This Post is Graphic in Nature: Upon Further Review, Georgia’s Defense Returns almost as much as South Carolina’s and LSU’s
Dude’s note: all statistics utilized here are courtesy of Bill Connelly of SBNation/Football Outsiders/Football Study Hall. Your team’s website may show different numbers (Heck, mine does!), but he’s organized the data I wanted already in three articles. Georgia’s data was drawn from this. South Carolina’s data was drawn from this. LSU’s data was drawn from this.
The Georgia Bulldogs will open the season against Clemson before entering SEC play against South Carolina at home. The Dawgs’ second conference bout comes at home against the LSU Tigers. To say that the Bulldogs better be ready to play from day one is an insult to stating the obvious. Hence the urgency around the Georgia football program this summer.
As we all know, Georgia lost a lot on defense. And with an opener against Clemson (a high powered offense), the absence of a dozen defensive stalwarts is magnified. But Georgia isn’t the only defense that is faced with a rebuilding project. In fact, the other two SEC teams I have already mentioned (South Carolina and LSU) have their work cut out for them as well. But you might not have heard about that.
The graph below shows (as a percentage) the amount of production from last season that is returning for each of those three schools in various areas. There is no intentional “cumulative” effect, the connection of the data points is simply for illustrative purposes.
As you can see, this data examines returning tackles, tackles for loss, sacks, pass break-ups, interceptions, forced fumbles and fumble recoveries. Those are seven significant categories of measure.
If you want a more thorough numerical breakdown, here is what each team returns in each area:
That picture is admittedly convoluted, so let’s just take a look at the raw average. I don’t know that this is necessarily accurate in totality (more on that in a moment), but if we average the totals the figures above this is the amount of defensive production that each school in question returns:
So, while Georgia has certainly lost a lot on defense (the Dawgs return just 34.22% of total productivity), they aren’t light-years behind South Carolina (returning 42.63%) and LSU (43.70%).
But, I don’t think this is an entirely accurate portrayal of each team.
Some of this data is over-valued.
Tackles, for instance, are a bit of a moot point. While tackles may demonstrate how effective an individual player is within a team, total team tackles don’t necessarily demonstrate team success. By their very nature, tackles need to happen. But, there are several distorting factors in tackle numbers:
- A team that forces a three-and-out series only has three chances to record a tackle. In contrast, a team that allows long, slow drives may be forced to make 15 tackles. Which defense is better? Not the one with more tackles.
- A team that intercepts a pass or recovers a fumble doesn’t get to make any more tackles on that possession.
- Regardless of personnel, somebody (especially on talent-loaded SEC teams) steps up to make tackles. Case in point: in 2009 the Georgia Bulldogs registered 877 total tackles according to Georgiadogs.com. They fired Willie Martinez, hired Todd Grantham and dumped the 4-3 for the 3-4. The next year, Georgia registered 877 tackles. Those numbers are coincidentally (freaky) similar, but you get the point.
So, I don’t think that tackles carry much weight. Especially since statistical measurements being used (tackles for loss, sacks, pass break-ups, interceptions, forced fumbles and fumble recoveries) represent much bigger plays – both in the stopping of individual drives and in the momentum of a game.
With tackles removed from the averages, here is how Georgia, South Carolina and LSU stack up:
Georgia declines slightly (from a 34.22% average returning production figure down to 34.11%). LSU declines a little bit more (from 43.70% down to 42.58%). South Carolina declines as well (from 42.63% to 41.54%).
I also assert that the passes broken-up category is disproportionally credited. Don’t get me wrong; I recognize the merits of PBUs on two-folds:
- A PBU prevents a catch (and therefore limits the yards that go with it and possibly the movement of the chains for a first down or even the changing of the scoreboard for a TD).
- A PBU renders a passing down useless.
However, a pass breakup benefits the team much moreso than an individual player. And, although this study is ultimately being done to analyze a team, it is reflective of individual stats. A PBU limits yards, first downs, touchdowns, points and in the process makes the offense waste a down. But, a PBU tells us that a receiver did not catch a ball because a defender stopped the ball’s trajectory; nothing more. It does not tell us what would have happened had the player caught the ball. It ultimately yields an incomplete pass – something that happens 45% of the time anyway. It does not offer a firm intrinsic value. (In contrast, a sack definitively keeps a QB from passing. A tackle for loss definitively keeps a player behind the line of scrimmage. An interception or fumble recovery definitively ends a drive, every single time.)
I suggest that passes broken-up should not be included in the returning figures either. The chart below shows the production without tackles and without PBUs:
As you can see, Georgia advances slightly (up to 34.59%) with this refinement. LSU barely declines (down 0.16% from the last measure to 42.42%). South Carolina moves down to 39.84%.
Dude’s note: Before I jump into analysis and final commentary, I want to head-off the “you’re being selective with data” arguments. 1. I did not remove either of Georgia’s two worst categories (Georgia returns just 28.13% of last year’s sacks and just 30.77% of last year’s tackles for loss). 2. I did not remove South Carolina’s best two categories (the Cocks return 52.63% of sacks and 52.13% of their tackles for loss). 3. I did not remove LSU’s best category (53.85% returning in fumble recoveries). I took out two categories that I felt had minimal bearing. Now for some random points of emphasis…
Georgia vs. South Carolina:
The narrative about Georgia’s defense this offseason has been, “They lost everyone!!!”
The South Carolina narrative has been, “CLOWNEY!!!!”
South Carolina’s defense returns marginally more than Georgia’s by these standards – just over 5%. Combine that with the fact that the Cocks have lost 15 of their 23 turnovers (65.22%) as represented by INTs and Fumble Recoveries, while Georgia lost just 18 of 29 (62.01%) and I’m not sure that South Carolina’s defense returns decidedly more than Georgia’s. Does Carolina have more back? Probably so. But not by a landslide.
South Carolina returns Jadeveon Clowney. Yes, he’s the best player in the country. But the returning figures I utilized feature Jadeveon Cloweny’s performance. And, I’m not sure that his production will be guaranteed to be as good with as much as 60% of the surrounding production gone. (Dude’s note: much to the chagrin of Gamecocks, I made this point earlier this summer. They hated me for it.)
As depleted as Georgia may be (and to be fair: I think they are more depleted than the other two teams in question – but not by much), the Dawgs are surprisingly balanced.
- Georgia’s most depleted unit is the linebacking unit. That unit will return just 30.35% of its TFL, Sack, INT, Forced Fumbles and Fumble Recovery production.
- South Carolina’s worst unit is also its linebackers. The Gamecocks return just 0.89% of their production from linebackers. You read that correctly: that’s less than 1%. That happens when you run a 4-3 defense and lose your top-five LBs.
- LSU returns just 11.12% of its defensive line production.
Similarly, Georgia’s poorest statistical category is better than South Carolina and LSU’s respective low tallies:
- Georgia returns just 28.13% of its sacks from 2012.
- South Carolina returns just 12.50% if its fumble recoveries from 2012.
- LSU returns just 21.43% of its sacks from 2012.
It’s not enough for me if I still have one reader with me after all of these numbers and graphs. So I’m taking it a step further. Let’s make an official DudeYouCrazy Returning Defense Formula…just for fun.
We have five ingredients all represented by percentages:
- Tackles for loss
- Forced Fumbles
- Fumbles Recovered
Let’s figure out some weightings.
Tackles for Loss and Sacks are essentially the same thing, but one is decisively more damaging. In 2012, Georgia, South Carolina and LSU combined for 276 tackles for a loss with a resulting yardage loss of 1090 yards. The three teams combined for 110 sacks resulting in a total loss of 702 yards. So, TFLs in this short study pushed teams back about 3.95 yards each while sacks did 6.38 yards of damage. So, in the formula returning tackles for loss should be worth about 60% less than sacks.
Additionally, both of those categories should be weighted significantly lower than actual turnovers (Interceptions and Fumbles). Interceptions and Fumble Recoveries always end drives. So, they are going to be arbitrarily worth 3x the amount of a sack.
As for forced fumbles, they can’t be overvalued as they are not turnovers until they are recovered. But, it takes a forced fumble to have a fumble recovery. So they will be weighted the same as a tackle for loss.
So the impromptu DudeYouCrazy Returning Defense Formula is…
% Returning TFL + % Returning FF + (1.6 x % Returning Sacks) + (4.8 x % Returning Fum Rec) + (4.8 x % Returning INTs)
With that in mind, here’s how the teams shake out:
A disproportionate bend is put on turnovers, but I don’t think that’s an inaccurate emphasis. Playmakers make a defense.
Thank you if you stuck around this long. Real preview articles start tomorrow!
That’s all I got/