Turnovers in college football, 2013

Submitted by dnak438 on January 15th, 2014 at 2:02 PM

We don't need more cold water splashed in our faces after the season we just endured, but Football Study Hall did an analysis of turnover differential in the past season and tried to get an idea of who were the "luckiest" teams -- on the assumption that fumble recovery is essentially random, and here are the results:

Team Off. TO Def. TO TO Margin Adj. Off. TO Adj. Def. TO Adj. TO Margin Diff TO Luck/Game
Houston 18 43 25 27.4 31.1 3.8 21.2 +8.17
Ball State 19 31 12 23.1 21.6 -1.6 13.6 +5.21
Mississippi State 18 25 7 25.3 20.3 -5.0 12.0 +4.61
Oregon State 26 29 3 29.3 21.2 -8.1 11.1 +4.25
Florida State 18 35 17 19.5 25.5 6.0 11.0 +3.93
Missouri 16 32 16 19.9 25.7 5.8 10.2 +3.65
South Carolina 17 30 13 16.3 20.1 3.8 9.2 +3.54
Buffalo 15 30 15 20.6 26.6 6.0 9.0 +3.45
East Carolina 16 24 8 24.9 24.7 -0.2 8.2 +3.15
Michigan 21 26 5 21.9 18.8 -3.1 8.1 +3.12

Essentially, what you have is a TO margin, based on real numbers. Then you have adjusted turnovers, assuming that 51% of all fumbles are recovered by the offense, and that 22% of all passes defended are intercepted. The adjusted turnover margin is thus an estimate of a turnover margin without as much luck built in.

This suggests that Michigan was fairly lucky in terms of turnovers in 2013. We recovered more fumbles than average (15 opponent fumbles, 9 of which we recovered) and lost fewer than average (22 fumbles 8 of which we lost), and we had fewer picks than expected based on passes defended. (If you don't like what you see, just close your eyes are repeat to yourself that turnovers aren't random).

We played some of the unluckiest teams this year. The top 10 unluckiest:

Team Off. TO Def. TO TO Margin Adj. Off. TO Adj. Def. TO Adj. TO Margin Diff. TO Luck/Game
Cincinnati 28 21 -7 21.5 21.9 0.4 -7.4 -2.84
Alabama 17 19 2 14.7 24.1 9.4 -7.4 -2.85
Vanderbilt 23 30 7 19.6 34.2 14.6 -7.6 -2.93
Kansas State 25 25 0 16.5 24.4 7.9 -7.9 -3.02
Pittsburgh 17 16 -1 14.9 22.2 7.2 -8.2 -3.16
Indiana 20 17 -3 18.3 24.8 6.6 -9.6 -3.98
Southern Miss 38 19 -19 28.1 19.4 -8.8 -10.2 -4.27
Troy 18 11 -7 18.6 22.9 4.3 -11.3 -4.72
Rutgers 30 18 -12 23.6 24.5 0.9 -12.9 -4.95
Utah 25 16 -9 21.4 25.0 3.6 -12.6 -5.25




January 15th, 2014 at 2:27 PM ^

We were fairly unlucky as a conference, according to this. Only Michigan, MSU and Purdue managed positive luck, and the conference average was -0.7567. If the average turnover margin of the conference is close to zero though (I believe that it was only barely above zero, at least when I did the calculation in December), then it isn't difficult to imagine though. Interesting stuff - thanks for sharing this!


January 15th, 2014 at 3:15 PM ^

Maybe it is because Devin Gardner has a 5 yard head start on the "Center randomly moonshots the ball 20 yards toward his own endzone" play Michigan ran a few times this year.

Two Hearted Ale

January 15th, 2014 at 2:51 PM ^

Is anyone aware of turnover stats by position? Do offenses recover 51% of QB turnovers? RB? WR?

I would guess that a QB loses fewer fumbles than backs and receivers lose the most fumbles based on the number of team members near the fumble.


January 15th, 2014 at 3:11 PM ^

I searched and can't seem to find it, but I remember reading somebody's turnover manifesto, and they had parsed out the different types of fumbles.  The thing I remember most was that QB fumble rate is quite high when they are sacked, and more importantly, the defense recovery rate of a fumble from a sack is way higher than 51%.  So the greatest contributor to fumble takeaways was sacks/pass rush. 

EDIT: see link above


January 15th, 2014 at 3:09 PM ^

but why is fumble recovery thought to be completely random?  Obviously time, place, bounce, and other factors play a part, but does ability to recover the ball and swarm to the ballcarrier make a difference? 


January 15th, 2014 at 5:31 PM ^

I don't have time to find them, but there have been studies done that show that there is very little correlation between fumble recovery rate from one year to the next. If recovery were truly a skill, there shouldn't be much change in a team's rate from year to year. That's the main support for the argument that recovery is basically random. Also, as you say, location and bounce are such huge factors, it really makes it hard to show that recovery is something other than mostly luck.


January 15th, 2014 at 6:33 PM ^

Recovery of a fumble, despite being the product of hard work, is almost entirely random.

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. The odds of recovery are based solely on the type of play involved, not the teams or any of their players.

Fans like to insist that specific coaches can teach their teams to recover more fumbles by swarming to the ball. Chicago's Lovie Smith, in particular, is supposed to have this ability. However, in Smith’s first three seasons as head coach of the Bears, their rate of fumble recovery on defense went from a league-best 76 percent in 2004 to a league-worst 33 percent in 2005, then back to 67a percent in 2006.

Fumble recovery is equally erratic on offense. In 2008, the Bears fumbled 12 times on offense and recovered only three of them. In 2009, the Bears fumbled 18 times on offense, but recovered 13 of them.

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. With this in mind, Football Outsiders stats treat all fumbles as equal, penalizing them based on the likelihood of each type of fumble (run, pass, sack, etc.) being recovered by the defense.

Other plays that qualify as "non-predictive events" include blocked kicks and touchdowns during turnover returns. These plays are not "lucky," per se, but they have no value whatsoever for predicting future performance.

Link: http://www.footballoutsiders.com/info/fo-basics