[A note, there are a lot of GIFs. In this piece, it's not as necessary to see them to get the point, but the post is better served with them than without]
Without question, Josh Robinson has been one of the worst cornerbacks in the NFL this past year.
It’s not a controversial statement, but even his biggest critics hold out hope for the young speedster from Central Florida. It’s a unique circumstance where a player finds himself maligned all season, but fans have a cautiously optimistic outlook on his future.
Robinson was miscast last year as a slot cornerback, required to defend space instead of the sideline. Often the optimism for Robinson stems not just from this fact (2013 was his first experience playing the slot), but also from the fact that Mike Zimmer, more than at any other position, is a defensive backs’ whisperer.
As a position coach or defensive coordinator, he’s turned otherwise marginal or castoff talents at defensive back into stars. At Dallas, Brock Marion went from a subpackage player to a ballhawk, and Larry Brown did the same, immediately doubling his career interception total in his first year with Zimmer as a positional coach. Darren Woodson didn’t earn his first Pro Bowl or All-Pro recognition until Mike Zimmer arrived. Despite being a three-year veteran, he hadn’t grabbed an interception until then.
His one year in Atlanta was the only year safety Chris Crocker grabbed three or more interceptions (until he did it in 2012 as a subpackage player… under Mike Zimmer in Cincinnati). After earning All-Pro and Pro Bowl honors working with Zimmer in Dallas, Roy Williams’ play fell off until he joined Zimmer again in Cincinnati, with two good years before retirement.
After disappointing years in Jacksonville, Reggie Nelson took Williams’ place and started playing as an above average starter in Zimmer’s defense. He reunited with Terence Newman in Cincinnati, too, after putting together two extremely disappointing seasons in Dallas in 2010 and 2011. He immediately become an impact starter for Cincinnati after joining Zimmer again.
Adam “Pacman” Jones had a number of off-field altercations that saw him leave the league more than once before joining Cincinnati where he played at a high level again, but without the off-field concerns. Combined with all that are the positive contributions from undrafted free agents and late-round picks, like safety George Iloka and Nedu Ndukwe.
If there’s anyone that Vikings fans can trust with turning around defensive backs, it’s Mike Zimmer. But Zimmer isn’t magic; he cannot create skill where there is none.
There is a small but reasonable list of mid-to-late round defensive backs that haven’t taken the league by storm that have arrived in Zimmer’s career. But, there’s a very good chance that if Robinson is talented that Mike Zimmer can extract it.
Unfortunately some of the ways used to evaluate players are better for some positions than others—it’s far easier, for example, to measure the impact of running backs or quarterbacks because of statistical data than it is for offensive linemen or defensive backs.
For the most part, the ability to derive talent from coverage statistics comes from schematic differences and sample size. While there are usually a reasonable number of coverage snaps to draw from, statistics that rely on rate by target are a bit wanting, as full-time cornerbacks will be targeted anywhere between 60-100 times a year—decent, but not robust.
Further, cornerback statistics are more team-dependent than almost any other positional grouping, despite the fact that corners are sometimes the most individual players on the field. Defensive backs who are targeted often may not necessarily be bad; they could be very good on a team of great players. So, too, for CBs that are not targeted often. They aren’t dispositively talented so much as better positioned than their teammates.
There are also different schematic responsibilities as well; cornerbacks who are lined up against a team’s top receiver often will find themselves subject to more targets than a cornerback who is not, and perhaps even more than a cornerback on the same team who is simply shadowing the second-best receiver.
In addition, zone players are expected to give up short-yardage passes and limit gains, which may make their target and reception total incomparable in significance to man coverage players.
Some quarterbacks will audible to a specific route combination based on coverage and could prefer “coverage-beaters” that don’t rely on individual cornerback skill in order to make a completion—a corner route against the Cover-2 is a good example, or bunch routes against man coverage.
Regardless, many cornerback statistics can still provide us with a clue as to the general quality of corners in the league. Often the best cornerbacks, regardless of scheme, will lead in these statistics. The same goes for those at the bottom of the barrel, like Josh Robinson.
Perhaps the best single statistic for cornerbacks has nothing to do with touchdowns or interceptions—those are too easily skewed by one piece of data at any given time—but yards. Because CBs often are asked to give up short-yardage passes to prevent a first down, catch rate isn’t always the best (though it has its uses as a red-flag indicator). The yards statistic does a decent job capturing that instead—a zone corner that gives up a short-yardage pass but gets off the field is better rewarded this way.
Yards per coverage snap is also probably superior to yards per target given up, simply because Y/CS gives credit to defensive backs who prevent targets from superior coverage. Beyond that, looking to adjusted yards per snap in coverage allows us to get a slightly more complete picture, because it counts touchdowns and interceptions—and adjusted yards per target allows us to figure out what happens when corners are in fact targeted.
With that, we can look to targets per snap in coverage to see how “boom-bust” a corner can be. A corner who gives up more yards per snap in coverage but sees a low target rate is still somewhat valuable because they only give up a few plays a game and require the quarterback to take risks to do it.
Listed are the cornerback ranks, out of 91, for Vikings cornerbacks in 2013 in each of the statistics mentioned, as well as the average number of snaps in coverage a cornerback took before allowing a reception (CS/R). I’ve bolded yards per coverage snap mostly to indicate my preference for it among all of the relatively incomplete statistics.
It might be a little surprising to see Cook rated so well in any particular category, but given that his diagnosis is often about his ability to react to the ball in the air, not his positioning, it fits what we know about him.
Particularly striking is the fact that while Robinson was not targeted more than the average cornerback—and not particularly often among Vikings cornerbacks—he still gave up a wealth of yards and a few touchdowns. This speaks to the same issue as Chris Cook, but to a far worse degree.
For Vikings fans however, there has been a common refrain—and a central argument to his prospects as a salvageable defensive back: that he’s been far better on the outside than in the slot. With a number of his snaps in the slot, aggregate statistics about his play may not capture the kind of player he is. If one takes a look at the 46 cornerbacks who have had 100 snaps in the slot, there’s definitely an argument to be made:
Again, that’s out of 46.
The biggest argument in favor of Robinson was that when he wasn’t miscast in his slot role, he was theoretically fine. Generally speaking, he’s done a better job staying in position and deterring quarterbacks from throwing his way, but even when playing outside he hasn’t been as good as what many say was a promising 2012 showing.
The evidence doesn’t quite bear out in favor of Robinson when his snaps are exclusively isolated to the outside (ranked out of 78 corners):
It’s better, but not by much (and certainly not by enough). This seems in stark contrast to what Robinson put together in 2012, where the statistics paint him in a much better light (out of 80 corners):
Generally speaking, these aren’t bad in the abstract, but perhaps there are warning flags in how often he was targeted relative to his teammates or the fact that he gave up many, many more receptions on the outside than his teammates.
Combining the two years isn’t great for Robinson, nor is it particularly instructive. He was functionally asked to play two completely different roles and had different comfort levels for the defense on either year. Still, on the outside alone, he combined to be a below average corner.
If he used his combined statistics for the two years, he would have ranked 62nd, 64th, 44th, 57th and 77th in those respective categories on the outside among 2013 corners. Not exactly confidence inspiring, and again, not as relevant as the fact that he was supposed to be an average corner his rookie year.
His 2012 statistics may be more misleading than most, however. Despite a relatively low number of yards given up when in coverage, he did give up the second-highest catch percentage of cornerbacks on the outside (Jefferson allowed the seventh-most, Winfield 25th, and Cook 37th)
And as long as we’re using Pro Football Focus data (where all the statistics come from), it’s relevant to point out that on a per-snap basis, PFF was none too impressed with Robinson. While his statistics are somewhat friendly, his grades were not. Of all corners with at least 100 coverage snaps (123), he ranked 120th in grade per snap in coverage in 2012. In 2013, he graded out 112th of 132 corners. Perhaps neither year was particularly good for him.
PFF tends to grade almost exclusively using broadcast information, which means they give a small bonus every time a cornerback is not targeted while also grading every snap they are targeted—taking care to assign credit and blame for the result of plays as necessary. That means they’ll often miss an opportunity to grade poor coverage when quarterbacks didn’t take advantage of them.
That means two things: 1) when quarterbacks targeted Robinson, bad things happened much more often than good things and 2) they really missed an opportunity to grade him as even worse than he was.
The ways that Robinson have found to give up coverage are stunning. There are times when he looks confused in coverage, but mostly he seems to be massively out of position, even on the outside.
He’s not good in transition and seems to be reading the receiver incorrectly, falling for receiver and quarterback fakes. This isn’t the entirety of what he does; you can take a look below at what looks to be a Cover-4 with a Tampa-2 element:
However, that’s rarer than it is common, and you’ll often see him bite. Sometimes this is because he’ll play man coverage like a zone corner, reading the quarterback instead of the receiver, but he’s not been particularly good reading receivers either. I much prefer Robinson to play in press coverage, be it press zone or press man, but it’s not great.
He also suffers from the old Vikings defensive back canard—when he has good positioning, he doesn’t locate the ball:
He played tighter coverage more consistently in 2012, which resolved his biggest issue (sudden route stems) to a small degree, but didn’t solve them. Often you’d see a disconnect between how he played shallow cuts (fades, posts and go routes) and how he played sharper cuts (comebacks, square-ins)… and apologies for the poorer quality; All-22 was in terrible definition until 2013:
Though, even corner routes had him beat at times—which is designed to attack the seam between the safety and the corner. If there isn’t any expected safety help, then the corner should not have as much difficulty against it:
This is a fairly classic Tampa-2 coverage call, and the issue isn’t the call—it’s Robinson’s read and adaptability. Four verticals is the call against the Tampa-2, which is one of the two most common Tampa-2 beaters and they’ve combined it with a late-breaking flag/corner route to put additional stress on the safeties (the other most common Tampa-2 beater). There are some different rules for defenders in this situation, but A.J. Jefferson (top of the screen) and the two safeties are playing it correctly while Robinson is not.
Because the middle linebacker cannot cover two seam threats and because the deep drop of the Sam (Greenway), Robinson needs to let the safety take the inside receiver (so that Greenway and Smith bracket him) and Robinson takes the outside receiver, just like Jefferson did on the other side of the field.
But the issue may be that Robinson also did not anticipate the receiver breaking to the corner, which of course creates massive problems that the Vikings didn’t happen to pay for. He had this issue several times in the game, including this more difficult to diagnose one:
There’s an argument to be had that Greenway carried his receiver a bit too far (it’s Henderson’s assignment after the receiver passes the 38 or so), but that doesn’t really affect the fact that Robinson needs to squeeze the zone and shift inwards. Zone assignments are not static; they’re fluid and need to adapt to the offense’s hand. In this case, moving the zone to squeeze the receiver is the right call, because it gives leverage to the other players on the field. Instead, the receiver breaks free.
In all honesty, it looks like Robinson knew that he had to move his zone landmark in response, but didn’t anticipate the receiver’s cut seven yards into the route.
Crafty receivers had a good day against Robinson, too:
It’s mostly in defending those routes with weak angles, or when receivers round out their cuts, that he excels:
That may not inspire confidence, and it shouldn’t, but at least his success rate on those kinds of routes were better than his peers at the position. That said, the small difference in ability against these easier-to-defend routes by no means makes up for the gulf in ability he has against moderately difficult or hard-to-defend routes by any stretch of the imagination.
To say that he’s been good on the outside is a bit of a deception. Certainly, he’s far better on the outside than he is in the slot, but that doesn’t make him “decent” on the outside, either—something I had thought was true until I did this work.
Further, Robinson hasn’t improved much between 2012 and 2013 on the outside.
Mostly outlined are faults of his in man coverage, but given the prevalence of man coverage concepts in Zimmer’s scheme, it’s worth emphasizing. Further, they do appear in zone coverage—route recognition and reading receivers is a failing of his, and a big part of what make Richard Sherman and Charles Tillman so good at playing in zone schemes is their ability to read routes and anticipate.
This isn’t the death of Robinson as a starting corner, or even as a high-level one; that part of his career is mostly on life support. He might simply be better served by playing even tighter coverage and allowing himself to feel the receiver throughout the route through constant contact, like Darrelle Revis.
A consistent critique of Revis is that he’s constantly holding when he’s not—he’s riding the line, but keeps a feel for the receiver by resting a hand on the receiver’s arm throughout the route. Revis is a master at using subtle signals, like changes in muscle tension, a receivers eyes, or shoulder alignment to determine where a receiver is going to go.
Bengals corners seem to do this fairly consistently. Both Terence Newman and Pacman Jones find ways to anticipate receivers in part due to their constant contact with receivers.
I think the adoption of this technique will be a positive for Robinson, who has the athleticism—both straight line and laterally—to work with the smaller margins of error a technique like this demands. It also allows him to use his physicality and strength to his advantage. From an attitude and baseline athletic talent perspective, it’s good for him and a fit for a corner who originally looked better in zone than man.
That said, it requires a lot of him that he’s proven not to have. It still requires great route recognition, and good play on the ball in the air. Robinson also hasn’t shown the patience to wait on flipping his hips, something that the Bengals defense asked its corners to do—only committing when the investment made sense.
Even Robinson’s impressive recovery speed doesn’t make up for the fact that he had opened the gate and should have allowed an easy completion for 17 or so yards in the air and anywhere between 5 and 38 yards on the ground. Opening his body to the wrong direction allowed the receiver an easy break inside—which is an issue he had in 2012, too.
It’s not often a simple thing to diagnose, because of the techniques encouraged by different schemes—attempting to funnel receivers inside or outside as the case may be. But it happens often enough in man coverage when Robinson’s incentives are clear. Zimmer’s defense would rather that corners be patient when they’re not in tight coverage.
You’ll also see a lot of mixed coverage concepts in these Bengals GIFs—man coverage on one side and zone on the other, with different sets of rules governing what players do when different route combinations appear. Given Robinson’s issues with developing chemistry and an inability to drop receivers off into the next zone at the appropriate time, mixed coverages may be quite the worry.
For the most part, it is extremely unlikely that Josh Robinson transitions into a starting-caliber corner from here on out. My defense of him as an outside corner was probably misguided and he’s had some of these issues for quite some time—issues that haven’t gone away or improved in the two years he’s been in the NFL.
Robinson is physically ideal for the slot. He has quick feet, excellent body control, all kinds of athletic ability and agility and a tackler’s frame. But he hasn’t put it together on the outside or in the slot, and his continued spot on the roster demands that he answer the question of whether or not he can be saved.
He’s in the perfect situation with the perfect coach to revive his career, but the odds are longer than they are short and it may be time to look towards the 2015 draft for cornerback help if Derek Cox, Jabari Price, Kendall James, Shaun Prater or even Julian Posey can’t step up in an enormous way.