Talking stats: SoCon football and turnovers

I was reading Jeff Hartsell’s review of The Citadel’s football season in The Post and Courier. In the second post of the three-part series, head coach Kevin Higgins had this to say about turnovers:

 We didn’t get as many turnovers as I would have liked. We just didn’t have that many opportunities. That’s something we’ll have to study in the off-season and address that. We need to be able to turn the ball over — one more turnover against Samford or Georgia Southern or App State could have meant the difference in any of those games.

I thought it might be an interesting idea to dig a little deeper into the statistical record to see what The Citadel could do to force more turnovers. However, that meant more than just going by the raw data.

First, I decided that it would be best to concentrate solely on Southern Conference play. Including games played against the likes of Virginia Tech and Virginia-Wise (just to name two SoCon opponents) would make the statistics something less than balanced. Besides, teams are ultimately judged on how they fare against league opponents. There is also the benefit of each team’s conference statistical summary including four home and four away games.

Another consideration was trying to account for the different types of offenses employed by SoCon teams, including three “true” triple option teams and several schools running the spread, and with varied paces of play. That is why I felt it was important to focus on certain percentage categories, rather than totals.

I compiled data (league play only) for a number of different statistics, both for offense and defense. After doing this, I put together a spreadsheet which you can access at the link below:

Southern Conference 2011 Football Statistics

I trust most of what is on the spreadsheet makes at least some sense.

Before I get to my conclusions about The Citadel’s issues with forcing turnovers (along with observations on some other SoCon schools), I want to make a few points:

– It is generally accepted that there is no real skill in recovering fumbles. Anyone who follows any of the websites that study professional football statistics/history is aware of this. Football Outsiders puts it best:

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, since Smith took over the Bears, their rate of fumble recovery on defense went from a league-best 76 percent to a league-worst 33 percent in 2005, then back to 67 percent in 2006. Last year, they recovered 57 percent of fumbles, close to the league average.

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.

Although this makes perfect sense, it is understandable that longtime football fans might not be so sure. I think the best way to illustrate the randomness of fumble recoveries is to highlight Pittsburgh Steelers legend Jack Lambert, who besides being a fantastic linebacker was one of my favorite players.

In the 1975 AFC championship game against the Oakland Raiders, Lambert recovered three fumbles. In the following year, 1976, he recovered an amazing eight fumbles (in fourteen games) for a remarkable Steelers defense. Lambert had a “nose for the football”, to say the least — and yet…

Those three fumble recoveries against the Raiders were the only recoveries he made in eighteen career playoff games. Those eight fumble recoveries in the ’76 regular season? They make up almost half of his career total (17).

Not everything about the NFL applies to college football, of course, particularly in FCS play, but there is no doubt that this particular observation does hold at the college level. Basically, when a ball is loose on the ground each team has a 50-50 shot at getting it. In 2011, there were 130 fumbles in Southern Conference play. The defense recovered 69 of those fumbles, or 53%. Congratulations to SoCon defenses!

No team in league play had a particularly unusual percentage when it came to recovered fumbles, either from an offensive or defensive perspective. It may be that an individual school was luckier or unluckier by a fumble or two, but that’s about it.

That isn’t to say that fumbles aren’t important, because they are. Often a fumble is more damaging to an offense than an interception (because of lost field position). However, they aren’t predictive events.

That doesn’t mean coaches shouldn’t be training their players to use the Lawrence Taylor “chop”, or continuing to have drills emphasizing fumble recoveries. It’s just that everybody does those things.

– Another thing to remember: interceptions (from a defensive perspective) tend to be random too.

This one isn’t quite as intuitive as the fumble recovery factoid, but think about it this way. Most interceptions result from a bad pass thrown by the quarterback. However, what has (normally) happened is that the QB has thrown a bad pass that was caught by a defender, instead of a bad pass that just hits the ground; there is an element of chance to this. That is why team defense interception totals can vary wildly from year to year even with similar personnel.

That isn’t to say that defenses can’t create situations where interception-prone offenses will toss the pigskin to the wrong players. I wanted to see what teams in the SoCon did the best job at pressuring the quarterback, which seemed to me to be a good way of forcing offensive errors.

I compiled sack percentage and interception percentage to see if they correlated. Again, I didn’t use raw totals, because there is a big difference when facing a team that throws the ball seven times per game (Wofford) versus forty times per game (Elon). The “pressures” statistic isn’t readily available for the SoCon; I suspect that there would have been similarities between team pressures and sacks. At least, I hope so.

You can see the numbers in the linked spreadsheet. Some observations:

– It is no accident that the three teams to make the playoffs (Appalachian State, Georgia Southern, and Wofford) are in the upper echelon when it comes to defensive sack percentage. Furman, which finished fourth in the league, finished second in the category. Leading the category was Chattanooga (more about the Mocs later).

– Defensive interception percentage does seem to at least have some correlation to defensive sack percentage. The exceptions: Samford (which intercepted more passes than it “should” have), and Wofford and UTC (each of which intercepted fewer passes than a correlation might suggest). The Terriers, in particular, seem to have been short-changed a few picks.

The Citadel’s defense finished last in interception percentage. The Bulldogs were seventh in sack percentage, ahead of only Samford and Western Carolina. I think it’s no coincidence that The Citadel didn’t intercept many passes after having less-than-stellar sack numbers. (Admittedly, that’s a rather obvious conclusion.)

– I also examined the offensive statistics for the same categories. The Citadel finished as the worst team in the league in both interceptions thrown (by percentage) and fumbles per play. The Bulldogs fumbled 23 times in SoCon action, losing ten of them. (Curiously, Georgia Southern also fumbled 23 times in league play, losing ten.)

While I tracked fumbles per play, I elected not to go through every game account to determine whether fumbles occurred on rushing or passing plays; that would have taken more hours and more days than I have, to be honest. In the NFL, the average rushing play results in a fumble 1.16% of the time, while a pass play will end with a fumble 2.04% of the time. Interestingly, 18% of all sacks in the NFL (2000-2009 time period) resulted in fumbles.

I’m not sure those numbers are quite as relevant at the college level; for one thing, there is a lot more fumbling in SoCon play than in the NFL (2.66% vs. 1.67%). There is also a lot more running than passing in the conference (almost a 2-to-1 differential).

Those sack/fumble stats are something to think about, however.

Meaningless trivia: there was only one game in the Southern Conference this season in which neither team fumbled: Georgia Southern-Appalachian State.

– The “luckiest” team, at least on the surface, appears to have been Samford, which finished eighth in defensive sack percentage but fourth in defensive interception percentage. The Birmingham Bulldogs also had the best rates for offensive interceptions thrown (with the fourth-best sacks against percentage), so it worked out both ways for Samford.

I have to wonder if Samford’s pace of play had something to do with that. Samford ran the most plays from scrimmage of any team in the league, and also faced the second-most plays on defense (Western Carolina drew the short straw in that category).

– Balance, as always, is overrated. Samford was by far the most balanced team on offense (305 runs, 301 passes) and finished 4-4 in league play. The second-most balanced team was WCU, which was 0-8. There is nothing offensively balanced about Georgia Southern and Wofford; those two playoff teams combined for a league record of 13-3.

– I don’t know what to make of Chattanooga. Usually a team that loses so many close games (including three by the same exact score, 28-27) doesn’t do itself any favors in the turnover battle, but the Mocs tied for the league lead in fumbles recovered and led the league in forced fumbles. UTC also finished second-best in the league in offensive lost fumbles.

UTC didn’t have the rate of return on defensive interceptions that might have been expected by its league-leading defensive sack percentage, but it wasn’t bad. The Mocs did have a higher average offensive interception percentage, but it wasn’t abysmal.

I think it would take a more detailed look at Chattanooga to figure out exactly how and when things went wrong for the Mocs, but I can safely say no team in the league was unluckier than UTC — just not as unlucky in the things you usually would associate with unlucky teams.

I guess my final conclusion, at least with regards to The Citadel, is that the Bulldogs must get more pressure on the quarterback if they expect to increase their defensive turnovers. However, it has to be remembered that defensive turnovers are an effect of good play, not a cause of good play.

I would also suggest the Bulldogs were a touch lucky on offense themselves when it came to turnovers, and need to continue to improve the consistency of execution on that side of the ball.

I admit my analysis of The Citadel (and some of the other teams in the league) may be flawed. That’s one reason I included the spreadsheet, in case anyone else wants to take a crack at what the numbers may mean.