How Boom or Bust was the 2013 LSU Rushing Attack?

Jeremy Hill is gone. Right guard Trai Turner along with him. And valuable reserve Alfred Blue will also explore his chances at the next level. The 2014 LSU rushing attack will still return many familiar faces, including 4/5 of a solid offensive line and a couple of backs we've seen great flashes from.

We'll also bring in arguably the most highly touted LSU recruit in history, Leonard Fournette, and another top 200 running back in Darrel Williams. There's plenty good reason to believe that even with the defections, the LSU rushing attack in 2014 could be just as, if not more, potent than the 5.0-ypc, 2,600+ rushing attack LSU featured in 2013.

Even if the passing attack regresses under a younger QB with a new set of receivers, we can generally assume the rushing attack will continue to produce at a high level, as it has most seasons in the Miles era. Passes? We don't need no stinkin' passes.

LSU recently parted ways with long time offensive line coach Greg Studrawa, a move that's been hoped for from some of the LSU faithful for some time. One of the oft-levied criticisms toward Stud was that the offensive lines underachieved during his watch. Though a hefty amount of the criticisms levied his way were likely unfair (let's be honest, you don't produce a 3,000-yard passer and 1,400+ rusher with an AWFUL line), it's also fair to examine the struggles of the line.

Of particular note, in 2013, is that the line really struggled to pick up stunts up front. LSU allowed 25 sacks (some of which can be laid squarely at the feet of Mettenberger), and since the numbers aren't officially tallied by the NCAA, who knows how many hurries and QB hits. For a Cam Cameron offense, one that would like to trend toward more balance, this is clearly unacceptable. Thus Studrawa is sent packing.

Beyond the issues in pass protection, LSU's offense seemed to have difficulty ironing out consistent yardage in the running game. Peruse the Game Threads and reaction pieces here at ATVS and what you'll see are numerous fans (myself included) commenting on the difficulty of the offense to consistently establish the run game. From a distance, it seemed as if the LSU offense was good for a handful of big gainers each game, but also a bevy of three yards or fewer carries in between. Having seen Jeremy Hill produce any number of spectacular runs, we're less inclined this to be a fault of the runner, so much as the offensive line struggling to get a push up front. But is this reality or perception?

My Data

For this study, I want to take a look at every SEC rushing carry LSU attempted in the 2013 season. QB kneels at the end of half/game will be discarded. As will any carry by Mettenberger, since he offers nothing as a rusher, and whatever few designed runs or scrambles he attempted wouldn't be statistically significant. What we're really looking at here is "How consistently did LSU run the ball when they tried to run the ball with their best ball carriers?" The best ball carriers in this study will be every RB and WR who carried the ball in 2013. All data is tabulated from ESPN's Play-by-Play data on ESPN.com box scores.

First let's acknowledge some inherent weaknesses in the data. Carries of 10-plus yards are all lumped together. Considering the pure probability of options, naturally there should be more carries of 10-plus (90 possible ways to get a 10-plus yard carry). Further, the data raises other questions. How much different is a one-yard carry to a two-yard carry, really? Especially when the totals are loosely estimated by nature. It's also irrespective of down and distance in all situations. So a carry from the one to goal cannot possibly be more than a yard and probably skews the data some. For those reasons acknowledging that LSU racked up more 10-plus yard carries than carries of zero and less combined may be slanted.

That said, I don't think the numbers are entirely meaningless. How you define "boom or bust" will greatly determine your perception of this data. If we operated off the assumption that every three-yard carry was a "success" because four consecutive three-yard carries equals a 1st down, then the consistency of the LSU rushing attack is staggeringly positive. 172 of LSU's 258 carries went for 3-plus yards. That's actually a perfect 2/3 of the time that LSU was "successful." The 3-yard carry is, though, the ultimate in-betweener. Three of them is not enough for a 1st down and most coaches, even the aggressive ones, trend toward punting in 4th in short situations, outside of late-game or obvious need for scoring situations.

Now, if we are to dive deeper into the data, I think the argument many made was that we basically hovered between two ranges 0-3 and 7-10-plus, rarely settling into that 4-5-6 with any consistency. The data doesn't agree:

While there were a greater number of carries from 7-10-plus range, the fact that the range of possibility there is 93 different carry types accounts for the discrepancy. As you can see from the first chart, 38 rushes tallied 10-plus yards last season. If anything, LSU proved alarmingly consistent at picking up positive yardage in 2013, netting only 25 total true rushing carries for zero or fewer yards.

But we return to the issue of grounding this data. It's interesting enough in a standalone, but how does it compare to past eras? There's obvious flaws to comparing it to the past, namely that we've employed quarterbacks with legitimate rushing capabilities (namely Jordan Jefferson). There's also the issue of turnover between not just personnel, but coaches. Yet, enough like variable exist for us to draw upon the similarities. Had LSU transitioned from, say, it's current offense to a Rich Rod read offense to a Mike Leach air raid, then we could probably toss out any data due to incomplete comparisons. But the offense has stuck to mostly some hybrid of spread rushing and traditional downhill looks, emphasizing a power rushing attack and down field passing.

So in turn, I aggregated all of the rushing data in the Miles era in the same manner as above. All carries tabulated by yards gained/loss, all kneel downs thrown out, all non-factor rushers also discarded. In this case, Matt Flynn, Ryan Perrilloux, Andrew Hatch and Jordan Jefferson counted as factoring runners, since the option was a legitimate portion of their repertoire. Russell, Lee, and Mettenberger's carries were thrown out. Russell did offer some break the pocket rushing potential, but the aim here is to capture how well LSU rushed the ball when they tried to run the ball with their best ball carriers.

Just by clicking through the charts you can see there's not a major discrepancy visually. Let's take a look at the data in a table form for comparison.

Seeing the numbers aligned gives you a little better perspective by year. You can more easily spot highs and lows across years. For example, 2011 and 2007 were particularly effective years for generating explosive plays (10-plus), while the offense struggled to break big runs in 2006, 2009 and 2012.

The ultimate idea here is to get a feel for mean carry of each season. Remember, we're trying to capture consistency. So let's take a look at this data in two ways. First, here's the actual mean carry for each season since 2005.

Year	Carry Type
2005	10+
2006	3
2007	10+
2008	10+
2009	2
2010	2
2011	10+
2012	2
2013	10+

So I had really hoped this would be useful, but I'm afraid it's not really. Half the seasons of the Miles era there were more carries of 10-plus yards than any other. As discussed earlier, the nature of how I tabulated the data makes it statistically more probable to aggregate 10+ yard carries. Though, it's quite interesting that's not the end result in every single season. When I discard the 10+ category for those years, it turns out as follows: 2005 (1 & 2 tied), 2007 (2), 2008 (3), 2011 (2), 2013 (1).

So essentially all we're seeing is that the type of carry easiest to aggregate (10-plus) and the type of category easiest to accomplish (1-2 yard carries) are the most regular. From the standpoint of determining a "consistency" this isn't overly helpful. So let's group the data a bit more collectively, using the groupings from earlier 0-3, 4-5-6, 7-10-plus.

Grouping gives us a little better picture. Let's look at these aggregates by percentage of total carries in a given year.

Unsurprisingly, most carries in a given season are of the "plodding" type (0-3). Discarding that, the variations in the other two types give us a picture of offenses that were able to more regularly generate big plays and those that weren't. 2007, 2009, 2010, and 2011 were offenses more capable of eating up huge chunks of yardage on the ground. 2008 is at a dead heat. 2005, 2006, 2012 and 2013 were offenses that couldn't turn 4-plus runs into 7-plus runs. Is that an issue of personnel? Play calling? Pure dumb luck? In the case of 2006 and 2013 we're talking about the difference in 7-8 carries. It's not uncommon for shoestring/grass-monkey tackles, especially not 7-8 plays which would turn potential 7-plus-yard carries into 5-6 yard gains.

What Does This All Mean?

In short, I'm not sure. Football Outsiders uses some advanced metrics to measure the "Success Rate" of an offense, but it's not broken out into run vs. pass (that I can find). To this point that's the only other methodology that I know that could possibly offer insight into what we're trying to tabulate here. So while I'm not certain about any of this, here are some "ideas" I have that may be takeaways.

1) Running QBs are Better For Big Play Rushing Attacks

That is, at least they have been under Les Miles. Look at the four seasons that produced more 7-10-plus carries than 4-6-yard carries: 2007, 2009, 2010 and 2011. Every single year there was a QB that not only possessed the ability to scramble, but was factored into the rushing attack through the option.

To be clear, this doesn't always translate into overall offensive success. Of those seasons, only 2007 and 2011 could be considered good offenses, and 2011 only marginally so (50th overall in Yards Per Play).

2) There is No Correlation Between Good/Bad Passing and Good/Bad Rushing

We often hear of "running to set up the pass" or "passing to set up the run" and what I'm finding in this data is that there may be zero correlation between the two. To me, for either of those to be true, at least in a larger sense, both portions of the offense would equally lift one another. So if you pass well, you ran well, or if you ran well, you passed well. I could be entirely incorrect in this assumption, but what the data suggests is that these may be entirely unrelated.

Take 2006, for example. It's one of the only two 3,000-yard passing seasons in the Miles era. It also produced the second fewest 7-plus yard runs in all of Miles tenure. In 2010 LSU barely topped the 2,000-yard mark, but also generated the 2nd highest number of 7-plus yard runs in the Miles era. In 2013, we see both: a 3,000-yard passer and a high number of big rushing plays. So three samples: 1) Throw well, don't run big. 2) Don't throw well, do run big. 3) Throw well and do run big.

3) The 2013 LSU Rushing Attack is Not Boom or Bust

A mere 10 percentage points separates the number of carries LSU totaled from 0-3 to the number of carries LSU totaled from 7-plus. It's by far the most evenly distributed of any season in the Miles era. For me, to see a boom or bust quality, I'd need to see a heavy number of 0-3 and a heavy number of 7-plus with an almost minuscule total of 4-5-6 yard carries. Instead what we see is a mere eight carry difference between 4-5-6 carries and 7-plus.

To me this exhibits that the LSU rushing attack was not only not boom or bust, they were the most stable and consistently successful rushing attack in the Les Miles era. In 2013, only 25 carries went for 0 yards or fewer. The second lowest season in the Miles era was 44 in 2005. In 2013 we managed to almost completely eliminate the negative rushing play, which is amazing in and of itself. Fewer than two rushes a game were going nowhere or backward. That's astounding.

We often put feathers in Cam Cameron's hat for bringing the LSU passing game to life, but we've all ignored, just a bit, how successful of a rushing attack he also engineered. So, from my view, by what the numbers say, the rushing attack was not boom or bust... our eyes deceived us. The 2013 LSU proved to have one great trait: the ability to go forward.