4
Georgia Tech 22 14 64% 5-3
Virginia Tech 10 7 70% 6-3
Maryland 10 8 80% 4-4

 

NC State was darn near the luckiest team in the nation when it comes to recovering fumbles.  13% is quite insane, and I think this might reflect the type of fumble the pack suffered (I'll speculate about this at the end of the article).  

Meanwhile, Maryland lost 8 of their 10 offensive fumbles in conference play.  That's a rate that would put them in the 100's nationally.  Florida State was about average here, recovering just over 50% of their own fumbles.  I sure wish the Seminoles could have recovered that Antone Smith or Preston Parker fumble against Boston College.  Or that Marcus Sims fumble..., but for every one we complain about, there were some they fell on that could have been disastrous, like Bert Reed's fumble against NCST right before half.

Does anyone find it odd that this category didn't really correlate to winning?  I do, but I also believe some of it has to do with being unable to differentiate 10 teams who went 4-4 or 5-3 in their 8 game ACC seasons.

Let's now have a look at Defensive Fumble Recoveries. 

Name Opponent Fumbles Recovered Opponent's Fumbles Recovered % ACC Record
Florida State 13 10 77% 5-3
Virginia 11 8 73% 3-5
Wake Forest 16 11 69% 4-4
North Carolina State 11 7 64% 4-4
Clemson 10 6 60% 4-4
Boston College 12 7 58% 5-4
Duke 13 7 54% 1-7
Virginia Tech 19 10 53% 6-3
Georgia Tech 15 6 40% 5-3
Miami (Florida) 18 7 39% 4-4
North Carolina 13 4 31% 4-4
Maryland 17 5 29% 4-4

 

Well hello there, Florida State.  Recovering 10 of 13 fumbles really made this defense seem much better than it was.  Let's reminisce:

In the Virginia Tech game, the Hokies were having one of their better days offensively despite using a tight end at quarterback after losing their first two.  With Tech up 10-7, the Hokies fumble at their own 46 and FSU recovers it on the Tech 46, kicks a field goal to tie the game at 10.  Then, with FSU up 24-13, the Noles kick off and VT fumbles on their own 26, FSU takes over and they get a field goal to go up 10 and effectively ice the game.

And sometimes, as against Boston College, fumble recoveries can mask poor defense.  Boston College had drives of 47 yards (9 plays) and 54 yards (5 plays) erased by FSU fumble recoveries.  In fact, because many people blame only the offense for the BC loss, I want to point out BC's last 5 drives of the game:

9 plays, 47 yards
14 plays, 65 yards
7 plays, 71 yards
5 plays, 54 yards
16 plays, 67 yards

And I'll remind you that BC had one of the worst offenses any of us have seen in a long time.  Two of those drives ended in fumbles for Boston College, and had the ball not bounced fortuitously for FSU, this game could have been a much bigger loss than it was.

The other game that really sticks out to me, as an FSU fan, was Maryland.  Maryland's first two drives were 7+ plays and 35+ yards.  They were running well on FSU.  But then "Da'Rel Scott rush, fumbled, recovered by FlaSt Derek Nicholson at the Mary 22, Derek Nicholson for 22 yards, to the Mary 0 for a TOUCHDOWN."  FSU would recover one more Maryland fumble late in the game, but at that point it was already over.

I'll venture a guess that UVA would have had a duke-like record had they not recovered 73% of their opponent's fumbles.  Wake always seems fortuitous in this area, but the numbers say they don't possess some special ability to recover fumbles that the other 119 teams do not.  (Their specialty is interceptions).

Finally, UNC played phenomenal defense for most of the year, yet the couldn't get the ball to bounce their way.  Ditto Maryland (discussed below).

Now let's combine the two

Name Total Fumbles (by and against) Total Recovered Total Recovery % ACC Record
North Carolina State 26 20 77% 4-4
Wake Forest 28 18 64% 4-4
Florida State 28 17 61% 5-3
Duke 24 14 58% 1-7
Virginia 23 13 57% 3-5
Clemson 26 14 54% 4-4
Miami (Florida) 26 12 46% 4-4
Boston College 35 16 46% 5-4
Virginia Tech 29 13 45% 6-3
Georgia Tech 37 14 38% 5-3
North Carolina 28 10 36% 4-4
Maryland 27 7 26% 4-4

Florida State was slightly above average overall in fumble recovery, or "FLuck" (Fumble Luck), but let's look at two of the countries extreme outliers.

I've done the review, and everyone else is correct in their research:  nobody is able to recover a greater percentage of fumbles from year to year.  It really is luck, and that makes sense because of how the ball bounces.  Here's a photo of the SEC's 2007 data:

Fumbleluckwx3_medium

via img71.imageshack.us

Something very interesting here:  nobody had higher than 63% or lower than 42%.  Wildly, 5 of the 7 ACC teams had a percentage outside that range!  Here's the

2007 data from the ACC

Name Total Fumbles (by and against) Total Recovered Total Recover % ACC Record
Miami (Florida) 25 15 60% 2-6
Wake Forest 22 13 59% 5-3
Virginia Tech

 

33

19 58% 7-1
Virginia 28 16 57% 6-2
Duke 26 14 54% 0-8
North Carolina 26 14 54% 3-5
Boston College 27 14 52% 6-2
Clemson 20 9 45% 5-3
Georgia Tech 34 15 44% 4-4
Florida State 30 13 43% 4-4
Maryland 23 8 35% 3-5
North Carolina State 16 5 31% 3-5

2007 was a much more reasonable year.


The two 2008 extremes

Maryland was just ridiculously unlucky.  Recovering only 7 of the 27 available fumbles in 8 games is like akin to having someone steal your credit card, destroy your credit, and then having your car repossessed (okay, I'm running low on examples).  I've never seen anything like Maryland's fumble luck.  They might have won the Atlantic division had they had even normal bounces of the ball. 

North Carolina State, on the other hand, is on the exact opposite end of the spectrum.  77%?!!?  That is even crazier than Maryland's 26%  NC State did get a lot better as the season wore on, but some of that is absolutely attributable to their fumble luck.  BC thoroughly outplayed the pack in their house, but NC State kept it close by recovering some of their own fumbles.  Further, the Pack housed UNC while recovering three UNC fumbles and losing none of their own!  Such was the crazy variance of the ACC in 2008. 

No Consistency from year to year.

This has been covered in the linked precursors to this piece, but the same idea holds true.  Recovering fumbles is not a skill.  No team is able to consistently recover a higher percentage of fumbles than any other from year to year.  Teaching the technique of fumble recovery is a waste of time because the outcome is simply too dependent on the bounce of the ball.  Now, if a coach had some way to determine how the ball will bounce and could teach their players to make that split second determination, they'd have something.  Unfortunately, no coach has discovered this method. 

If you compare the 2007 percentages to the 2008 percentages, you'll see wild swings.

Name Total Recovery % 2007 Total Recovery % 2008 2007 Rank/ 2008 Rank
Boston College 52% 46% 7th/ 8th
Clemson 45% 54% 8th/ 6th
Duke 54% 58% 5th/ 4th
Florida State 43% 61% 10th/ 3rd
Georgia Tech 44% 38% 9th/ 10th
Maryland 35% 26% 11th/ 12th
Miami (Florida) 60% 46% 1st/ 7th
North Carolina 54% 36% 6th/ 11th
North Carolina State 31% 77% Last/ 1st
Virginia 57% 57% 4th/ 5th
Virginia Tech 58% 45% 3rd/ 9th
Wake Forest 59% 64% 2nd/ 2nd

NC State makes our case for us.  From worst to first, from sitting home in December to playing in a bowl game.

In thinking about this, I was initially skeptical.  It seems to me that there are still problems with this and they relate to scorekeeping.  Scorer's just don't keep detailed enough information for us to be able to separate things like fumbled snaps, from other types of fumbles.  I'd be very interested to see how often a QB recovers a snap under center versus a snap from the shotgun versus another type of fumble.  I'm betting there would be differences.

Another issue I'd like to look at but am sure it won't happen because of the lack of available data is the type of defensive being played.  Does a zone defense have a greater chance of receovering a fumble than a man defense?  In a zone defense, everyone is looking at the ball, and it's concieveable that all 11 defensive players could see the fumble and only one offensive player (the fumbler) would know of or see the fumble.  This might be a benefit on forced fumbles, but a zone might hurt the defense on unforced fumbles as the defender would probably be farther away from the fumble.

Editor's Note:  I have to give a lot of credit to author OutsideTheLines at the fabulous Alabama blog, RollBamaRoll.com for his piece on this very subject.  I've read others elsewhere but can't remember the links. 

I can only hope that the 'Noles keep up their 2008 pace, but realistically, they have just as good of a chance to regress to their 2007 level or lead the conference.  Such is life when dealing with FLuck. 

 

Go Noles!

Become a fan of us on facebook

http://www.facebook.com/home.php#/pages/TomahawkNationcom/63264950770

Follow us on Twitter!  https://twitter.com/TomahawkNation

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Tomahawk Nation

You must be a member of Tomahawk Nation to participate.

We have our own Community Guidelines at Tomahawk Nation. You should read them.

Join Tomahawk Nation

You must be a member of Tomahawk Nation to participate.

We have our own Community Guidelines at Tomahawk Nation. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker