And The Valley Shook: An SB Nation Community

Navigation: Jump to content areas:


Pro Quality. Fan Perspective.
Around SBN: Has Kentucky Improved Since the Non-Conference Season?

Rating the Lines

I’m a stats guy.  I like the objective measure of things that have happened.  Now, you don’t want to go overboard and I’m a fan of Benjamin Disraeli’s adage “there are three types of lies: lies, damn lies, and statistics.”  (No, Mark Twain didn’t say that) 

But I think a lot can be learned from looking at the numbers.   Our lying eyes sometimes do lie, and we often remember just the data which supports our own hypothesis.  I don’t think it’s malicious, but I think without stats, we tend to just talk out of our asses.  It’s just always important to remember stats merely told us what happened, not what is going to happen.  They can give us a good idea, but just because a running back went for 100 yards against a team last year, it doesn’t mean he will again.

The biggest problem with the statistical record of football (of which there are many) is offensive line play.  There simply isn’t a stat which measures line play in one quick and dirty reference.  So lines get praised or derided based on little more than press clippings and a few highlights here and there.  I’m trying to add something to the debate, even if the metric isn’t perfect. 

I start with a simple premise.  An offensive line has two jobs: (1) protect the quarterback and (2) open up holes for the running game.  A line can be good at one task and not the other.  For example, Tennessee allowed four sacks last year.  Four.  That’s pretty damn awesome.  But, that same line had problems run blocking.  I wanted a metric that took both of those skills in account. 

So, for pass blocking, I looked at sacks allowed.  But sacks alone don’t tell the story, it has to be tied to pass attempts.  A team shouldn’t be rewarded for simply running a lot.  So I divided the number of sacks by pass attempts.  And to make the number look nicer, I multiplied it by 100.  So the number is sacks allowed per 100 pass attempts.  The average SEC team had 410.9 attempts, allowed 21.7 sacks, and had a sack/100 attempts of 5.29.  Here’s how the SEC stacked up:


Tennessee

0.749

Florida

3.601

Georgia

4.110

Arkansas

4.153

Vanderbilt

4.789

AVERAGE

5.294

Alabama

5.297

Mississippi St

5.645

LSU

6.787

S Carolina

6.998

Auburn

7.022

Ole Miss

7.179

Kentucky

7.197



Rush blocking was even easier: yards per attempt.  That’s it.  In case you were wondering, the average SEC team had 496.7 attempts at 4.32 yards per carry.  Here’s the SEC in rush blocking:

Arkansas

5.96

Florida

5.34

LSU

4.90

Georgia

4.53

AVERAGE

4.32

Tennessee

4.24

Kentucky

4.17

Ole Miss

4.16

Alabama

3.96

Auburn

3.76

Vanderbilt

3.75

S Carolina

3.57

Mississippi St

3.51



Now, certainly, I understand some of the flaws of the system.  Arkansas, for example.  How much of their gaudy 5.96 yards/attempt is a function of having Jones and McFadden?  Well, that’s simply unknowable.  But we’ve been given running backs credit for accomplishments without making a discount for offensive line play, so I see no reason why we can’t do the opposite.  All of the performances in football are inter-related.  A quarterback’s completions are a function of their receivers as well.  Isolating performance is simply impossible and I’m not going to try. 

Secondly, simply adding the pass rating with the rush rating together would not work because they are on different scales.  And what does the difference in sacks per 100 attempt really mean?  I don’t know.  So I made a tweak to put everything in the same scale and also make sense of these two lists of numbers. 

I computed the standard deviation and rated each team based on the standard deviations from the mean.  It gave me a sense of how much a deviation really mattered and it placed the run blocking and pass blocking on the same scale.  And then I multiplied each number by 100 for ease of reading.  Simply adding the two scores together gave us an overall OLINE score for each team. 

Is this perfect?  Of course not, but it at least gives me an idea how each offensive line performed.  And a slightly flawed metric is better than nothing at all.  Here’s the SEC OLINE Ratings:

TEAM

PSCORE

RSCORE

OLINE

Arkansas

0.610

2.289

289.922

Florida

0.906

1.427

233.339

Tennessee

2.432

-0.113

231.964

Georgia

0.634

0.288

92.170

LSU

-0.799

0.807

0.821

Alabama

-0.001

-0.504

-50.547

Vanderbilt

0.270

-0.801

-53.075

Ole Miss

-1.009

-0.219

-122.776

Kentucky

-1.018

-0.214

-123.234

Mississippi St

-0.188

-1.129

-131.647

Auburn

-0.925

-0.786

-171.141

S Carolina

-0.912

-1.046

-195.813

 

A score of zero would be perfectly average.  Which is about where LSU is, a fairly surprising result.  Which could speak to the methodology, but we really did let up a lot of sacks.  Anyway, looking at each team:

Star-divide

Arkansas An interesting case, as their gaudy 5.96 yards/carry completely skews the rankings. Only four teams have an RSCORE above 0 because the Hogs so raised the average. They even did a really good job in pass blocking, allowing only 4.15 sacks per 100 attempts. Considering 4 Hogs were 1st or 2nd team All-SEC, I think we’ve identified the SEC’s best line last season.

Florida Great balance. Florida’s line was both terrific in protecting the quarterback and in run blocking. Even more amazing is that running quarterbacks tend to get sacked more (I know, it’s counter-intuitive), but Tebow didn’t get sacked a lot. Maybe it’s because he’s got the body type of a stegosaurus.

Tennessee Kind of an illusion. The pass blocking was brilliant as they allowed only 4 sacks in 534 attempts. No team passed more often, and no one got sacked less. The run blocking was decidedly mediocre, but the pass blocking was just so amazing that they end up near the top of the rankings.

Georgia Quietly above average in both phases of the game. No one really throws out UGa when talking about the best lines in the SEC, but they probably belong in the discussion just for their quiet consistency.

LSU Is our line overrated? The run blocking was really good, but the pass blocking? 30 sacks allowed puts LSU near the bottom of the conference. It’s also not like Flynn was a statue back there.

Alabama About as average as a team can be. Which isn’t meant as an insult.

Vanderbilt Their slightly above average pass blocking makes up for some of the worst rush blocking in the league. Which makes sense, as the knock on Vandy has always been they can’t recruit those mammoths other SEC schools can, but they at least have good technique.

Ole Miss If Michael Oher is so great, how come Ole Miss doesn’t at least have a league average offensive line? Just asking.

Kentucky A prolific passing team that gave up a ton of sacks. Before you think this system is biased against passing teams, remember Tennessee had even more pass attempts and their PSCORE is ridiculously awesome. They throw a lot. They get sacked a lot. And they get sacked at a high rate.

Mississippi St. 3.51 yards per carry. Ouch. It’s almost like they aren’t trying. And State used to have really good offensive lines. What happened?

Auburn OK, here’s the big surprise. Auburn’s line was bad every way a line can be bad. If you want to point to one thing which led to Auburn’s disappointing year, it’s probably this. And, as a huge believer in line play, if Auburn can’t fix this, they aren’t going to Atlanta.

South Carolina Honestly? Simply awful.

0 recs  |  Comment 4 comments

Story-email Email Printer Print

Comments

Display:

The good news is,

the weakest link of the line, Carnell Stewart, is gone. I know he was good in run blocking but he probably committed more than his share of O-line penalties and gave up more than his share of sacks.

If we are going to go 10-2 this year, our O-line numbers will need to be better.

by LSU Jonno on Aug 16, 2008 11:19 AM CDT reply actions   0 recs

What a difference a year makes...

Great post – but you’re describing the line that was, not the line that is. If they stay healthy, they will be considerably better than last year. And last year they were pretty damned good. Average in the SEC is pretty above average IMO.

lsutigerbait.blogspot.com

by TigerBait on Aug 17, 2008 7:48 PM CDT reply actions   0 recs

Next year

Well, that’s the problem with any stat. Every stat only covers what has happened and not what will happen. I think our line will improve, but I am surprised our line was statistically average because it was treated as a well-above average line.

by Poseur on Aug 17, 2008 8:04 PM CDT up reply actions   0 recs

Nice post...

I’ve been doing a lot of stat wonky stuff at Rock M Nation and SMQ, and I wanted to recommend another way to look at the OL’s impact on rushing yards (via Football Outsiders): line yards. Here’s the definition:

Line Yards (or, Adjusted Line Yards) – Statistic which attempts to, even to a small extent, separate the ability of a running back from the ability of the offensive line. Line Yards are average rushing yards per play by running backs only, adjusted in the following way:

    * 0-4 yards: 100% strength
    * 5-10 yards: 50% strength
    * 11+ yards: not included
    * runs for a loss: 120% strength

It’s a pretty cool way to look at things. If Darren McFadden busts off an 80-yard gain because he’s Darren McFadden, the OL doesn’t get 80 yards credit—it gets 7.5 yards. It’s definitely not perfect, but it’s a great way to separate the credit a bit.

However, you need play-by-play level data to do that, so I thought I’d just spit it out here for you:

LINE YARDS PER GAME

1. Arkansas (152.3)
2. LSU (135.8)
3. Florida (126.5)
4. Georgia (112.1)
5. Kentucky (112.0)
6. Auburn (111.6)
7. Alabama (106.9)
8. Vanderbilt (105.5)
9. Mississippi State (102.6)
10. Tennessee (92.6)
11. Ole Miss (91.4)
12. South Carolina (86.3)

LINE YARDS PER CARRY

1. Florida (3.47)
2. LSU (3.27)
3. Kentucky (3.25)
4. Arkansas (3.23)
5. Ole Miss (3.12)
6. Alabama (2.99)
7. Georgia (2.95)
8. South Carolina (2.95)
9. Tennessee (2.85)
10. Auburn (2.81)
11. Mississippi State (2.74)
12. Vandy (2.73)

LINE YARDS PER CARRY, CLOSE GAMES ONLY (i.e. while the score was within two possessions or less)

1. Florida (3.57)
2. LSU (3.25)
3. Kentucky (3.23)
4. Arkansas (3.14)
5. Ole Miss (3.12)
6. Georgia (3.06)
7. Alabama (2.96)
8. Mississippi State (2.90)
9. Vandy (2.88)
10. Auburn (2.79)
11. South Carolina (2.76)
12. Tennessee (2.73)

If you have any questions or need any other data, either shoot me a message at RMN or e-mail me at billconnelly1 @ gmail . com.

http://www.rockmnation.com
Thrust nunchuk upward!

by Bill C. on Aug 25, 2008 8:00 AM CDT reply actions   0 recs

Comments For This Post Are Closed


User Tools

Welcome to the SB Nation blog about LSU Tigers.
Start posting about the Tigers »

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

FanPosts

Community blog posts and discussion.

Recent FanPosts

Fiddler_on_the_roof_fiddler_1__small
From RockyTopTalk.com-Tennessee reportedly loses J.C. Copeland to LSU
Small
If you could bring back one Tiger from the 2000's for next year's team, who would it be?
Alabama_logo2_small
Really ESPN?
Got_les_small
NSD Notice to LSU Recruits: This Is the Real Les Miles!
Small
GIANT slap in the face ESPN
Got_les_small
Does Negative Fan Speak Really Impact Recruiting?
Got_les_small
Dooley to UT, Crowton to La Tech??
Lsu_small
I Am Absolutley Dumbfounded By This
Small
ANOTHER TIGERT CHOKE STARTS A YEAR
Small
2007 outback bowl anyone?

+ New FanPost All FanPosts >

ATVS Twitter Link

follow Les on Twitter


Managers

Pb140006_small PodKATT

Editors

Gse_multipart30441_small Richard Pittman

Me_and_beer_small Poseur

Lsu_logo_small Purple Reign

Photo_small Dane Noble

Authors

Dduzcaz8lg8xca1e9umxcayb6h89casec7jycanko951caqq3246cavdxhrccam82axacat2kr9qca2oz3rsca64takzcaw0toprca4hx60mcahrmqq0cad7ixvgcayr1dn9cahuxjuicaeieutgca3t3udbcaeieq7pca2s0zi6_small Tortfeasor

Koolaid_small Jrlz

Official Partner of CBS Sports