Stay Positive: Using ATVSQBPI2 to Rate a Sophomore QB's Second Year

We talk a lot around these parts that there's no need to panic over Jordan Jefferson.  He's just a sophomore, and sophomore quarterbacks tend to improve.  Of course, we say that and it feels like it should be true, but I started to ask myself: is that really true? 

So I went on over to the indispensible cfbstats.com and pulled all of the underclassmen QB starters in the SEC over the past five years and then how they did as juniors.  Why five years?  Because their stats only go back to 2004.  Hey, at least I'm honest. 

I limited the study to any quarterback who averaged over 100 yards/game, which would at least indicate significant playing time.  Not everybody started every game, but if you're averaging that many yards/game, you saw a lot of live action.  Last year was actually a banner year for sophomore starters, as four quarterbacks qualified for the study: Jordan Jefferson, Ryan Mallet, Stephen Garcia, and Larry Smith.  Obviously, since none have played their junior season yet, we can't look at how they do relative to their "rookie" year for a few more months.  But it bears watching.

From 2004-2008, the five year period we're going to use, there have been 18 underclassmen who met the 100 yards/game benchmark.  It bears noting, only two freshman qualified, Jarrett Lee and Wesley Carroll.  Both failed to keep their starting gig.  In fact, one of the big surprises is not that 18 underclassmen were their team's nominal starter from 2004-2008, but that of that group, seven of them did not return as the starter.  That leaves us with exactly 11 quarterbacks who qualified for our study.

The seven who didn't keep their jobs and the likely reason why they lost it: 

Syvelle Newton (USC 2004), ATVSQBPI2 - 5.936: Spurrier.  Moved to wide receiver.
Ethan Flatt (OM 2004), ATVSQBPI2 - 5.331: A case of the sucks.
Wesley Carroll (MSU 2007), ATVSQBPI2 - 4.777: Hiding under the bed after the abuse suffered.
Chris Smelley (USC 2008), ATVSQBPI2 - 5.117: Spurrier.
Mike Hartline (Kentucky 2008), ATVSQBPI2 - 4.767: Randall Cobb's sort of emergence. 
Kodi Burns (Auburn 2008), ATVSQBPI2 - 5.022: A new offense installed.
Jarrett Lee (LSU 2008), ATVSQBPI2 - 5.629: Interceptions.  Lots and lots of interceptions.

This is Year Two of seeing so many sophomores start, and it didn't turn out well in 2008 to 2009.  Only Jevan Snead kept his job, and he wasn't a real success story.  So it's an open question how the four sophomores from 2009 will do in 2010.  Let's look at the 11 who qualified for the study.

THE OUTLIERS: Tim Tebow and Omarr Conner.

Neither guy is very helpful to the study, for opposite reasons.  Tebow came in and as a sophomore, posted an ATVSQBPI2 of 9.109.  That is off the charts good.  Omarr Conner started for MSU in 2004 and 1005, posting ATVSQBPI2's of 4.648 and then 4.256.  Some guys are just great and some guys are just terrible, and experience is irrelevant.  We can't really learn lessons from either of the two.

Now, let's get down to business.  First, let's look at the sophomores from last year.  Larry Smith seems to fall in the Omarr Conner group, putting up a dismal 4.003 while Ryan Mallet was already awesome as a sophomore, posting a 8.731.  Yes, he's already in the Tebow group.  Yes, that is depressing.  Our two "normal" quarterbacks are Jordan Jefferson and Stephen Garcia.  Jefferson's efficiency balanced with his penchant for taking sacks rates him at 6.096 yards/play.  Garcia was far less accurate but gained more yards and was more prone to big plays, rating at 5.739 yards/play.  Both QB's are about even, with Jefferson coming out slightly ahead. 

So, looking at past SEC quarterbacks, are these two likely to improve, decline, or stay the same?  Unfortunately, if history is any guide, the answer is an unequivocal "we don't know".  In fact, a sophomore starter seems just as likely to decline or hold steady as he is to improve in his second year, and that's not even counting the guys who lost their starting job.

GROUP 1: DECLINERS
(Chris Leak, Jevan Snead, Chris Nickson)

Of our 9 quarterbacks, a full third saw their average yards per play decrease by at least one full yard.  Leak saw his average go from 7.687 to 6.388.  The encouraging sign is that Leak improved his efficiency, threw less picks, and scored at about the same rate.  His decline really did come from throwing for less yards/attempt, and poor running numbers.  Snead dropped from 7.844 to 6.488, primarily caused by a TD/INT ratio going from 26/13 to 20/20.  He simply stopped taking care of the football.  Nickson went from 6.550 to 5.403, but his decline was almost entirely due to his rushing numbers.  Nickson went from 146-694-9 to 58-178-2.  He also had a negative TD/INT ratio in both seasons.

GROUP 2: IMPROVERS
(Andre Woodson, Matthew Stafford, Blake Mitchell)

While only three sophomores improved by at least a yard/play in their junior year, two of the quarterbacks who did improve, improved by leaps and bounds.  Stafford went from a respectable 6.783 to a terrific 8.491.  He threw for about 1.7 yards/attempt, improved his completion percentage to above 60%, took less sacks, and even slightly improved an already excellent TD/INT ratio.  Quite simply, he blossomed into a star.  But the truly remarkable case is Andre Woodson, who went from 4.745 to 7.787, an improvement of 3 yards/play.  He went from a liability to one of the top players in the SEC.  He doubled his yardage, improved to a 63% completion rate, and improved his TD/INT ratio from 6/6 to a whopping 31/7.  He still took sacks, but this was the look of a player growing up.  The third quarterback, believe it or not, is the oft-maligned Blake Mitchell.  Spurrier would eventually bench Mitchell for "ineffectiveness", but he improved from 6.222 to 8.075, which is pretty good.  Mitchell completed over two-thirds of his passes, and he dramatically cut down on his sacks taken.

GROUP 3: THE HOLD STEADY
(Brandon Cox, JaMarcus Russell, John Parker Wilson)

Of our group Stuck Between Stations, one of them was actually just waiting a year to explode.  JaMarcus went from 6.424 to 6.696.  But that is actually Russell going from his freshman to sophomore years.  After one and half years of starting, he took the job as a junior and improved from 6.696 to 9.165.  He could just as easily be in the group listed above.  Russell never threw many picks, but he saw his TD's increase from 9 to 15 to 28.  He saw his completion percentage improve by nearly 10 points a season.  He even saw his rushing numbers steadily improve.  He simply got better in every single category.  Cox (6.454 to 6.370) and Wilson (6.139 to 5.617) both slightly declined, lending credence to the theory if the guy doesn't have his lightbulb moment before his junior year, it simply isn't going to happen.  There has to at least be some improvement in the second season as a starter. 

So what can we take from this?

Well, Jefferson can really go any which way, as can Garcia.  Mallet and Smith are likely already the players they are going to be, but players who post an ATVSQBPI2 around 6.0 are a good candidate to improve.  Of the three decliners, two had yards/play above 7.0, and the third was a running quarterback. 

The most likely scenario is that Jefferson either improves or holds steady at his established level of performance.  He already has a high completion percentage and a good TD/INT ratio, so really, the best way for him to improve is either slight improvement in yards/completion or, of course, to stop taking sacks. 

Jefferson has already cleared the first hurdle: keeping his job.  Now, all he has to do is top the performance of Blake Mitchell.  That doesn't seem so hard when you say it like that, huh?