Notes: Data depicts Lions WRs with very distinct types

Seth Walder at ESPN Analytics posted a very cool plot on Wednesday with a Twitter thread explaining the process by which he arrived at it. First, here is the chart:

The Y-axis is a normalized scale showing how much more or less than other receivers in comparable situations would be expected to catch the ball. A major issue with figuring out if a receiver making contested catches is being contested because he could not lose his man with a route or move is situational. What if the defense is defending third-and-4 in a tight game situation between the 40-yard lines and they are crowding in expectation of a quick route to the sticks? What if they are defending the goal line? In some situations, the defense is going to deliberately crowd receivers and in other situations they will give more cushion.

Walder points out in his thread that it’s not just the depth of target/air yards of the throw that matters, but also the “air yards to sticks.” In attempting to create an expected amount of separation that a receiver should be getting, Walder tries to account for the situational yards the offense should be trying to pick up for a first down conversion. Using data from the last three seasons (back to 2017), he gets rid of stuff behind the line of scrimmage like tunnel screens because the defense can’t pre-jam the receiver on their side of the neutral zone and converts that to a normalized measure along the X-axis of his chart.

Naturally, this is a Detroit Lions blog, so our readers were no doubt looking for a particular name in the chart the second they laid eyes on it:

When you locate Golladay’s name clutching the left edge of the chart but way up the side in positive catch-over-expectation territory, this agrees with the conventional wisdom around his fantastic 2019 season. But then look a little further up, and you see the other outside receiver on the Lions:

The 64 dollar question is what is going on here. Are Marvin and Kenny G actually bad at getting separation or is there something about the kinds of targets they are getting that makes defenses crowd them harder than perhaps even Walder’s model is accounting for? This is, the analyst admits, an unfinished project so we really don’t know:

That’s not a lot of sample size and I have no idea what the standard errors (or if an adjusted R-square would even be meaningful) are, so insert shrug emoji here. Most of our readers are going to roll their eyes at how obvious the next sentence is, but it is important that new analytical tools make sense and agree with what we observe — and this does, which is good! The Lions appear to have two distinct types of wide receivers on their roster generating data in Walder’s setup: the pair of outside gladiators fighting for 50-50 balls and a pair of slot guys who are schemed into vacant lots via solid play design.

This goes back to the air yards to sticks plot that Walder put up in in an early part of his thread. If you are throwing deep and way beyond the sticks, then you’re talking about much more room to work with and green space the defender has to worry about; that’s what he’s talking about when he mentions the downward slope below zero and the flat line after the first down marker. If you’re hitting the goal line up close or aiming for first downs, that’s a lot less space to make moves in. Marvin Hall didn’t have a lot of targets and did not qualify, but consider Danny Amendola way further to the right of where the contested catch bros are and the kinds of plays we saw worked to him. Obviously, even the throws to the crafty slot veteran in front of the sticks are designed for more schemed separation: you probably don’t want to use Kenny Golladay as a shallow crosser off a bunch rub setup.

As mentioned, Walder’s chart is a work-in-progress, but still neat to see stuff we’ve pondered sliced and diced in the data. Now, on to the rest of today’s Notes:

• Speaking of wide receivers:

• Dave Birkett at the Detroit Free Press posted his first 53-man roster projection, and it is missing the player who has spent the most seasons in Detroit entering training camp. Kyle Meinke from MLive, meanwhile, has a less-controversial 53-man roster prediction that stays with the tried-and-true.

• From Max Gerber at Lions Wire, we have praise from former NFL scout Daniel Kelly on the Split Division podcast (of which Max is one of the hosts) for several Lions players and head coach Matt Patricia.

• As if you needed another reason to be excited about the potential of Detroit’s young defensive backs like Tracy Walker, Amani Oruwariye, and Jeff Okudah, here’s another. James Light reminds us of the meticulous detail work emphasized by coaches like Matt Patricia and Cory Undlin, who came up in the Parcells/Belichick/Crennel traditions:

• ESPN’s Michael Rothstein wrote about COVID precautions taken by the Lions organization leading up to training camp.

• Video training isn’t just for professional football players! The Lions organization set up an instructional site for youth football players to keep making progress during this very weird year: