Yesterday on TJB, Bassett posted a link to an interesting article by the Star Ledger’s Michael Fensom. Fensom wrote about adjustments that Marty Mornhinweg made to the offense over the last month of the season that contributed to the improved success for the offense as a whole and Geno Smith individually.
The part that caught my attention was this:
…one noticeable adjustment over the season’s concluding month was the Jets operated quite a lot out of the shotgun formation.
There’s no question that the Jets did operate out of the shotgun a lot over the last month of the season and that Smith showed some progress in such situations. However, did it really represent an adjustment that one could attribute Smith’s improvements to? We analyze in more detail after the jump.
How often did the Jets pass out of the shotgun?
I’m going to start by giving consideration not to the total number of shotgun snaps, but limiting my focus specifically to when Smith actually passed the ball. I’ll discuss running the ball from the shotgun later on.
The easiest way to evaluate this is to isolate the numbers for when Smith was under center but still threw the ball. All information is compiled from NFL data, so any inaccuracies are theirs, not mine and for the sake of consistency I’ve removed any plays that ended up being negated by a penalty from the data. If Fensom’s perception that the Jets helped Smith by getting him to work out of the shotgun more over the last month, you’d expect to see a reduction in the number of times he threw from under center.
However, this was not the case. Over the first four games, he threw 30 times from under center. This dropped to 15 over the next four games, then 14 for the next four. Over the last four games, this rose up to 21 again. So, actually Fensom is wrong and the Jets passed less out of the shotgun over that last month? Actually, that’s not correct either.
During that last four games, Smith threw 116 passes. However, in the previous four, he had only thrown 74 due to a combination of the Jets paring back their passing game and the fact that he was benched twice. Therefore, as a percentage, he was throwing from the shotgun about the same amount over the last four games as he had while struggling during the previous four.
Looking back to earlier in the season, this is actually where there was a clear increase in how often he worked out of the shotgun from the first four games to the next four, which didn’t see him make any on-field improvements. There’s one outlier in there, as Smith actually passed from under center 11 times in the first game, more than in any other game all year and well above his average (five times per game).
One final point to note about those last four games is that the number of times he threw from under center did show some signs of reduction. It was six in week 14 and eight in week 15, but only four in week 16 and three in week 17. That still wasn’t less than earlier in the year though – Smith had three or fewer passes from under center in six of the first 12 games. In addition, the fact that there was an above average number of attempts from under center in weeks 14 and 15 would appear to suggest that the assumption more shotgun was a factor in Smith’s improvements is an over-simplification at best.
How often do teams generally pass out of the shotgun?
These numbers are slightly meaningless without context, so I ran the numbers for this weekend’s playoff games to see how they compare with the teams that played this week. As you might expect, five pass attempts per game from under center seems to fall within the bounds of normality.
Alex Smith – 15
Kaepernick – 12
Brees – 11
Foles, Dalton – 4
Rodgers, Rivers – 3
Luck – 1
It’s perhaps not surprising that two run-heavy teams in San Francisco and Kansas City passed out of the shotgun more than the rest, but it’s more of a surprise to see Drew Brees up there. That may have been gameplan specific, though, because the Eagles are weak against the run. At the other end of the scale, you can see that most of the quarterbacks passed from under center about the same as Smith over those last couple of games. Clearly the Colts use the shotgun almost all the time.
Success when passing from the shotgun
When we talk about Smith operating more out of the shotgun and seeing more success as a result, we’re really talking about eliminating those under-center plays where he might not be as comfortable or might struggle to see the field. The effect of this is two-fold. Not only are you removing plays where he might be statistically efficient and therefore improving his numbers overall, you’re also perhaps helping him to get into a better rhythm. The downside could be that the offense becomes more one dimensional and predictable, but at the same time, if you pass from under center less, then those plays in isolation perhaps have more chance of being successful.
When passing from under center, Smith completed just 40-of-80 passes for 414 yards, two touchdowns and five interceptions. That’s a quarterback rating of just 47.6. On that basis, eradicating such plays from the playbook perhaps would have had a positive effect upon his numbers. From the shotgun, Smith had a quarterback rating of over 70 during the season, a modest improvement on his 66.5 final rating. If he was operating solely out of the shotgun, would we see improvements of this order (perhaps further augmented by the rhythm factor mentioned above) or would the increased predictability of the offense outweigh this?
Over that last month, Smith’s numbers while passing from under center were not so bad. He had a quarterback rating of 60.8 having had a quarterback rating of just 9.8 in the previous four games when passing from under center. We’re getting into small sample sizes here though. For what it’s worth, from games five to eight – where we’ve identified that there was an unsuccessful shift towards more shotgun – his rating while passing from under center was just under 70, lending weight to the “less is more” theory noted earlier.
It gets even more interesting when you consider the outlier mentioned earlier where he passed a season-high 11 times from under center in week one. In that game, he actually went 9-of-11 for 77 yards and a touchdown. He did also have an interception, but that was easily the most success he had passing from under center all year and rather than build on this, the Jets opted to go away from having him pass from under center as the year went on. The result, of course, was that he saw less success when he did that. Removing that first game from the numbers when passing from under center drops his rating down to just 40.5. So maybe less isn’t more.
Clearly passing from under center was something they placed an emphasis on during the offseason and also in that first game they were perhaps trying to capitalize on the fact that maybe the Bucs wouldn’t be expecting him to pass in those situations. If this was something that he needed to work at, obviously he would need to reinforce that work over the course of the season and with the Jets shifting towards more shotgun over the rest of the first half of the year, perhaps this had an adverse affect on Smith’s development by not enabling him to continue to work at that aspect as much as he perhaps needed to.
In order to place these numbers in the proper context, we can look at the difference between quarterback ratings and the ratings when in the shotgun for other players around the league. For many of them, these two numbers will be close. That’s not just a sign of their overall consistency, but also a sign that many quarterbacks operate primarily from the shotgun and therefore their overall rating shouldn’t differ too significantly from their rating when in the shotgun. Luck, for example, has an 87.0 overall rating and an 86.9 rating when in the shotgun.
First, let’s consider those quarterbacks whose overall rating is markedly better than their shotgun rating. The overall ratings for Russell Wilson, Peyton Manning and Cam Newton are all about five points higher overall than when in the shotgun. Maybe this is because those three are more effective with play-action. Let’s come back to that thought later.
Now, let’s consider which quarterbacks were better in the shotgun and saw their overall rating reduced by their passing from under center. The biggest discrepancies belong to the likes of Robert Griffin, Kellen Clemens and Colin Kaepernick. Clemens’ rating was almost ten points higher than his overall average when passing from the shotgun, but maybe this is just a product of the fact that the Rams were trailing a lot and therefore went into the hurry-up against some soft zone and prevent defenses. Kaepernick, of course, comes from an offense where a bigger discrepancy is less surprising because they do have more plays where they pass from run formations. He continued to struggle more when under center on Sunday though, completing just four of 12 passes for 47 yards and a pick (Alex Smith was the most effective when under center this weekend – 11-of-15 for 180 yards). As for Griffin, a point worth noting is that NFL data doesn’t differentiate between shotgun and pistol, so all those plays would be included within the shotgun data.
Back to Smith. We can see that there are some other examples of players that seem to do better when throwing from the shotgun and obviously their coaches will be weighing up how best to overcome this (including simply throwing from the shotgun more), just like the Jets will.
One might expect there to be a correlation between the numbers when throwing from under center and in play-action. If you’re lined up in a run formation, but end up throwing the ball, that’s something we would naturally associate with play-action passing. However, once again this may be an over-simplification.
Many teams run play-action from the shotgun just as much as they do from under center. In fact, there is less of a need to make a play-action fake when you’re in a run formation, because you’re likely to have already affected the alignment of the defense before the ball is snapped.
I’ve often thought that those quarterbacks who are inexperienced at operating from under center would struggle with play-action in particular because you usually have to turn away from looking downfield and then look up and make a read. From the shotgun, at least the quarterback can keep his head up and read the field more easily before and after the fake. I’ve theorized in the past that this is why Brian Schottenheimer seemed to favor shotgun sets in short yardage situations while Mark Sanchez was his quarterback (although this doesn’t explain his propensity for empty backfields).
Let’s go back to my previous comment that perhaps the improvements in overall rating as against rating in the shotgun for Wilson, Manning and Newton could be attributed to play-action. If that theory were correct, then each of them would have a higher rating when making a play-action fake than when not making a fake. Sure enough, based on data from PFF they all have significantly better numbers on play-action passes than non-play-action. Wilson’s rating is 17 points better, Manning’s is almost 30 points better and Newton’s is even more than that.
It gets even more interesting when we look at 2012’s numbers, though.
Starting with Newton, he had a 109.4 rating on play-action passes, much better than his rating for non-play-action (75.4). So far, so good. Manning also shows some improvement, albeit not quite so pronounced (108.8 as against 104.8). Wilson, however, actually had a better rating on non-play-action passes (102.6 as against 95.0). Clearly, therefore, it’s not always that simple and maybe this is a sign of where Wilson has grown this year and the next logical step for Smith if he wants to get better.
How about Smith this year, though? His quarterback rating was 68.2 on non-play-action passes but just below 60 on play-action throws. It would seem that this is somewhere he needs to improve, but the discrepancy is not as big as that between shotgun and non-shotgun. The sample size is slightly more – he threw 98 play-action passes and made a play-fake 22.4% of the time, the 12th most in the league.
Aside from Smith, the only other quarterbacks with a lower rating on non-play-action passes were Cutler, Roethlisberger and Griffin, all three of whom also had a better rating from the shotgun than when under center, so there does seem to be some correlation.
Running from the shotgun
Earlier on, I admitted that I was restricting my analysis to actual passes from the shotgun (or, more accurately, not from the shotgun). However, Fensom’s article (and Bassett’s focus) talked about the Jets operating more out of the shotgun, not just Smith passing from the shotgun. Could it be that they were in the shotgun more and although the percentage of throws that were shotgun throws didn’t really change, this was reflected in more runs from the shotgun formation? Then, taking that one step further, is there any reason that this would help Smith? Maybe it could in terms of helping him to stay in rhythm and read the defense better over a series of plays, but there’s nothing tangible or obvious.
Looking at the final game of the season, the Jets ran the ball a surprising 33 times from the shotgun in that game. They only had 42 carries in total, although the 33 runs noted above does incorporate penalties this time. That would seem to represent a significant increase from earlier in the season. If we go back to a run-heavy game (36 carries) from earlier in the year – the Saints game – then they only ran 18 times from the shotgun in that game. There was one significant factor in that last game though and that’s Chris Ivory’s injury. Maybe the Jets simply have more shotgun packages for Bilal Powell, who ended up being the lead back in that game, and more under center packages for Ivory. Sure enough, if we look back to the previous week against the Browns, when Ivory was available, the Jets had 39 carries, but only ran 20 times from the shotgun. Again this would suggest that if they were running more from the shotgun over that last month, it’s perhaps not as pronounced as you might think.
Another factor at the end of the season was that Smith ran the ball more, which was also acknowledged by Fensom in his article. While I’ve included these in the run-from-the-shotgun data in the above paragraph, some of these would have been designed as pass plays with Smith making the decision to take off. That being the case, perhaps we can account for a slight uptick in the amount of shotgun pass plays dialed up over the last month and should also factor some of Smith’s running statistics into any evaluation of the improvements he made as a passer.
Does it matter?
If Smith is struggling to be as effective when operating from under center, couldn’t the Jets just give consideration to getting him to operate almost exclusively from the shotgun?
While this works well for Luck and others, there are limitations. Marty Mornhinweg runs a complex offense which has multiple formations, personnel groupings and different variations of plays. A play like the sprint draw, a Mornhinweg running-game staple, simply wouldn’t be as effective if you moved the quarterback into the backfield. That would affect all variations of that play, including tendency busting passes. You can’t therefore just make the adjustment of moving the quarterback into the backfield unless you stop running certain plays, which removes some of the benefits of having a complex offense in the first place. Only passing from the shotgun and running the ball every time when the quarterback is under center isn’t a long-term option either (although it’s evident they did this at times). That would make the play selection too obvious for the opposing defense, even if you broke tendency from time to time.
Mornhinweg has shown a willingness to tweak his offenses based on personnel in the past – running the ball more often at times this year and employing fullbacks to varying degrees over the years, dependent on his team’s personnel strengths and weaknesses. Perhaps he would give consideration to running a consecutive series of plays from the shotgun in order to get the benefit of Smith finding a rhythm, without compromising the effectiveness of the running game or increasing how predictable the offense is.
To what else can we attribute any improvements?
So, having reached the conclusion that any adjustment to operate more out of the shotgun was perhaps not as significant as first thought, but at the same time acknowledged that this is still likely a factor in his improvements over the last month, what were the other factors?
Down the stretch, I identified improvements in Smith’s decision making and decisiveness, but also his pocket presence, confidence and consistency. While all of those are things that could be improved through him having a better comfort level within the offense, I’d also attribute them to the personnel around him. There was a marked difference in how much he trusted his receivers, which manifested himself in some well timed throws, whereas the same timing patterns would almost always be thrown late earlier in the season, with Smith unwilling to release the pass until after his receiver had made his break. His pocket presence was likely helped out by the Jets tweaking their protection schemes over the last month and Brian Winters finally starting to find his feet.
You also have to give some credit to David Lee. While Mornhinweg, the offensive line settling down and the return of the injured players may have made Smith’s job easier for him, he still had to do that job well. While his technique was still inconsistent at times, he did show progress throughout the year and it will be interesting to see if he develops faster with a full offseason to refine that.
Sometimes the media will oversimplify things based on gut feel or, just as bad, by relying on statistics that they fail to place within the correct context. Some fans do this too, often when the article in question does give consideration to all the factors. I also don’t doubt that there are coaches and maybe even players around the league that are guilty of doing the same thing.
Adjusting to more shotgun sets alone didn’t magically turn Geno Smith into the solid quarterback we saw over the last month of the season. If it were that simple, we’d have seen improvements from him when they did that (to a more pronounced degree) after the first four weeks of the season. We’d also know for certain that he could remain at that level and develop from there if they continued with that approach in 2014. Of course, there could well still be ups and downs in store for Smith in the years ahead. He might not even end up as the starter next year.
Whatever the reasons, there were improvements over the last month and Jets fans will be hoping Smith can pick up where he left off.