Shotgun Wedding: Alex Smith Meets the Offense of His Dreams

AUTHOR'S NOTE: The time I usually devote to team and players stats was instead devoted last week to this very special article I've put together with the help of Aaron Schatz from Football Outsiders (FO). He provided the individual stats for Alex Smith and Shaun Hill, and I did the writing, but we're both in general agreement about what the stats say as they relate to Smith and the shotgun. If you want to show your appreciation for Aaron's entirely unobligated attention to the whims of our blog, and have an interest in accessing FO's statistical database for yourself, head over to his site and get the premium membership. Obviously, I highly recommend it.

Besides whether or not Mike Nolan was going to make it through an entire press conference without saying, "look it," the fate of Alex Smith has been the hottest topic of Niner debate over the past 5 years. I don't need to recap all the gory details, but suffice it to say that he's a polarizing figure among Niner fans. My goal for this post is not to settle the debate. Rather, what I'd like to do here is answer a Dennis-Green-esque question about Smith: Is he who we thought he was?

Whether you love him or hate him, and whether you think he's the 49ers' "QB of the future" or not, there's no denying that what we've seen from Smith this season constitutes a remarkable turnaround for a guy who, just 9 months ago, was in serious danger of being released. In this post, my intention is to show - as crisply and clearly as possible - what the statistics say about the reason(s) for his resurrection. I know the guy's already married and all, but, as you'll see, this season has been something of a shotgun wedding for Smith.

A ROCKY RELATIONSHIP

There's really no mistaking the fact that Smith's career has been seriously hampered by injury. Indeed, he's only had 2 seasons in which he was actually healthy: 2006 and 2009. So, the first clue about the reason for his renaissance is that he's actually healthy. To drive the point home, here's a table showing the 49ers' Pass OFF DVOA for games since 2005 that Smith did not play, games he played while coming off of an injury, and games he played while totally healthy (I've excluded 5 games in which multiple QBs played):

As you can see from this table, the 49ers' pass OFF has been considerably better in games when Smith is healthy when compared to games in which he didn't play or played coming off of an injury. What's most interesting to me is that, despite the ridiculously small sample size for "Smith coming off injury," the 58.1% difference is statistically significant within 95% certainty. In other words, it's beyond a reasonable statistical doubt that the Niners' OFF plays better when Smith is healthy.

After the jump, I'll show why health is just one of many things that are the same between 2006 and 2009, and therefore can't explain why Smith is better in 2009 than 2006. And, of course, I'll offer my suggestion for what CAN explain it (See story title)...

Of course, not all "Smith healthy" games are created equally. For instance, 16 of these 25 games occurred in 2006, 3 occurred in 2007, and 6 have occurred in 2009. Each of these seasons was qualitatively different for various reasons:

1. Different offensive coordinators (OCs)
2. Different offensive lines (OLs)
3. Different receiving corps

So, to further isolate the reason(s) for Smith's renaissance, we need to see if any of these qualitative differences have exhibited any quantitative (i.e., statistical) difference to support the theory that they are a meaningful contributor to Smith's turnaround this season. Again, remember that we're comparing "healthy Alex" to "healthy Alex" here so that we partial out health as a factor.

First, the OCs. It's conventional wisdom around NinerLand that Norv Turner turned Smith into a competent NFL QB during the 2006 season. However, as we've just seen, the fact that Smith was healthy in 2006 is as important - or more important - a reason for why he played so well that season. Interestingly enough, the fact of the matter is that Smith - and the Niners' overall Pass OFF in games he started* - is considerably better this season than he was in 2006 (2009 stats are through Week 13, Pass OFF Game DVOAs only include games Smith started):

So, if Smith is better now, how much of it has to do with the fact that Jimmy Raye is his OC rather than Turner. Here's a table comparing the Raye's and Turner's career DVOA stats and rankings prior to being hired by the 49ers:

While I'd say that Raye's track record was slightly better when he got hired, I wouldn't say that it necessarily accounts for Alex Smith's improvement above and beyond what we already know about the impact of health. For instance, a 3.5% improvement in Pass OFF DVOA -which is what we'd expect given Raye's and Turner's track records - is hardly a drop in the bucket of Smith's 14.9% individual DVOA improvement. Also, keep in mind that Turner's stats don't include his pre-DVOA-era stint as OC of the early-90s Dallas Cowboys. So, all in all, I'd say the difference in OCs doesn't explain Smith's renaissance.

Onto the OLs. How much of Smith's improvement this season can be attributed to the inherent ability of his OL to block pass rushers? Here's the comparison:

OK, so actually, the Niners' OL is worse this season when it comes to pass-blocking. Of course, FO's research shows that ASR is more dependent on the QB than people realize, so it may just be the case that the OL's worse ASR this season is a byproduct of Shaun Hill starting at QB for about half the games. As I'll show later on, ASR has, in fact, decreased since Smith took over at QB. However, what we're trying to explain here is whether Smith's renaissance is due to having a better OL in 2009. Suffice it to say that, even acknowledging the OL's ASR improvement with Smith at QB this season, their ASR is still not better than it was in 2006. In other words, even if you account for the fact that a sack-prone Hill was the QB earlier in the year, the OL is still not pass-blocking better this season than in 2006, and so "better OL pass-blocking" can't be the reason for Smith being better this season than he was in 2006.

The third component of potential qualitative differences between Smith's 2006 and 2009 seasons is his supporting cast in the passing game. Here's a table showing the stats for his stable of receivers in 2006 and 2009:

From the averages at the bottom of this table - which are weighted by number of targets - it's pretty clear that Smith's 2009 receiving corps is not much better - if at all - than his 2006 receiving corps. It's true that Vernon Davis is better now than when he split time with Eric Johnson back in 2006, but Frank Gore's receiving stats are worse this season, and both of Smith's starting WRs (Michael Crabtree and Josh Morgan) are worse statistically than their counterparts in 2006 (Antonio Bryant and Arnaz Battle). Therefore, it ends up being a wash.

So, just to recap this section, we've determined a few things. First, Alex Smith is significantly better when he's healthy than when he's not (aka duh!). Second, when comparing "healthy Alex" to "healthy Alex," he's actually better this season than he was in 2006. Third, this improvement is not due to (a) having a better OC, (b) having a better OL, or (c) having a better receiving corps.

I NOW PRONOUNCE YOU QB AND OFFENSE

Based on the above, Alex is having his best season this year, but it's not because of health (i.e., he was worse while healthy in 2006), it's not because of a better OC, it's not because of an improved OL, and it's not because of an improved receiving corps. What the hell is it then?

Well, perhaps there's something about this season in particular that doesn't have to do with any of the factors that I've already ruled out. Oh, I know, it's called "the elephant in the room," aka shotgun OFF. Because the only thing different - in terms of offensive football - between this season and his other healthy season, 2006, is the implementation of a shotgun OFF, the only stat-related** factor left to look at is whether or not the reason for Smith's renaissance is due to him being a better shotgun QB.

Thanks to Aaron Schatz over at Football Outsiders***, I got my hands on Smith's individual shotgun and non-shotgun stats through 11 games:

Eureka! When Smith is in the shotgun, he averages half-a-yard more per pass, his interception rate (IR) drops by half, and his QB DVOA is 38.0% better! So, yeah, Smith is waaaaay better in the shotgun.

Of course, it could just be that all QBs show this kind of discrepancy between shotgun and non-shotgun performance (after all, FO's research shows as much), and so what I'm attributing to shotgun might actually just be a statistical artifact of that formation being more efficient in general. To solve this, we need to do a little experiment.

If only there was a QB to compare Smith to that's had the same OC, the same OL, and the same supporting cast this season. Oh wait, there is. His name is Shaun Hill. The question at hand, then, is, "Was Shaun Hill waaaaay better in the shotgun this season just like Alex Smith has been?" Again, thanks to Aaron, here's Hill's stats:

As is plainly obvious, Hill did not enjoy anywhere near the amount of improvement in shotgun that Smith has. Hill's Shotgun GB DVOA was over 30% worse than Smith's is, and his "benefit from the shotgun" difference is over 3 times smaller. Furthermore, as you can see from the differences between DVOA and unadjusted VOA for each QB, Hill's horrible shotgun efficiency actually came against an easier shotgun schedule. In other words, Smith's massive benefit from shotgun as compared to Hill was totally independent of any difference in strength of schedule.****

So, essentially, if we put another QB in practically the exact same situation as Alex Smith in 2009, that QB doesn't benefit from shotgun nearly as much. In other words, all else being equal, Smith has been helped by the shotgun offense waaaaaay more than Hill, who had the same OC, same OL, and same stable of receivers (p.s. In case you're wondering, Bruce stats = Crabtree stats, and Morgan with Smith < Morgan with Hill, so, if anything, Hill had a better supporting cast according to DVOA).

MARRIED WITH CHILDREN

Just to put an exclamation point on things, let me provide some evidence showing that the increase in shotgun with Smith at QB has borne fruit outside of Smith's own renaissance. Here's a table showing some 49er stats as they were after Hill's last full game, and as they are after 12 games:

As you can see, there's been absolute improvement across the board: better overall OFF, better pass OFF, better rush OFF, better run-blocking, and better pass-blocking. In other words, an increase in the use of shotgun underlies both Smith's specific improvement and a general improvement in the 49ers' OFF overall.

BOTTOM LINE

If you put everything I've presented in this post together, you see a clear picture of why Alex Smith appears to be a better QB all of a sudden. The stats-based logic goes something like this:

1. Smith plays better when healthy, and the 49ers' OFF plays better with a healthy Smith.
2. Smith has only had 2 healthy seasons with the Niners: 2006 and 2009.
3. Smith is playing better while healthy in 2009 than he did while healthy in 2006.
4. Smith does not have a better OC, a better OL, or a better receiving corps than he did in 2006.
5. Therefore, Smith's improvement in 2009 cannot be attributed to health, OC, OL, or receivers.
6. Smith plays waaaaay better in the shotgun.
7. Hill, who had the same OC, OL, and receivers as Smith, and even played an easier shotgun schedule than Smith, did not play waaaaay better in the shotgun.
8. Smith has taken about 60% of his snaps from shotgun this season.
9. Therefore, the main reason for Smith's renaissance is more shotgun.

In other words, it's been a shotgun wedding for Smith and the 49ers' OFF.

*If you add in the 83.8% Pass OFF Game DVOA for this season's Texans game, the 2009 average increases to 13.0%.

**I realize that there are non-statistical factors that could explain Smith's improvement. He could just be a "smarter QB" now after having basically done nothing but "mental reps" for the past 2 years. He could be "more mature" given the adversity he's had to fight through so far in his career and personal life. However, these are intangibles that no one can - or should - try to measure from my distance. So, I'm limited to tangibles that we can measure statistically. I will say, however, that, if it's truly the case that Smith is "smarter" or "more mature" in 2009, then why do these things not show up when he's under center? Can it really be that moving a few steps forward and putting his hands under his C's rear end makes his brain go soft and turns him into an NFL toddler? I seriously doubt that.  

***DVOA, DYAR, EYds, ALY, & ASR statistics used to produce this article were obtained from Football Outsiders.

****And one more thing, don't think that Hill has been worse than Smith from shotgun simply by virtue of taking a lower percentage of snaps in that generally-more-efficient formation. Remember, I've presented their "per play" stats, i.e., I've already accounted for the difference in shotgun percentage, and, therefore, eliminated it as a confounding variable.