O-Line Numbers or Why LSU's Offense Sucked in 2009

If there is one thing I enjoy around these parts, it's inventing stats.  The ATVSQBPI2 is not only a fun tool, it's pretty useful.  We can make comparisons week-to-week and season-to-season.  It allows us to compare QB's in a spread or a west coast offense to ones in a vertical passing game.  We're pretty proud of it, and you'll be seeing it throughout the year.

ED. NOTE - If you want to skip the long boring introduction, with math, just go straight to the jump and we'll start rating the lines.  All you need to know is that 0 is perfectly average.  Positive numbers good, negative numbers bad.

I'm also a huge believer in line play.  Simply put, games are won up front.  Without a good line, there's no time for the quarterback to throw and no holes for the running back to run through.  It's been one of my quests to develop some sort of reliable measurement for line play, given its importance to building a winning team.  I'd like to say I have the absolute perfect stat, but I don't.  I'd say I'm about 75% satisfied with my OLINE metric, but I'll take an admitted flawed number over absolutely no data whatsoever.  My advice: pay attention to the underlying numbers.

I rate the offensive line on two basic skills: pass blocking and run blocking.  We can look just at total sacks because that would reward teams for not passing as much: the key is the rate at which a team allows sacks.  Instead, I rank teams on sacks allowed per 100 modified pass attempts (attempts + sacks).  I rank rushing on the familiar yards per attempt metric, once again so we're looking at rates not total yards.  Since a line has already been punished for sacks, I remove sack yards from the rushing totals and sacks allowed from rushing attempts.

Both rates are on slightly different scales, so I use standard deviation to come up with an RSCORE and a PSCORE for rushing and passing, respectively.  Add the two numbers and multiply by 100 for the OLINE score.   Now, I like the stat and it gives us a pretty fair rating of lines.  There are a few drawbacks which I freely admit:

1.  It's nice useful on a game to game level.  It's an end of the year stat only.

2.  Since it's standard deviation, the stat is dependent on the rest of the league. 

3.  The stat can't really be used to compare across seasons. 

4.  Yes, a quarterback is partially responsible for the number of sacks allowed and running backs are partially responsible for their yards/attempt, but we've been given those guys credit for the work of the offensive line for years, so turnabout is fair play.

In the end, I'm kind of happy the OLINE stat is so dependent on others.  Metaphorically, it's perfect for offensive line play. 

OK, long boring introduction over.  Let's look at precisely how friggin' terrible LSU's line was last season. 

LSU allowed 37 sacks last year.  That sucks.  South Carolina also allowed 37 sacks, which also sucks, but they allowed those 37 sacks in over 100 more pass attempts.  As bad as USC's line was at pass blocking, it wasn't even on the same planet as LSU. 

Georgia had the best pass blocking unit in the SEC, allowing only 3.468 sacks per 100 pass attempts.  The average SEC line allowed 6.391 sacks per 100 attempts.  LSU? 11.012.  That's nearly twice as often as a league average line and over three times as often as Georgia's. 

Hey, but there's always run blocking... 

Oh.

Well, at least LSU wasn't last in run blocking.  South Carolina and LSU are both hoping to make big noise this season, but both teams need to show huge improvements on the line.  Both were the two worst lines in both run and pass blocking.  That's not a good sign.    

Florida, who had a dismal pass blocking unit, had a great run unit, but it's an open question how much of that was Tebow.  We'll find out this year.  Auburn is banking on their line this season, and it's easy to see why given their efficiency last year.  The real surprise is that Alabama was only #3 in run blocking, given the two studs they have in the backfield to help out the 'ol yards per attempt.

Let's put it all together...

How bad is -336.78?  I have OLINE stats going back five years, and the worst OLINE number posted was -229.48 by Kentucky in 2006.  LSU's 2009 offensive line was historically awful.  It wasn't bad for LSU or just comparatively bad, it was the absolute worst offensive line in the SEC over the past five seasons.

You can't win squat that way.  The news gets worse.  Teams that have posted a -200 OLINE in one season, as a group, do not improve much.  However, I've never seen a team go from above average, like LSU was in 2008, to absolutely terrible the next season.  

The key to 2010 is up front.  Let's not sugarcoat it.  The offensive line absolutely must improve for this team to win anything this year.  I think they can improve, but let's be clear where we are starting from.  The 2009 LSU Tigers had the worst offensive line in the SEC in half a decade.  That is simply unacceptable. 

Last year is last year.  Time to put last year behind us.