How (Un)Healthy Is Connor Shaw?: A Statistical Analysis

After Connor Shaw left Saturday's game against the UAB Blazers and subsequently revealed that he had a broken bone in his throwing shoulder, I assumed that there was absolutely no way he would be ready to go for the SEC home opener against the Missouri Tigers. Apparently I was wrong. After briefly addressing D.J. Swearinger's suspension in his weekly press conference, Steve Spurrier said, in a very matter-of-fact manner:

Connor threw the ball around well last night. He'll start. He's fine... No issue on who the starter is this week.

My first reaction was that this is a horrible decision. Honestly, I couldn't believe that Connor was even allowed to play a down in the UAB game, given that he has a two-week-old fracture in a bone that is a vital cog in his ability to chuck an oddly shaped ball through a maze of 215-pound man-children. That requires an unbelievable amount of precision, and any mechanical weakness in his throwing motion is likely to have an adverse effect on the accuracy and velocity of his passes. This is, of course, without considering the potential risk to Shaw's health by continuing to have him play. I have no medical knowledge to use as a basis to question the decision-making here (the medical staff has told Spurrier there's no risk of making the injury any worse), but the fact that the extent of Shaw's injury was concealed until Saturday night and that he was permitted to play in a game that we would have (and did) win easily without him causes me to question the judgment of the decision-makers here.

But Connor Shaw will be the starting quarterback on Saturday, whether that knowledge has quadrupled my annual Pepto-Bismol budget or not, so I set out to determine just what percentage of the healthy Connor Shaw we currently have. I started researching for this post expecting to confirm my belief that Connor's injury has so diminished the quality of his performance that starting Dylan Thompson should be a no-brainer, but then a funny thing happened: the numbers and a repeat viewing of Shaw's performance against UAB told a different story.

First, let's look at what we got last year from a fully healthy Connor Shaw. Number 14 showed in 2011 that he can be a highly efficient passer, completing 65% of his 188 pass attempts in 2011, good for 23rd in the FBS and 2nd in the SEC (behind only A.J McCarron). This was a skill that he displayed from the moment that he stepped on the field (62%, 5.94 yards per attempt through his first 7 games), but it wasn't until the final three games of the season that he displayed the ability to make explosive plays in the downfield passing game while becoming even more efficient (75%, 11.9 ypa).

The one impediment to Shaw's growth as a passer was that he could never seem to shake was his tendency to hold onto the ball too long when receivers aren't getting open or the protection breaks down. Connor was sacked 23 times in 2011, which means 10.9% of his passing plays ended in sacks. I know I'd be OK with a 2-3 percent drop in his completion rate (via chucking the ball into the student section) if we saw a corresponding drop in his sack rate. There was a lot of talk in the offseason about Shaw studying Drew Brees' pocket presence to improve in this area, but so far his 10.7% sack rate is nearly identical to what it was last season.

An area where Connor has regressed in 2012 is his completion rate, currently sitting right at 60%, which is awfully close to his pre-Citadel game levels in 2011. (But it's worth pointing out that if D.L. Moore had caught the ball where Shaw hit him right on the numbers, his season completion rate would be 64%, thus illustrating the dangers of reading too much into early season stats.) He also hasn't hit any big plays in the passing game (6.96 ypa), with his longest completion against some pretty weak secondaries being the 25 yard pass Ace Sanders against on a short throw that was lengthened by some nice running after the catch.

Despite the shoulder injury, Shaw has actually been a more effective runner so far this season (6.5 pre-sack rushing yards per carry in 2012 versus 5.93 in 2011), which can likely be attributed to the added threat of handing it off to Marcus Lattimore when running the zone read.

The problem with Shaw's throws so far has been a combination of not having enough velocity on the ball and throwing the ball too late (specifically, waiting until a receiver is open rather than in anticipation of him getting open out of his break). This is a recipe for a passing game that is very difficult to watch.

This is what Spurrier had to say about Shaw's throws during Tuesday's press conference:

Connor is a better passer than he's shown. He really is. He sometimes holds the ball a little too long. He was a little deliberate last night - or two nights, three nights ago. Tentative was the word I used after the game. He's just a little hesitant. So he needs to get out of that. He needs to just take his steps and fire it in there a little better at times.

Here's a video of his pass attempts against UAB:

In going back and looking at the plays consecutively, I was struck by how well the passing offense seemed to be clicking at the beginning of the game, prior to the interception where Ace Sanders and Connor Shaw appeared to have a disagreement about what route the former should have run. Another thing that stands out, especially on the pass bogus pass interference call that extended our last drive before halftime, is that Cody Gibson is consistently getting beaten off the edge. I initially thought the post route to Ellington on the play with the PI call was badly underthrown, but you can see in the replay that Shaw was being hurried by the blitzing defensive end and had to make an off-balance throw.

While the offensive line is not blameless, there are clearly some bad throws there. So the question is, "How much of this regression is due to Connor not being 100% and how much is due to the fact that he might just not be as good as he looked during the last three games of 2011?" This is a question that's very difficult to answer, given our imperfect knowledge about the extent of Shaw's injury, how much his injury is affecting his throws, and that he has yet to be confronted with an above average defense in 2012. But for the purposes this discussion, let's just assume that what he's shown so far this year is what we'll continue to get for the foreseeable future. How does it compare to what we'd get from Dylan Thompson?

Dylan Thompson's skill profile has been much closer to the all-or-nothing variety (52% completion rate, 10.14 yards per attempt). He either completes a deep pass for a large gain or throws a bounce pass on a slant route or throws an interception that gets wiped out by a roughing penalty. He's also been credited with not holding onto the ball as long as Connor, but Dylan's identical 10.7% sack rate begs to differ. His pre-sack rushing YPC of 5.6 is almost a full yard per carry below what Connor is doing.

So, really, the primary statistical difference between the two is completion rate and yards per attempt. Between the two, Dylan's skill profile seems most unlikely to make a favorable transition into SEC play. While the sophomore from Boiling Springs has made some excellent throws, he has also made a lot of bad throws, and most of his yardage has come on plays where South Carolina's receivers have simply taken advantage of overmatched C-USA defensive backs.

Connor Shaw also has also shown that he has skills that simply cannot be measured. The ballsiness of his performance against Vanderbilt doesn't produce anything that shows up in a stat line, but it showed that he's as vital a cog in this offense as Marcus Lattimore and that he would sooner make an extremely questionable decision about his personal health and safety than watch this team lose. That's not to say that Dylan Thompson wouldn't do the same if put in that position, but Connor is one of the undisputed leaders of this Gamecock team and, if the difference between Shaw and Thompson is anywhere close to a toss-up, you have to err on the side of Connor.

I say all of this about Connor even though he blocked me on Twitter because I light-heartedly complained about him tweeting a Dark Knight Rises spolier immediately after he saw it on opening night. (I still haven't seen it yet, so shut up.)