In my re-review of Derek Carr, I was fairly positive I would change my mind on him. A lot of the things I thought about him were definitely wrong, but for the most part I’m OK with him as a Vikings fit. I think he requires a lot more caution than Bridgewater or Bortles in terms of evaluation, and as you’ll see, there are a lot of pitfalls.

Carr is best known for being an extremely productive, statistics-friendly passer with over 5000 yards this season, and a 50:8 TD/INT ratio. He has one of the strongest arms in the draft and a brother who was a previous flame-out in the NFL. The statistics won’t matter, but he’s certainly an intriguing prospect.

Again, we’ll use the lens of Accuracy, Delivery, Ball Velocity, Field Sense, Pocket Presence, Scrambling/Running and Intangibles.

Accuracy: In terms of getting the ball to the receiver, Carr is a very accurate quarterback. I’ll be talking about his system a lot, and it makes sense to add the normal caveat that most statistics will be skewed by the high degree of screen passes. Knowing that, it’s important to find ways to adjust for his scheme.

The high amount of screen passes (28.6% of his passes were behind the LOS) was sort of a neurotic decision by Dave Schramm to reduce sacks and turnovers, something he highlighted before the season as the focus. The year before, only 19% of his passes were behind the LOS (same as Manziel this year, and less than Bortles), which is relatively typical for a college passer. (It helped—Fresno State finished first in the country in sack rate and eighth in interception rate. The year before, they finished 39th and 11th).

People forget that even though Carr threw nearly 30% of his passes before the LOS, he still threw more passes at every distance than any other quarterback considered the top of the draftable class. Isolating those plays gives people a good idea of his anticipation, accuracy and timing.

We can go back to the passing charts to get a general sense of his ability at different distances. While Bortles and Bridgewater have better numbers at the “sweet spot” of 5-14 yards, Carr’s numbers are decidedly average and Manziel’s are subpar.

Here is a case where I think that the numbers tracked and the general accuracy do not quite match up. There are two specific reasons that he and Bortles differ in terms of accuracy (and what it means for an evaluator). The first is related to ball placement and the second is schematic.

As much as I think Carr’s issues with pressure are overblown, not enough is said about his ball placement. I understand that, because it’s not extremely damning most time, but it is an issue enough times that I would call it an “area of concern” even if it’s not exactly a weakness. It’s still, however, superior to Bortles’ on a general basis.

Carr overcorrects for ball safety, and it creates far too many issues in terms of enabling his receivers to gain extra yards, but it does mean they are almost always the only ones with access to the ball. His cautiousness from this perspective is distinctly at odds with his aggressiveness in other areas, and it will be interesting for any coach to reconcile.

The ball will be given to the receiver in a catchable position where only they can get it, but they often have to stretch far more than they should in order to get to it.

In looking at 2012 tape to see if the new emphasis by Schramm caused the majority of this problem, I was left with some signs for positivity as well as some signs for negativity. His ball placement was much friendlier to receivers that year, but that doesn’t mean it was necessarily good on a consistent basis.

The second concern—schematic—mostly has to do with the type of intermediate and deep throws you’d see in his offense instead of the distribution of throws by distance. He’s better than his peers on a number of routes in terms of his accuracy: slants, fades, backfield routes and several more. On intermediate crossers he’s very iffy (a place where Bortles has him beat, hands down) and his plays in seam routes are incredibly inconsistent. Sometimes they are extremely precise, while at others wildly off.

Finally, it is important to note that Carr throws with anticipation but doesn’t have the same ability that David Fales, Keith Wenning and especially Teddy Bridgewater show on a consistent basis, or Bortles flashes when his mechanics are sound enough. That said, he’s clearly currently better than all of the “upside” quarterbacks—Manziel, Bortles, Thomas, Savage, Boyd, etc. Whether or not he will continue to be better than those quarterbacks when they develop is a question that still needs to be answered, but I think only Bortles and maybe Savage can pass him in this department.

Carr clearly does better on routes that have a clear timing element or break than he does with free-wheeling routes or when there aren’t well-established timing elements. He would have been a better fit in a rhythm offense, where he can use his other skills to artificially (but effectively) boost his accuracy.

Carr happens to be an extremely poor thrower on the run, especially to his left. I’ll talk about this in the next section.

Delivery: Carr can demonstrate ideal mechanics, but usually doesn’t. Generally, he almost always has good rotation with his shoulders and through his hips, but doesn’t consistently step into his throws.

The largest issue with his mechanics has more to do with what he does with his lead leg than anything else; it will occasionally lock and remove some power from his throws. At these times, he can reduce his shoulder/hip rotation as well. It’s really a simple fix—throw on a bended knee. Working against an elastic band may be all that’s necessary.

Sometimes he does set a more solid base and the ball comes out like a cannon.

His release is a little inconsistent, which is odd. It doesn’t get worse as the game goes on, or under pressure. It mostly comes when he feels he has to throw over a defender on a sideline route (generally this is a positive as the quarterback will be throwing from variable platforms, but here it is not).

Normally, he keeps the ball high and tight, and when he cocks back to throw, keeps the tip pointed up. But on those deeper sideline throws against tight coverage, the ball will drop. He doesn’t need to do that and it hurts his ball placement as well as his release time.

Sometimes (not often and usually on screens) he’ll throw outside of his frame. He needs to keep the ball closer to his body throughout the throw to reduce stress on the elbow so that the release is over the top instead of angled.

From a mechanical standpoint, his issues are more nitpicky than they are significant problems.

There are situational mechanical issues as well. His base will get narrower within ten yards of the end zone, though it may contribute significantly to what looks like excellent touch in the red zone—he throws the end zone fade as well as anyone in the draft class.

In fact, this gets to some of the other aspects of his delivery, and something that may intrigue people: Derek Carr has the best completion percentage by ten percent of any of the top quarterbacks in the draft class within ten yards of the end zone. He also has the highest adjusted net yards per attempt by an entire yard. In the 2013 class, the best quarterback (drafted in the first three rounds) was Geno Smith. In 2012 it was Russell Wilson (second was Andrew Luck) and in 2011 it was Colin Kaepernick (followed by Cam Newton).

If you include rushing data, Carr remains at the top in AY/A, with Wilson and Smith. Dalton jumps ahead of Kaepernick in the 2011 class.

If you expanded the 2014 field to include every quarterback with at least 18 passes within ten yards (the median within the set) and is draftable, Carr ranks second behind Brett Smith in AY/A and completion rate. The same holds true if you expand it to include rushing and 25 attempts.

There’s clear sample issue here to make definitive conclusions from the data alone, but it really does look like from a film perspective that Carr has unusually good touch and presence in the red zone. He also happens to have the most action-plays (passes or runs), so his number is more stable (sort of) despite being at the top.

If Carr is allowed to slow down and play in a rhythm offense, I think a lot of these issues will generally go away. An offense that will force him to install three-step and five-step drops before his throws will massively improve his footwork and stop him from rushing his throws.

In fact, in 2011, his footwork on the drop was excellent and showed a talent for hitting those routes that he had trouble with in 2013 despite being a clearly worse overall passer.

Ball Velocity: Has a superlative arm. Sometimes, the ball leaves his hand so quickly that a batted pass barely alters the trajectory of the ball, and he’ll throw the ball on a rope at any distance. He also understands when high-arc passes are necessary to get over defenders, without putting the ball at risk from aggressive defensive backs.

Like any strong-armed quarterback, there are moments where he fires throws into his receivers when he doesn’t need to. It doesn’t happens as often with him as it does with those other passers, but it’s something to watch out for. He still throws the “bucket” passes that receivers love, as well as the “frozen rope” passes that are so difficult to defend.

The bigger issue is at intermediate distances, as the short and deep speeds are contextually appropriate, and it usually isn’t on a crossing route (which were relatively rare in the offense anyway) that you’ll see too much velocity, and more likely a comeback route.

When in a mechanically poor position, Carr’s throws still exceed NFL speed when they need to, and it will probably bail him out for years to come.

Field Sense: Carr clearly demonstrates situational awareness, and he knows to avoid passes if they won’t get the yardage he needs.

He regularly goes through progressions and hits the second read. I have never seen a quarterback who plays so aggressively and overeager also pull back so often on bad throws. He absolutely won’t force it, but he loves taking shots against favorable looks. Carr’s play is marked by an odd mix of aggressive and cautious play.

By that I mean he’ll get too excited when he sees a defensive hole to exploit, rush his throws, and hurt his accuracy. But that aggressiveness turns into cautiousness when he sees a hole close or a defensive look change, and he shows extremely impressive ball control when he pulls back on a throw to go somewhere else.

I do think that despite his ability to get to his second read, he locks on to his primary receiver far too often, and like Garoppolo will overrely on pump fakes in order to prevent defenses from keying in on that tendency.

The thing is, he knows when someone is open, which is why he so rarely throws interceptions and bad passes. But without a natural instinct to move on to other parts of the field, he’ll simply wait for his receiver to beat his defender until it becomes obvious he won’t.

Far too often, it seems like Carr plays against an opposing defense instead of with his current offense. In many ways, I think this is frustration with his offensive system and coordinator.

He does try to extend plays a little bit too often, like Rodgers (this is not meant to be a good comparison) but without the escapability of players that are typically as mobile as he is. To make one positive Rodgers comparison, he does throw where there are likely to be flags, meaning he’ll intentionally throw an inaccurate ball in order to get the penalty yardage. This is an uncommon presence of mind for a young quarterback if I’m right about that.

Carr is brilliant at preventing negative plays, pressure or not. The fact that he was pressure on over 30% of his snaps but only took 12 sacks (on 694 passing attempts) is crazy. It doesn’t matter that he had a lot of screen passes—the vast majority of screen passes I logged, he didn’t have pressure. Almost all of the pressure recorded came from more traditional passing distances with the real threat of a sack.

He was one of the most pressured quarterbacks in the FBS and still produced gaudy numbers. Regardless of competition, that’s an incredible feat and well worth noting. Add to that the fact that he had one of the highest drop rates among draft-eligible quarterbacks and it’s fairly clear that Carr’s statistics don’t speak to who he is, even after taking into account average time in pocket (2.4 seconds) and distance of passes thrown. On snaps thrown past the LOS, Carr’s time in pocket averaged 2.7 seconds, the same as any NFL quarterback.

Carr is judicious in his use of third-read checkdowns and throwaways to prevent those turnovers and interceptions, and even quarterbacks with impressive pocket presence produce many more negative plays under pressure.

When you adjust his numbers for distance thrown to a traditional NFL offense, Carr still leads the class in INT rate, an astonishing 1.7%.

In general, Carr is very good at diagnosing coverages, even when they grow complex and specific. He adjusts his throws in man vs. zone coverage, and keeps his route checks specific to the habits and general look he’ll see on the field. Carr is very, very good against zone coverage but needs to do more to help his receiver against man coverage. That said, playing man coverage against him doesn’t automatically put him at a disadvantage, he’s simply very good at finding seams.

He has full control of the protections, and has done an OK job with that despite the high number of pressures coming his way—he usually correctly identifies the blitzer or blitzers and can even recognize line twists before they happen. He has a lot of control at the line in terms of audibles and route adjustments, but the nature of his offense didn’t seem to include route packages—meaning it seems like he could change individual routes but not whole plays, which really limits the effectiveness of his on-field smarts.

His ability within that context is fantastic, though. His ability to recognize defenses has allowed him to attack things he specifically sees and eviscerate defenses if he feels he has a matchup advantage. This enthusiasm for receivers he feel will be open creates the issue with his eye discipline and locking on to receivers.

Carr diagnoses defenses very well and will go for the throat when he sees something, but he doesn’t force passes or make careless football decisions. He makes his decisions quickly in some contexts (his presnap read is almost always right and very quick) but slowly in others (waiting for a player to get open instead of moving on).

In the NFL, it will be important for him to be able to look off of safeties and make the play. I’m somewhat confident Carr can do it; he occasionally reads the middle-of-field coverage in his dropback post-snap before picking a side (when he hasn’t been predelegated a presnap read) and can hold the safety. Because his offense asks him to throw when the opportunity rises instead of in time with things, it doesn’t behoove him to look off the safety as much as for most quarterbacks and he uses the general defensive read to determine his throws. That said, he needs to do hold his eyes much more.

Sometimes, he’ll stare down one receiver and throw to another, which at least says something to his peripheral ability to hold a defense, and may mean that his ability to hold the middle of the field with his eyes is underrated. When he was a redshirt sophomore, he was slightly better at this, which at least implies he doesn’t need to see his receiver to see if he’s open.

Like Bridgewater, Carr plays in a progression system that prioritizes the short routes and climbs up the field, instead of the reverse or outside/in or inside/out reads (or clear key progressions). I really am beginning to dislike the system in terms of what it does for NFL evaluation, but I understand why teams do it: it’s easy and creates high-percentage plays.

For Carr, because the offense is so substantially different than Bridgewater’s, the evaluation challenges are different as well. Teddy didn’t have as many screens and 1-4 yard passes in the route packages and therefore had to make more difficult reads. For Carr, the screen was almost always open and almost always the first read. When he pulls back on that read, though, his quick decision to move to the next receiver is on-point.

To belabor the argument, if Carr is in a rhythm offense and his reads systemized, I imagine he’ll be much better.

Pocket Presence: I don’t really understand the meme involving his issues under pressure. He clearly has some issues with pressure, but they are well overstated. He has more snaps with a poor reaction to pressure than most draftable quarterbacks, but he also has more snaps with an excellent reaction to pressure. There are two reasons for this: he threw more passes than anyone else in the draft class, and his offensive line was a dumpster fire.

It’s rare to see top college quarterbacks under that much pressure that often. It does happen at around that rate in the NFL, so it’s not worthy of “excuse-making,” but it does provide context.

The issues are there, but he often leans back as he throws, and is willing to make throws as he’s walking backwards. A lot of his throws under pressure look like the following:

It’s a bad process that puts too much air on the ball and could be ripe for interceptions. His response to this issue is to be extremely risk-averse (see above, in accuracy) and throw away from the defender, though this doesn’t always create a catchable ball for the receiver. In general, leaning back is bad because it hurts velocity, creates inaccuracy and is a proxy for other issues with a QB who is scared of pressure.

In addition to that lean, he’ll narrow his base and throw with a locked front leg against pressure far more often than he will without pressure and throw off his back foot much more often.

Carr also occasionally rushes the throw. Many times he will do the same thing without pressure, which to me is more a symptom of the fast-paced, hurry-up offense than it is the desire to remove pressure from the equation (he will do it when it is obvious that there can be no pressure). These hurried passes are often his most inaccurate.

He also has plays like this:

That’s absurdly impressive and it happens from time to time. More often his “good” plays against pressure look like this:

Has run away from phantom pressure, but it isn’t as frequent as Garoppolo, Manziel or other QBs projected in the first three rounds (Bridgewater and Bortles aside). It was easy to see twice for sure in the Rutgers game (one was an end zone interception, one was a thrown away pass after extending the play), and one or two times where he rushed a throw, but could simply be overeager to attack really favorable defensive looks.

He did it thrice against San Diego State, the worst regular-season game in terms of pressures given up. He produced one mediocre play (run for two yards), one spectacular play (44 yard sideline throw that bounces in and out of his receiver’s hands) and one good play (overtime catch-and-run for 27 yards).

In most games, he does not do it at all.

The frequency with which he abandons clean pockets with free receivers is nothing like Manziel, who many have pointed out since I wrote my report has a record of constantly abandoning kill-shots or good open receivers across the middle in favor of scrambling.

Derek Carr is difficult to diagnose in this context. For all the trouble Manziel caused in evaluation, it was easier to argue that he was good at diagnosing pressure but overreacted. For Carr, I don’t think he is scared of pressure (although he certainly was as a redshirt sophomore), so much as hurried by game context. When the game is moving at his pace, his reaction to pressure is pretty good. When the game is moving very quickly for no real reason, he accelerates the clock in his head. Many times, watching the film, I feel that If Carr went 5% slower, the offense would move 10% more efficiently.

Further evidence for his variable pocket presence can be captured in a small part with this image gallery, which includes every snap Carr took under pressure in three games, when they would upload (about 75% of the GIFs recorded were uploaded).

They are in chronological order in those games, and the general sense I’m getting is improved pocket presence over the course of the game. Later in games, he’ll step up into the pocket (with a good base and a bended front knee), throw down the barrel of a gun, and even take the time to reholster and choose a different target before throwing with defensive tackles nearly draped over him.  Those aren’t the qualities of a quarterback scared of pressure.

This doesn’t mean it’s not a concern. You don’t want your quarterback to react poorly to the two-minute drill with the game on the line. But it is far more manageable and correctable than a quarterback who is simply afraid. In the first case, you tell the quarterback to stop dropping his back shoulder. In the second case, you draft a new quarterback.

The fact that Carr had the highest percentage of his throws dropped of the quarterbacks I’ve taken a look at in this class really kills the ability to parse out specific numbers and look at why he’s necessarily done worse under pressure.

For a long time, when breaking down Carr’s specific skills under pressure, I dogged his ability to sense edge pressure. Upon further analysis, it’s more complicated than that. He actually does a fairly good job at sensing pressure from the edge and the interior and reacts organically to it.

He manages edge and interior pressure with mixed results, and is probably worse than the average NFL quarterback in this way. While he’ll occasionally have brilliant moments of sidestepping or stepping up, at times it seems like his film is dominated with less than ideal responses to that pressure.

Carr isn’t particularly talented at resetting before his throws, unlike Bortles and Bridgewater. If he gets moved off of his spot, he’ll rely on his arm strength to bail him out, and it hurts his accuracy. In that sense, his pressure management is subpar. Giving himself that kind of time could have hurt him at Fresno State given the porous line, but that’s no excuse—he needs to do a better job of figuring out when he needs to fire the ball away and when he has time to re-establish good mechanics.

One last note: Carr’s progression in terms of pocket presence has improved massively every year he’s started. In 2011, he was almost cowardly. In 2012, he was tense and far too nervous. This last year, he was inconsistent, but definitely willing to take hits. He’s tough in the pocket, has developed a sense for rushers and gets more comfortable over the course of games.

Scrambling/Running: Derek Carr is an extremely underrated runner, with speed, functional agility and decent vision. The best example comes from when he was slower, but just looks so damn nice:

Carr - On the Run

Carr’s athleticism has been extremely underutilized in his time at Fresno State, which is fine because he was a deadly efficient pocket passer. His scrambles downfield and designed runs have been largely very effective, and he could integrate option running or packaged play concepts as well as any other quarterback in the class.

He doesn’t possess the insane vision and preternatural running ability of Johnny Manziel or the fundamental build and balance of Blake Bortles, but his running ability could very well be the focus of an offense (although I would really recommend against it, just because of the efficacy of those offenses).

At the NFL level, this could make him the kind of quarterback that can be given the permission to damage a defense that sells out too much for the pass or drops too deep to respect his arm.

Unfortunately, his inability to throw on the run really limits the use value of this trait and he doesn’t really move laterally with a great amount of speed, so it’s an occasional weapon.

Intangibles: Carr’s intangibles are evidently through the roof. There’s not much to say besides the fact that he continues to will his team, even in completely lost games (2012 Oregon, 2013 USC) and let his team to a win in overtime against SDSU. His work ethic is supposed to be one of the best in the class; better than Bridgewater, Bortles, Murray or McCarron. Of course, it is sort of an impossible thing to quantify and compare, so let’s just stick with a much more reasonable evaluation of “great.”

It is difficult to speak much to Carr’s on-field intangibles largely because his off-field character positives overshadow most reports and discussions of his on-field ability to lead teammates. In case you didn’t know, he’s very Christian and is extremely devoted to his family. This doesn’t likely translate into any additional on-field success, but at least his character is probably not a concern.

Primary Concerns: The Las Vegas Bowl against USC seems to be the first thing that detractors talk about in regards to Derek Carr. With some high profile misses and inaccuracies, it certainly looks bad. His statistics from the game: 30 for 54 (54!?), 55.6%, 217 yards, 4.0 yards per attempt, 2 touchdowns, 1 interception and 3.9 adjusted yards per attempt (same as adjusted net yards per attempt because there were no sacks).

I’ve catalogued a lot of Derek Carr’s passes under pressure, like I mentioned before. Again, not all the GIFs uploaded, but if you look at the USC game, 7 of the 11 passes (that uploaded) hit his receivers in the hands (and one was a throwaway). That doesn’t mean he had the best ball placement (he didn’t), but his performance under pressure was far better than his detractors say and better than some starting quarterbacks in the NFL.

In total, 11 of the 16 aimed passes under pressure in that game hit his receiver in a catchable position. 13 of all of his catchable passes throughout the game were not hauled in—the highest total number and percentage of dropped passes I’ve seen in a single game I’ve tracked. 41 of the 50 eligible passes (one throwaway, two receiver miscommunications and one rocket sweep that is really an extended handoff but is counted as a pass in the box score) hit their receiver.

That isn’t quite the proportion of passes that had good ball placement in a position that is easy to catch, and some of the drops were more his fault than his receiver, but the story of the game is overstated. Hitting receivers on 82% of throws in a game where they were under pressure as often as they were is impressive, not poor.

There are a lot of issues people have with his system, and it makes sense. Like Manziel, Carr’s system did not include the same keys or progressions as an NFL offense typically would. He didn’t have the same rhythm or footwork system he will likely be asked to operate out of at the next level and theoretically little experience commanding those systems.

He did, however, play in a pro-style offense in 2011 under Pat Hill. His 2011 film is kind of what you would expect for his first season as a starter (it was his redshirt sophomore year). His issues with pressure were much worse (and consistently worse) and his accuracy was frankly awful.

But his ability to go through reads was impressive and he regularly hit his second and third read, making the correct decision at a rate that far exceeds those with his experience level. He wasn’t Peyton Manning, but it was good, though he trusted his arm too much in tight windows. He no longer forces throws, and hopefully a system change won’t bring him back to that.

Carr is generally accurate, and he has long stretches of play with surgical strikes, but when he misses, sometimes they are enormous misses. The nature of the misses makes you think he is more inaccurate in general than he usually is, but wild misses still raise eyebrows. The good thing is he always errs towards safety and away from defenders.

Conclusion: Carr is a bad fit for the offense he was in, despite the gaudy stats Fresno State put up. It’s not just that they hid his strengths, they magnified some weaknesses, removing rhythm from the game and creating an awkward read system without implementing an easy-to-audible route package system.

That’s not to say it was a bad offense in general; it clearly worked for Fresno State and Dave Schramm. But Carr in a system where the coaches can manage him for a year or two will be good. If given clear directives on his reads and priorities, he can do very well before being let go and potentially have a lot more leeway.

This is somewhat of a knock on Carr, but there are honestly very few unmanaged quarterbacks in the NFL. Brady and Belichick meet weekly to discuss changes and Brady therefore has a lot of control over the way he makes progressions (though he had very little control early on in his first two Super Bowl wins). Peyton Manning famously manages himself and Rodgers is a highly managed quarterback that might do better when unmanaged.

But there is a long history of quarterbacks who have been built within their systems and managed, like Joe Montana. Notably, Turner quarterbacks (and in general, Coryell passers) were highly managed as well. Carr also projects well, to me, to an Erhardt-Perkins offense with modular route packages that he can sub in an out at will with clear keys.

For Bridgewater, it is easy to see that an NFL system is a good fit, even if it comes in a different flavor. But Carr requires an offense that runs with completely different principles than the one he’s currently in; Carr is not a fit for his own offense at the NFL level.

His talent-level depends on your interpretation of my characterization of his weaknesses. If you agree with what I said, he’s a first-round level talent (after correcting for the fact that quarterbacks are pushed up in the draft), but not a top-ten type prospect. But if you don’t think there’s a good explanation for his pressure issues, he’s a late-second round prospect.

If you also happen to think my conjecture about the improvements to be gained from a system change are wrong, then he is probably a fourth-round prospect. If you agree with my take on his pocket presence but not about the immediate benefits of a system change, then I think he is a high-second round prospect for the right team and a high third-round prospect for a generic team.