/cdn.vox-cdn.com/uploads/chorus_image/image/49247705/usa-today-9005300.0.jpg)
Most aspects of football are mostly skill-based, but luck in large part governs turnovers. Or, aspects of turnovers at least. Today, I'm going over those random aspects. All stats are sourced from the excellent CFBStats.com.
Ball security is a skill. So is forcing fumbles. But recovering fumbles? It's basically all luck. The football has a funny shape and bounces in ways that people can't predict in real time. Besides that, it's not possible to control what mix of players from the two teams will be near the ball when it comes loose.
Here is the distribution of fumble recovery percentages from 2015 rounded to the nearest percentage:
It's in the shape of a bell curve, which means it's very likely to be random. And that point way at the top that was the most frequent fumble recovery percentage? It's 50%.
Here are all the teams from 2015 that recovered at least 60% of the fumbles in their games:
Name | Total Fumbles | Total Recovered | Recovered Pct. |
---|---|---|---|
Arkansas | 29 | 19 | 65.5% |
Houston | 40 | 26 | 65.0% |
Wisconsin | 28 | 18 | 64.3% |
Fresno State | 30 | 19 | 63.3% |
Oklahoma State | 38 | 24 | 63.2% |
Army | 48 | 30 | 62.5% |
USC | 26 | 16 | 61.5% |
Oregon | 36 | 22 | 61.1% |
South Carolina | 38 | 23 | 60.5% |
Kansas | 40 | 24 | 60.0% |
Navy | 40 | 24 | 60.0% |
North Carolina | 45 | 27 | 60.0% |
UNLV | 30 | 18 | 60.0% |
There is quite the mix here, from New Year's Six participants Houston and Oklahoma State down to Kansas and Army. Heading the other way, these are the teams that recovered fewer than 40% of their fumbles:
Name | Total Fumbles | Total Recovered | Recovered Pct. |
---|---|---|---|
Colorado | 48 | 19 | 39.6% |
Louisiana Tech | 38 | 15 | 39.5% |
Nebraska | 26 | 10 | 38.5% |
NC State | 43 | 16 | 37.2% |
Cincinnati | 35 | 13 | 37.1% |
Alabama | 37 | 13 | 35.1% |
Oregon State | 26 | 9 | 34.6% |
Duke | 38 | 13 | 34.2% |
Penn State | 53 | 18 | 34.0% |
Mississippi State | 36 | 11 | 30.6% |
Middle Tennessee | 44 | 13 | 29.5% |
Once again, the spread here is enormous from national champion Alabama to a pair of teams in Oregon State and Colorado that went a combined 1-17 in Pac-12 play.
The other category of turnovers is interceptions. Avoiding bad throws as a quarterback is a skill. Calling defensive sets to get defenders in the way of passes is a skill for coaches, as is knowing when to take risks to get picks for a player. Actually hauling in the interception does seem to be luck, though, which shouldn't be surprising considering that the players with the best hands tend to get put on offense.
Here is the distribution of interceptions as a percentage of passes defended (passes broken up plus INTs) rounded to the nearest percentage:
Again, it looks a lot like a random distribution. The percentage for all teams put together is 21.1%, and the median percentage was 20.7%.
Here are the teams that intercepted at least 30% of their passes defended:
Name | Int. | Passes Defended | INT Pct. |
---|---|---|---|
Connecticut | 18 | 49 | 36.7% |
Oklahoma State | 17 | 49 | 34.7% |
Appalachian State | 18 | 52 | 34.6% |
Arkansas State | 26 | 77 | 33.8% |
Bowling Green | 20 | 60 | 33.3% |
Oklahoma | 20 | 61 | 32.8% |
Boston College | 13 | 41 | 31.7% |
New Mexico State | 12 | 38 | 31.6% |
San Diego State | 23 | 75 | 30.7% |
Boise State | 22 | 73 | 30.1% |
There is a crazy mix here too, again from NY6 bowl teams like Oklahoma State and Oklahoma to BC, who didn't win an ACC game, and 3-9 New Mexico State. Boston College did have the No. 3 S&P+ defense, but NMSU's was fourth from the bottom and actually rated worse against the pass than the run. And now for the end of the list:
Name | Int. | Passes Defended | INT Pct. |
---|---|---|---|
Old Dominion | 8 | 62 | 12.9% |
North Texas | 4 | 33 | 12.1% |
Eastern Michigan | 4 | 34 | 11.8% |
UTEP | 5 | 43 | 11.6% |
TCU | 7 | 63 | 11.1% |
Iowa State | 5 | 45 | 11.1% |
Texas State | 3 | 28 | 10.7% |
Kansas State | 5 | 47 | 10.6% |
Vanderbilt | 6 | 60 | 10.0% |
UCF | 4 | 40 | 10.0% |
Louisiana-Lafayette | 5 | 53 | 9.4% |
Virginia | 4 | 44 | 9.1% |
Hawai'i | 3 | 43 | 7.0% |
Rice | 2 | 32 | 6.3% |
Down here is mostly bad teams, though we do get TCU sitting there at 11.1%. Vanderbilt may have missed a bowl, but it still had a top 20 defense.
Finally, coaches love to talk about turnovers like they're something within their control. To test that notion, I put every team's ranks in fumble recovery percentage and interception percentages in a table. I ran the correlation between the ranks, and it came out to 0.02. As a refresher: correlation is a number between 1 and -1, and the closer to zero it is, the less correlation there is between two factors. In other words, how good (or not) a team was at recovering fumbles tells us nothing about how good it was at intercepting passes it got a hand on and vice versa.
The implication here is that teams that were highly ranked in both were exceptionally lucky and those that were lowly ranked in both were exceptionally unlucky. Here are the lucky ones:
Team | Fumble Rank | INT Rank | Average Rank |
---|---|---|---|
Oklahoma State | 5 | 2 | 4 |
South Carolina | 9 | 13 | 11 |
Boston College | 19 | 7 | 13 |
New Mexico | 14 | 20 | 17 |
USC | 7 | 28 | 18 |
Kansas | 10 | 27 | 19 |
New Mexico State | 34 | 8 | 21 |
Houston | 2 | 43 | 23 |
South Florida | 37 | 11 | 24 |
Miami (Florida) | 17 | 33 | 25 |
Bowling Green | 50 | 5 | 28 |
Arkansas State | 55 | 4 | 30 |
Syracuse | 44 | 15 | 30 |
Arizona State | 30 | 29 | 30 |
Navy | 11 | 49 | 30 |
And here are the snake bit teams:
Team | Fumble Rank | INT Rank | Average Rank |
---|---|---|---|
Eastern Michigan | 80 | 117 | 99 |
Louisiana Tech | 119 | 79 | 99 |
Alabama | 123 | 77 | 100 |
Penn State | 126 | 76 | 101 |
Mississippi State | 127 | 78 | 103 |
Wyoming | 94 | 111 | 103 |
Michigan | 104 | 103 | 104 |
UTEP | 89 | 118 | 104 |
Virginia | 90 | 126 | 108 |
Rice | 88 | 128 | 108 |
Cincinnati | 122 | 97 | 110 |
Nebraska | 120 | 99 | 110 |
Hawai'i | 92 | 127 | 110 |
Texas State | 107 | 121 | 114 |
Louisiana-Lafayette | 116 | 125 | 121 |
Just imagine how good the Crimson Tide would've been with even just neutral luck in turnovers, much less good luck.
So those are the turnover luck extremes for 2015. Because this is an SEC-focused site, I'm going to put the tables for just the conference's teams below. If you're not an SEC fan or otherwise don't care, you're dismissed.
Fumble recovery percentages:
National Rank | Name | Total Fumbles | Total Recovered | Recovered Pct. |
---|---|---|---|---|
1 | Arkansas | 29 | 19 | 65.5% |
9 | South Carolina | 38 | 23 | 60.5% |
25 | Missouri | 36 | 20 | 55.6% |
27 | Tennessee | 36 | 20 | 55.6% |
40 | LSU | 36 | 19 | 52.8% |
49 | Georgia | 46 | 24 | 52.2% |
54 | Vanderbilt | 43 | 22 | 51.2% |
61 | Florida | 44 | 22 | 50.0% |
65 | Kentucky | 32 | 16 | 50.0% |
72 | Ole Miss | 49 | 24 | 49.0% |
74 | Auburn | 39 | 19 | 48.7% |
117 | Texas A&M | 42 | 17 | 40.5% |
123 | Alabama | 37 | 13 | 35.1% |
127 | Mississippi State | 36 | 11 | 30.6% |
Interception percentages:
National Rank | Name | Int. | Passes Defended | INT Pct. |
---|---|---|---|---|
13 | South Carolina | 12 | 42 | 28.6% |
21 | Georgia | 12 | 45 | 26.7% |
62 | Auburn | 14 | 67 | 20.9% |
67 | Florida | 14 | 68 | 20.6% |
69 | Arkansas | 11 | 54 | 20.4% |
77 | Alabama | 19 | 99 | 19.2% |
78 | Mississippi State | 13 | 68 | 19.1% |
80 | Texas A&M | 11 | 58 | 19.0% |
83 | Kentucky | 11 | 59 | 18.6% |
84 | Tennessee | 12 | 65 | 18.5% |
91 | LSU | 10 | 57 | 17.5% |
94 | Ole Miss | 15 | 88 | 17.0% |
96 | Missouri | 9 | 53 | 17.0% |
123 | Vanderbilt | 6 | 60 | 10.0% |
National ranks in both:
Team | Fumble Rank | INT Rank | Average Rank |
---|---|---|---|
South Carolina | 9 | 13 | 11 |
Georgia | 49 | 21 | 35 |
Arkansas | 1 | 69 | 35 |
Tennessee | 27 | 84 | 56 |
Missouri | 25 | 96 | 61 |
Florida | 61 | 67 | 64 |
LSU | 40 | 91 | 66 |
Auburn | 74 | 62 | 68 |
Kentucky | 65 | 83 | 74 |
Mississippi | 72 | 94 | 83 |
Vanderbilt | 54 | 123 | 89 |
Texas A&M | 117 | 80 | 99 |
Alabama | 123 | 77 | 100 |
Mississippi State | 127 | 78 | 103 |