I have watched three rookie QBs thus far -- Darnold, and before him, Allen and Mayfield -- in my tour through the controversial 2018 QB class. Definitely with Allen, probably with Darnold, and maybe with Mayfield, I noticed the same thing across their respective games:
They're all better outside of structure than they are within structure.
Now, we have to define "within structure" pretty stringently for this to work, when it comes to Darnold. But while Darnold's short game was how he made his money in college and again, where he was successful with the Jets, Darnold does not win in the short game like most quarterbacks.
Typically, strong short game QBs have an inherent sense of timing and rhythm. Their drops are measured, balanced, and consistent. They jive with the timing of the route concept, so that they can land on their back foot and immediately drive into a throw, or softly hitch into their next progression. Watch Drew Brees, Tom Brady, or Matt Ryan -- the bottom half is active, but balanced and under control; the top half is quiet, cocked, and ready.
It is not so with Darnold, who sometimes flashes a nice three-step-and-hitch sequence, and other times, just completely surrenders his base and sense of pace. This is why we might say Darnold is out his best "outside of structure" -- he regularly makes quality, off-platform throws, whether because pressure put him off-platform, or because he did himself.
It may even be a stretch to call these reps "off-platform." Darnold is a confusing and unique quarterback, and his game is accordingly difficult to fit within the typical paradigms of QB play. That's why he got Aaron Rodgers comps during the process -- I'd even say he does some Mahomes-y things, in terms of generating velocity without a throwing base. He does the irregular things that these QBs do, and that makes him exceptional if he's successful.
I don't think he's there yet.
So while these plays may be "outside of structure" or "off-platform" in a traditional sense, given Darnold's play style and mechanical preferences, they're within his structure.
Take that first rep, the quick pivot off of play-action. Most QBs shouldn't be asked to attempt this style of pass, because of how quick your release needs to be, and how hard it is to set your feet to the target. But because Darnold often doesn't need his feet set into his target to be successful, and because of how lightning fast that over-the-top release is (Darnold's upper-body mechanics are great), he has no problem here.
Second rep: hop steps into his drop, is late with his footwork, has to plant and drive with basically no lower-body power. Still has the heat to jam that ball in there, but you can see the inaccuracy flash. Much the same could be said on the third rep, wherein Darnold resets his base backwards before making a throw that goes...you know, forward.
My challenge to you is this. On some of these reps, pause the clips in the end zone view, as Darnold is releasing the ball, but before it begins flying to the target. Guess to which area of the field he's throwing it -- be precise. Play the clip and see if you were right.
I'm often wrong, when I play this little game on Darnold's film. Plainly speaking, he does not set his throwing hallway the way you are traditionally taught to do. The biggest time this becomes an issue is on routes to his right.
If you think about it, as a right-handed quarterback, you tend to set your throwing base a little to the left naturally. You're pulling that front shoulder back to open your stance a little bit and see the field more easily. Darnold is guilty of this -- his drops are straight bad -- and accordingly his hallways are often set improperly to the right side of the field, because he tails off left to begin with.
This is stepping into the bucket in broadcaster parlance, and it is a problem, no matter how weirdly successful he is with his poor mechanics elsewhere. Darnold is not a consistent thrower at this point in his career, and even his flashy games (GB Week 16 was awesome) will be too few and far between to manufacture a playoff offense.
Can you fix lower-body problems like this? Or at least increase accuracy somehow? Yes, but that was the concern I had with Darnold coming out, who I ranked below Mayfield, Lamar Jackson, and Josh Rosen. Despite all of his velocity and his quick release, Darnold regularly ranked as an average to below-average thrower on the 2018 Contextualized Quarterbacking's accuracy and placement scores. Fixing mechanics isn't easy, it doesn't happen overnight, and sometimes, it doesn't work.
What you have to be excited about is the deep ball, for Darnold and the Jets. The dude has an absolute whip, and given how little lower-body power he recruits, some of the downfield throws he makes are even stupider. This is where Darnold was successful in college, with over 20% of his throws going more than 20 yards down the field, and his best accuracy/placement scores relative to the class.
As I said, Darnold has a hose, and that velocity allows him to attack deep and intermediate windows -- something he didn't do often at USC -- even when his footwork puts him behind, or when he's operating outside of the pocket.
There are some great examples in here of good footwork and bad footwork alike, just to show you what Darnold is capable of with just his arm.
As I said, Darnold is a peculiar quarterback. I ripped through almost 200 passing attempts and I couldn't tell you what exactly his fixable problems are and how I'd go about it. But he is still quite young, and was raw coming out of USC. The Jets weren't gonna be winning in 2018, and they probably won't win in 2019, either.
Nothing wrong with needing development, but Darnold must show a more normalized process mentally and physically in Year 2 to give me hope that he will stick as a playoff-level starter in the league. Those players can churn out known quantities, expected and predictable levels of performance, week in and week out. Darnold isn't able to do that yet -- most rookies can't -- and will only be able to if he starts introducing more structure into his process.