Anatomy of an -ffense: Tennessee Volunteers in 2008 (so far)

If you've learned anything about me at this point in time, it's that I like to play around with numbers.  My first-ever RTT post came last year before the Arkansas game, where I gave Joel some charts to show that the magic mark for beating Arkansas was to hold McFadden under about 150 yards.  In fact, it's been a little frustrating for me during these last few months because there have been very few numbers to tinker around with; having a new offensive coordinator and a completely unknown system will tend to do that to a guy.

But now that we have 4 games under the belt (a third of the season already!  Well, 33.333%; repeating, of course...), we have some honest-to-goodness numbers to examine.  Naturally, they're not exactly the numbers we had hoped to look at, but it's better than nothing.  (Well, sorta.)

I had hoped to be able to do some fancy analysis of the Clawfense, but I can't honestly do that.  The offense has been so mistake-prone and ugly, that I don't think we can rightly say we've even seen the Clawfense.  (After all, how many offenses never run any kind of a misdirection play?  The only one I remember was the dropped pass from the G-Gun against Auburn.)  So instead, I'll look at the basic numbers: QB performance and a little run/pass analysis.

So without further gab, let's take a look at a few of the stats that UT's offense has accumulated to date.  Pretty pictures and tables follow after the jump, but there's no fancy mathy stuff to worry about.  We'll keep it straightforward because I don't think we can honestly do anything more.

Oh, and the first chart is ugly - all messy and spaghetti-y.  I intentionally violated a rule of chart-making (don't put too much on one chart) because I don't plan on using it a whole lot.  Explanation will follow, so don't panic if you're not a numbers type.

Qb_charts_i_medium

(Ok.  Remember, kids: don't put 8 different plots on the same chart unless you're only making generalizations.)

In the first chart, you see the QB ratings for the Vols for the first four games of 2002 - 2008.  You also see the QB ratings for the Vols during the 4-game losing streak in 2005 (labeled as 2005 Worst 4).  2008 is the thick black line.  The NCAA QB ratings tend to run higher than the NFL ratings, so I kept all the lines to help give a sense of perspective.  But you'll notice that 2008 keeps company with 2 other lines.  Any guesses where the other lines belong?  If you guessed 2005 (1st 4 games) and 2005 (4-game losing streak), you'd be one heck of a bitter Vol.  And you'd be absolutely correct.  In fact, the second chart is a close-up of those three 4-game series.  Oh, that game against Auburn this weekend?  Worst QB rating for a Vol as far back as 2002.  2nd worst is 2004 against Auburn.  Yeesh.

In the second chart, you see the 4 worst showings for QB ratings:  2008, 2005 and 2005 (4-game losing streak).  It's a little easier to see in this graph than the first graph that UT QBs tend to improve over these 4-game spans, except for 2008.  (Again, ugh.)  But contrary to the immediate opinion, Crompton's ratings are not worse than those in 2005 - at least not measurably so.  They are effectively equivalent, even if you factor in the level of competition for those 12 games.  So by the NCAA QB rating metric, we have seen a passing game roughly equivalent to the 2005 season.

In the third chart, you see the percentage of offensive playcall that were pass plays.  For example: in 2007 Game 1, over 60% of UT's plays were pass plays.  What you can see here is that Clawson is calling fewer pass plays than what we've seen over the last 3 previous years.  This is no surprise: Clawson has always run about 55% of the time, while Cutcliffe preferred the quick passing game and 2005 was basically a year-long QB audition.  Oh, time for a quiz question!

QUIZ QUESTION!  QUIZ QUESTION!
How can you tell, by the passing playcall percentage, when UT was playing against
SEC opponents and when UT was playing against nonconference opponents?

The answer is after the pretty picture:

345986669_2f6b750bd7_medium

Vols fan: Approx. 4:30 PM EST Saturday (via mandj98)

Answer:  In SEC games, UT has a higher pass-play percentage than in non-SEC games.    This holds true even for non-Florida games.  In fact, this trend is iron-clad all the way back through 2002 except for Florida 2003, where UT won and pounded Florida off the field.  I was rather surprised to see just how consist... wait, what? 

Are  you serious? 

Ok, let me take a look...

...

Well, I'll be!  In 2008, UT saw a steady decline in the passing ratio, even in nonconference games.  In fact, the trend is very linear with the exception of the UAB game.  When you remember that the UAB game was about a kajillion degrees and UT spend the second half killing the clock, that makes sense.  If we could somehow factor that out of the equation, I think you'd see a consistent decrease in the pass playcalls in 2008.  Not even 2005 shows such a trend.  For those of you who are looking for evidence of a reduction of Crompton's responsibilities in a game, here you go.

Alright.  There's more in those charts, but those are the major finds.  The rest is small stuff, so let's move on to other pretty charts.  Let's compare a few QB stats directly.  The following 2 charts show Yards-per-Attempt (YPA), Completion percentage (Comp. %), and QB Rating, comparing 2008 to various years.  Note that YPA is multiplied by 10 to make it easier to read.  No, UT was never getting 50 yards per pass attempt.  If they were, we wouldn't be having this discussion right now.

Qb_charts_ii_medium

In the first chart, you see the overall values of 2008 compared to the beginning of 2005 and the 4-game losing streak of 2005.  That is the statistical definition of equality, right there.  In short, 2008 QB play is equivalent to much of 2005's QB play.  In 2005, UT was waffling between Clausen and Ainge for much of the season (in fact, the 4-game losing streak started with Clausen and ended with Ainge), so inconsistency might not be surprising. 

In the second chart, you see 2008 compared to the Ainge-Cutcliffe years.  You don't need a numbers freak to tell you that there's no good comparison there; 2008's passing performance doesn't come close to previous performance.

Now, you didn't need a pretty chart to tell you that QB play this year hasn't been a pretty picture.  But it does help give a sense of perspective.  Obviously, something is amiss with the passing game, and general observation will put most of the blame on Crompton's inaccuracy so far.  Can he recover?  Will he even be given a chance?  We'll find out.  We just know that it's not up to par.

Ok, moving on.  Here is the last set of charts (tables, really) that I'll post here.  These are passing and rushing stats for the first 4 games for 2008 through 2002, and those dreaded 4 losses in 2005.  The games that UT lost are highlighted in red.  (In 2002, I guessed at the QB because the stats start to flake out there.  Going further back would have required intensive research (ie. non-internet stuff) so I stopped there.)  I'll put a few observations in bullet points below, but there's more to be noted than what I say.  Feel free to add in the comments anything else that stands out.

Ut_offense_charts_-_1st_4_games_medium

A few observations:

  • The rushing game has improved considerably.  The rushing stats in 2008 are ecplipsed only by the start in 2004.  But 2004 had considerably better QB play to keep the defenses honest.  We're a running team again, folks, and it's a great running attack!  Now we just need to have enough passing to keep opposing D's from stacking 9 in the box.
  • Erik Ainge had better stats than you thought.  Ainge's first four games were pretty good overall in 2006 and in 2007.  I know there was a lot of frustration with him when he was here, but he does deserve his due on this one.  Thanks, Erik.
  • 2008 was the second-best rushing effort against Florida since 2002.  2004 was the best, and we won that game.
  • The only other year UT started with more INTs than TD passes was, gulp, 2005.  Enough said.
  • We won't be successful with the run unless we can pass.  Duh, I know.  Buy you can see that the running production dropped off significantly in SEC play.  Part of that is due to playing better defenses, but part is due to those defense playing the run and daring Crompton to pass.
  • The number of offensive plays has been steadily decreasing in 2008.  Against UCLA, UT ran 41 passes and 34 rushes.  Even if you subtract the 3 OT plays, that's an Enduro compared to the 23 passes and 33 carries against Auburn.  The clock is flying, folks, and the team can't expect 4th-quarter comebacks.  Conversely, if UT has a lead, the rushing game should be good enough to kill the clock.  This again emphasizes the need for mistake-free football.

Well, that's probaby long enough for one post.  There's more in the data than what I outlined, but I'm sure Joel's eyes are glazing over right now.  If you see anything else, let me know.  If you see any glaring errors, let me know.  If you need a shoulder to cry on, well, I can't help you there.  My wife would get awfully suspicious...

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Rocky Top Talk

You must be a member of Rocky Top Talk to participate.

We have our own Community Guidelines at Rocky Top Talk. You should read them.

Join Rocky Top Talk

You must be a member of Rocky Top Talk to participate.

We have our own Community Guidelines at Rocky Top Talk. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9347_tracker