Roll 'Bama Roll: An SB Nation Community

Navigation: Jump to content areas:


Pro Quality. Fan Perspective.
Login-facebook
Network Message: 50% Off: CBS/SB Nation Fantasy Baseball

Fumble Luck

We all know that luck has an impact on sporting events, and in some individual situations it is easy to see that some team or another got "lucky." In the aggregate, though, it's much more difficult to say. There is no concrete rule that all good breaks and bad breaks must equal out in the end, and the truth is that some teams do get luckier than others. Again, though, it's generally pretty hard to quantify in many areas.

One of the areas we can quantify luck, however, is in regard to fumbles. Though fumbles are huge, game-changing type plays, they often come down more to luck than anything else. The research has been done, fumble recovery is nothing more than a random event -- causing fumbles, whether by coughing up the football or stripping the ball on defense, is a skill, but fumble recovery is not. It's basically little more than luck, it all mainly boils down to whether or not someone wearing your color jersey happens to be near the fumble, see it, and be able to pounce it.

The experts at Football Outsiders had this say on the subject:

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.


The research at FO, of course, is done for the National Football League, but I've crunched the numbers for SEC football, and the ultimate conclusion is still the same. Just as with the NFL, fumble recoveries in the SEC are entirely random events, and there is no correlation whatsoever with future performance. For example, as we will see further in a moment, the two teams that finished last in the conference in fumble luck in 2006 -- LSU and Arkansas -- finished first and second in fumble luck in 2007. On the other hand, Auburn finished second in fumble luck in 2006, and yet finished dead last in 2007.

Click here for the full conference data in 2007.

A few notes on several teams:

LSU: You could easily tell by following LSU this past season that they were about as lucky as you could possibly get, and the fumble luck data on further backs up that initial observation. But there is more to it than just that they merely led the conference in fumble luck. Their fumble recovery percentage of 63.64% is simply the highest I've seen to yet. The data I've crunched goes back to 1998 -- and though a few official web sites so lamely do not list some official statistics for some years, so it's a bit incomplete -- and that includes a little over 80 teams. Of those 80+ teams that I have fumble recovery data far, LSU is number one overall in fumble luck. Their 83% fumble recovery rate for their own fumbles, I will bet good money, has to be the highest we have seen in this conference, or will see in the conference, in ages.

Alabama: As we found with Pythagorean Wins, there was no real silver lining for the Tide. The same goes for fumble luck, too. We were right in the middle of the pack in terms of fumble luck, and you simply cannot say bad fumble luck cost us games. We weren't a 9-4 team that ended up 7-6 because of bad fumble luck. We were a 7-6 team that ended up with 7-6 with average fumble luck.

Auburn: Unlike their fellow Tigers, these Tigers finished dead last in the conference in fumble luck, and that really makes you wonder. I mean they were far from being a bad team to begin with. They finished 9-4, and would have made it to Atlanta had LSU not completed a long, last-second touchdown pass. It was obviously going to be a bit of a down year for them, but they still accomplished all that they did with the worst fumble luck in the conference. I've been big on Auburn for months now, and I see no reason to change that now. This team was 9-4 last year with bad fumble luck, and now they will have just about everyone returning in 2008, the schedule gets easier, and they should have a regression to the mean in fumble luck. If I'm delusional enough to think that it truly is great to be an Auburn Tiger, I would really have love what all of those things collectively portend.

Arkansas: The Hogs had the second best fumble luck in the conference, and in just about any other year they would have led the conference by coming in at approximately a 58% rate. Yet they still finished a mediocre 8-5, and clearly regressed. And now as if they didn't have enough problems anyway, they are likely facing a regression to the mean in fumble luck. Any time you have an 8-5 team with great fumble luck lose a ton of starters, including all of its impact players, plus you add in a coaching change... that one sounds about like a pending rectal probe. I don't hate Petrino like some, and I don't think he's the devil in disguise, but good grief does he have his work cut out for him. Far from hating him, I actually feel a bit sorry for the guy with the task he has at hand.

Tennessee: The Vols, as we know from Pythagorean Wins, were a huge overachiever, and had the fewest Pythagorean Wins of any team to win an SEC division since going to its modern format in 1992. Not surprisingly, the Vols finished third in the conference in fumble luck. I don't think there is any doubt that there is a direct link between fumble luck and overachieving / underachieving in regard  to Pythagorean Wins.

South Carolina: The Gamecocks were expected to improve after underachieving their Pythagorean projection in 2006, but improvement did not come, and actually a bit of regression reared its ugly head. A partial explanation of that oddity certainly has to come from their fumble luck, as they finished second to last in the conference in fumble luck. Again, I don't think there is any doubt that there is a direct link between fumble luck and overachieving / underachieving in regard to Pythagorean Wins.

0 recs  |  Comment 7 comments

Story-email Email Printer Print

Comments

Display:

I don’t really buy the premise of the article. Anyone who says that recovering a loose football is luck has never actually tried it. There’s a tremendous amount of skill involved.

Further, any event whose probabilities you can change by changing the inputs is not random. The odds of recovering a football increase dramatically when someone jumps on it versus trying to scoop it up and run with it. Similarly, when there are more defenders around the ball, the defense is more likely to recover that fumble.

There might not be a correlation between fumble recovery and wins, but that doesn’t mean that the event itself is random.

by PeteHoliday on Jun 18, 2008 10:23 AM CDT reply actions   0 recs

Fumbling For Thoughts

Is there any thought given to the type of fumble? There’s a world of difference between a fumble where the ball is caroming around the field like an ice cube on a tile floor and a fumble where a QB drops back to pass, drops the ball at his feet, and he falls on top of it to kill the down. Statistically speaking, the former fumble requires a significant amount of luck, while the latter requires gravity. It would be nice if “fumble difficulty” or some such modifier was factored in.

Also, I think replays should be taken into account. Let’s say a RB stretches for a first down and loses control of the ball. Upon review, however, the “fumble” is overturned because the officials rule that his knee marked the play dead. However, the players didn’t know that at the time. As far as I’m concerned, if the ball is considered to be live while the play is happening, then regardless of the later ruling it should count for statistical purposes. In fact, having a fumble overturned due to replay is itself a kind of fumble luck. On a tangential note, I also think replay can inadvertently reward sloppy play on the part of the offense … the offensive player shouldn’t put the possession in the hands of the officials … but that is an argument for another day.

"YouTube has destroyed our ability to know when we're entertained. It has turned us into a nation of deranged Roman emperors. We're a continent of Caligulas, sitting around in our bathrobes saying, 'Hmm ... I'm bored. I'd like to see something poop or sneeze.'"
--Patton Oswalt

by pantsfucious on Jun 18, 2008 10:24 AM CDT reply actions   0 recs

Excellent work, OTS, and I'm not just saying that because of what it might mean for Auburn

Though of course it doesn’t hurt. Regardless, between this post and the SMQ/RockM down-and-distance work, it’s been a good week for college football stat wonkery.

Pete, forgive me, but I have to ask: how do you look at the top team in fumble luck in each SEC division winning that division and conclude “there might not be a correlation between fumble recovery and wins”? How do you read “the two teams that finished last in the conference in fumble luck in 2006 - LSU and Arkansas - finished first and second in fumble luck in 2007. On the other hand, Auburn finished second in fumble luck in 2006, and yet finished dead last in 2007” and conclude fumble recovery is a repeatable skill?

by JCCW Jerry on Jun 18, 2008 2:01 PM CDT reply actions   0 recs

Correct...

You are correct, Jerry.

First and foremost, we’re not talking about wins here. The correlation between wins and fumble luck is not relevant in any way… we may as well be talking about wins and the head coach’s horoscope on the day of the game.

What is relevant is fumble luck and future performance, and that is what we are analyzing here. When you do that analysis, there is simply no correlation whatsoever between fumble luck in one year and fumble luck in another. In other words, there is simply no indication whatsoever that those things are connected. Again, it’s all random, the data points are all over the places. I’ve got 10 years of data at my disposal, and there is just nothing to it. No team has been able to consistently have good or bad fumble luck, you generally just bounce all over the place with nothing but white noise.

Don’t get me wrong, no one is saying that fumble recoveries are not the product of hard work and effort because they certainly are. However, in the aggregate it is a "lucky" event, meaning there is no correlation whatsoever to future performance, and meaning that at the end of the day it is not an actual skill. Successfully arguing that it is a skill can only be legitimately concluded once we have evidence of a particular team being able to consistently overachieve or underachieve in terms of fumble luck, and with 10 years of data at our disposal, we have no evidence of that.

by outsidethesidelines on Jun 18, 2008 4:42 PM CDT up reply actions   0 recs

Yet another great article!

Great job outsidethesidelines! I always enjoy reading your articles, they are better than 99% of the other football pieces out there … I mean it!

by RisingTideUA on Jun 18, 2008 4:53 PM CDT reply actions   0 recs

Great Article

I’m curious, did field position factor at all? Or the fumbler? As in, did 90% of fumbles by a QB get recovered by the offense, 50% by a running back, and 10% by a WR?

by Bobby Briggs on Jun 18, 2008 6:27 PM CDT reply actions   0 recs

Excellent article

Yet, I wonder how game situations factor into the equation. For example, I (and others) can argue pretty effectively that JPW’s fumble against LSU cost us that game, as well as his fumbles against Auburn the year before.

(And who can forget his fumble against LSU two years ago in the red-zone right before halftime….oh, and the game-changing redzone INT’s against AU and MSU last year….and also the momentum shifting INT’s against HOUSTON AS WELL AS IN THE FUCKING BOWL GAME!!! ARRRRRRRRRRRRRRRRGGGGGGGGGHHHHHH)

Ahem…anyway, I would think situational fumble recovery would be a pretty good indicatior of victory, although I’m not sure how the “measureables” would work. Maybe looking at it from the other end: since we know that there’s skill involved in HANGING ON TO THE DAMN BALL JOHN PARKER!!!...sorry…then that takes us right back to the ol’ giveaway-takeaway stat, which is a pretty good indication of success…

by sandman227 on Jun 19, 2008 1:04 PM CDT reply actions   0 recs

Comments For This Post Are Closed


User Tools

Welcome to the SB Nation blog covering the Alabama Crimson Tide.
Start posting about the Crimson Tide »

Join SB Nation and dive into communities focused on all your favorite teams.

Connect_with_facebook

FanPosts

Community blog posts and discussion.

Recommended FanPosts

26935420b_small
NCAA Balks on Head-Coach-in-Waiting Rule

Recent FanPosts

Nickcave_small
Classic Music Friday: THE KINKS
Twitter_small
SEC looking to give Alabama some schedule relief
S6300204_small
Bobby Humphrey Nominated to CFB HoF
Small
Rare audio of man who played for Wallace Wade
Nickcave_small
Music Monday: DAN DEACON
Script_a_small
Star players at the gun range
Me_in_sunny_clai_small
Awesome Article....
Nickcave_small
Classic Music Friday: BUDDY HOLLY
Small
Ten greats who never made All-America

+ New FanPost All FanPosts >

Roll Bama Roll on Twitter


Managers

Large_hankwilliams_small Todd

Miltonf-788904_small outsidethesidelines

Nickcave_small Nico2.0

Author

Kleph_logo_copy_small kleph

26935420b_small PeteHoliday

Oguchionyewu_1511662_small Bama07

Official Partner of CBS Sports