Jake Locker was sacked hardly at all and very frequently in 2012, which means what for 2013?

Jake Locker's Adjusted Sack Rate as a rookie was 8.3%. Jake Locker's Adjusted Sack Rate in his second season was 8.2%. The league-averaged Adjusted Sack Rate, Football Outsiders' statistic that accounts for the quality of the opposing pass rush and situations (e.g., quarterbacks are sacked more often on third down and long), in 2012 was 6.5%. Therefore, Jake Locker is one of those quarterbacks who tends to get sacked a lot, so we should expect Jake Locker's Adjusted Sack Rate in 2013 to be about 8.3%, right? 

Well, it's complicated.

One of the things I did here last offseason was write about how the Titans, which meant mostly Matt Hasselbeck, were sacked in 2011. Sacks are rare events; per Football Outsiders charting data, less than one quarter of total pressures in 2012 (hurries, hits, and sacks) ended up with the quarterback on the ground with the ball and not gaining any yardage. Still, they happen and they're important, so they're worth taking a look at in greater detail.

Sacks seem to be mostly a function of the offense, which in fact means a function of the quarterback and perhaps the offensive system. When a team changes quarterbacks, the team's sack rate changes. I've noted this before, but Jason Lisk found there's no correlation between quarterback A and quarterback B's sack rate even when A and B play for the same team in the same season. Moreover, when a quarterback changes teams, sack rate tends to remain as much or more constant than other rate stats. The Denver Broncos provide excellent recent examples of both of these phenomena. First, Tim Tebow replaced Kyle Orton during the 2011 season. Orton was sacked 5.4% of the time. Tebow was sacked 11.2% of the time (both numbers by ASR). Second, Peyton Manning became the Broncos' starter in 2012. His sack rate was 4.2%, in line with his ASR in Indianapolis, which fluctuated between 2.8% and 4.2% from 2002-10.

Jake Locker's Adjusted Sack Rate in the Titans' first 11 games was 3.7%.

Jake Locker's Adjusted Sack Rate in the Titans' final 5 games was 12.9%.

In other words, Jake Locker spent half the season getting sacked less often than Peyton Manning and the other half of the season getting sacked more often than Tim Tebow.

Quarterbacks are the biggest determinant of sack rate. Same quarterback, same season, wildly different sack rates. This makes no sense. At all. What the heck is going on here, and what should we expect from Jake Locker in 2013?

At this point I now try to explain something I regard as essentially confounding and completely inexplicable. So, here goes nothing.

Possibility 1: The Titans are changing basically the entire passing offense to something new (at least to Jake Locker) in 2013, so 2012 (and for that matter, 2011) data is entirely useless in making predictions for 2013.
Evaluation: There's something to this idea. I think offensive scheme can have a lot to do with sack rate. Buffalo is a great example of that-Ryan Fitzpatrick was hardly ever sacked not because the Bills had the league's most accomplished offensive line but because the design of the offense was for him to get the ball out quickly. The Titans threw out Chris Palmer's playbook (which they were using at least in some form even the second half of 2012), and we've never seen Jake Locker in another NFL offense. If you want to say "we can't judge Jake Locker at all, so talk to me again at midseason or next offseason," I get where you're coming from. I don't agree with you, since I think there's some information to be gained in over 400 NFL dropbacks, but I do get you.

Possibility 2: The Titans' offensive line the second half of the season was just that bad.
Evaluation: Yes, once Steve Hutchinson and David Stewart went down in Week 13, the line was really bad. Mitch Petrus, who started Week 16 against the Packers, may have been the worst player to start a game in the NFL in 2012. (I don't want to single Petrus out too much-he's probably still at least a trillion times better than me at football and got thrown into the lineup quickly, which was not his fault and "unfair" to him.) But there were still three offensive linemen in the lineup who started in the stretch where Locker was sacked rarely and one of the backups was a tackle who'd been in the system for at least two years. That wasn't a good offensive line, but it also wasn't the worst line we've seen in the NFL since the 2002 Texans. It affected Locker, for sure, but I don't buy it explains all the change.

Possibility 3: Fewer dumpoffs means more sacks.
Evaluation: We're in the realm of small sample size here, but FO charting has Jake Locker with 19 dumpoffs on the season. Of those dumpoffs, 14 came in the first eleven games and only 5 in the final five games. Locker had roughly equal numbers of dropbacks in each "half" I'm looking at, so 9 fewer dumpoffs could mean 9 more sacks. One thing both Mike Munchak and Dowell Loggains have talked about this offseason is better distribution of the wide receivers and running backs, so this could be implicated here. The only problem was, this doesn't seem to be an issue until the second half of the season.

Possibility 4: The loss of Jared Cook.
Evaluation: Cook missed the last three weeks, which was when things were at their worst. So, maybe? The problem I have with this attempt at an explanation is there's no causal mechanism. There was no Jared Cook, so so what? If this is what changed things, then it's likely because Cook's absence caused something else to changed. If so, what else was it that changed? If there is something meaningful here, I would expect it to show up in something else.

Possibility 5: Too many blockers means not enough targets.
Evaluation: I've noted before Chris Palmer philosophically believed in having lots of eligible receivers out in routes. The blockers were the 5 linemen, maybe plus the back. Dowell Loggains did use more blockers, but not a whole lot more and we still have the Titans with 5 blockers 40% of the time. I don't see a change from 5.56 to 5.69 blockers as a particularly compelling case for this. Then again, the league as a whole averaged 5.64 blockers per play so that change does mean the Titans went from a protector-light team to a protector-heavy one. Plus even with just 5 blockers Locker didn't stay as clean as he was the first half of the season.

Possibility 6: Opponents brought more rushers and had more success because of it.
Evaluation: This is at least half-right; teams did in fact bring heavier rushes against Locker in the final five games. Opponents brought five or more rushers 43% of the time after only doing so 29% of the time the first 11 games, and they had some success when they did so. The problem is, even on those less common occasions they brought four rushers they still had a lot more success than they did earlier in the season. This is a seductive explanation, but I don't think what happened actually supports it as much as it seems at first glance.

Possibility 7: Jake Locker's sack rate is particularly subject to the quality of the offensive line he's playing behind.
Evaluation: Unless you're dealing with a large and robust sample size that cannot be explained in any way, my position is to be deeply, deeply skeptical of any sort of snowflake theory claiming "X is different from all the things similar to X that have come before it." It's possible this is right, but I'm putting my money into Powerball before I put it here if only because of the payoff.

Possibility 8: Arbitrary endpoints. You're looking at two relatively modest sample sizes, each of less than 200 passing attempts, and suggesting both of them are independently meaningful. Hogwash. Stop overreading this "evidence" and just use his full-season number as the best predictor for 2013.
Evaluation: Even if only for space-related reasons, this more or less was the approach I took in Football Outsiders Almanac 2013, where I don't have 1300 words (and counting) to write just about what to make or not make of the split in Jake Locker's Adjusted Sack Rate.

I'm not sure what Locker's Adjusted Sack Rate is going to be in 2013. Subjectively, I haven't been impressed with his pocket presence. I think he didn't take many sacks the first half of the year because he threw a bunch of short passes and is athletic.

The Titans' offensive line should be better than it was the first half of 2012, though I think you've gotten the right idea if you read me as saying the offensive line isn't as important as you think it is.

We've seen quarterbacks change who they are before, to some degree-one of the reasons for That Brief and Magical Period Where Vince Young Seemed Like a Real Starting Quarterback was he didn't take nearly as many sacks as Kerry Collins did or he used to (and would again in the future). I think the Titans think Locker took more sacks than they'd like because he wasn't comfortable in Chris Palmer's offense, and he'll be more comfortable and play faster in the new offense (assuming the Titans even agree with me in thinking Locker took more sacks than they'd like, all things considered). We'll see.

Proverbial gun to my head, I'm putting my guess at Jake Locker's Adjusted Sack Rate in 2013 at 7.3%. Now tell me why I'm wrong and what I'm missing.

UPDATE (2013-06-024, 0846 CT): Prodded by Paul Kuharsky, I added a brief description of what Adjusted Sack Rate is in the first paragraph.

Quantcast