Writers of Pro Football Prospectus 2008

Most Recent FO Features

Flavortown.jpg

» The Week In Quotes: August 18, 2017

In the world of NFL quotes, Marshawn Lynch is the gift that keeps on giving. Plus: some great plays in practice, some terrible plays in practice, wicked rhymes, and Guy Fieri.

27 Jul 2017

Jarvis Landry's Value, Part II

by Scott Kacsmar

In Part I of our study of Jarvis Landry's value, we compared him to 27 other top wide receivers since 2014, looking at correlation between each receiver's production and his team's passing DVOA. Landry was one of two players with negative correlation (-0.14) between his receptions and his team's passing DVOA, and the Dolphins have actually performed much better when he wasn't a focal point in the offense.

Today we are going to look at the root cause for why Landry's catches are less valuable than those of his peers, and present some new charting data that can be found in Football Outsiders Almanac 2017.

Receiving DYAR Since 2014

We have never claimed that our individual stats are perfect tools to isolate individual performance. Something like receiving DYAR for a wide receiver is obviously affected by the quality of his quarterback. Still, it's not like Ryan Tannehill has been terrible in Miami.

In 2016, Landry's career-high 170 DYAR was barely more than his teammates DeVante Parker (139 DYAR) or Kenny Stills (120 DYAR) had on more than 40 fewer targets. In 2015, Rishard Matthews (235 DYAR) and Parker (93 DYAR) had more DYAR than Landry (72 DYAR) even though Landry led the team in targets again. A similar event happened in 2014, when Mike Wallace (221 DYAR) more than doubled up a rookie Landry's 102 DYAR.

It's not that Miami's passing game is incapable of producing receiver value. It's just that Landry's value has been limited despite a large number of targets over the last three years. The following table compares Landry to those same 27 wideouts from Part I, looking at their receiving DYAR since 2014. We also included average passing DVOA for each receiver's team in the games he has played during the last three seasons. This is not weighted by attempts and treats each game equally. This is also passing DVOA at the team level rather than an individual quarterback's DVOA, so the average is actually above 0% (because runs and penalties average below 0%).

WRs: Average Passing Offense DVOA and Receiving DYAR Since 2014
Player Team Games Avg Pass
Off DVOA
Rk DYAR Rk 2014 2015 2016
Antonio Brown PIT 47 36.4% 2 1,347 1 554 517 276
Julio Jones ATL 45 27.1% 8 1,157 2 356 343 458
Odell Beckham NYG 43 14.0% 18 862 3 396 304 162
Doug Baldwin SEA 48 27.6% 7 823 4 137 414 272
A.J. Green CIN 39 30.7% 5 820 5 158 414 248
Mike Evans TB 46 9.3% 22 716 6 222 187 307
T.Y. Hilton IND 47 10.2% 20 715 7 303 52 360
Randall Cobb GB 45 33.1% 3 679 8 479 77 123
Emmanuel Sanders DEN 47 16.4% 16 673 9 481 90 102
DeSean Jackson WAS 40 19.6% 10 645 10 306 97 242
DeAndre Hopkins HOU 48 -1.3% 25 557 11 237 268 52
Brandin Cooks NO 42 30.9% 4 540 12 124 192 224
Alshon Jeffery CHI 37 8.4% 23 538 13 278 126 134
Demaryius Thomas DEN 48 16.9% 15 534 14 317 56 161
Player Team Games Avg Pass
Off DVOA
Rk DYAR Rk 2014 2015 2016
Eric Decker NYJ 33 10.0% 21 525 15 199 278 48
Dez Bryant DAL 38 29.2% 6 497 16 430 -86 153
Larry Fitzgerald ARI 46 19.0% 12 494 17 54 363 77
Kenny Britt LARM 47 -17.5% 28 469 18 163 139 167
Golden Tate DET 48 15.6% 17 441 19 214 113 114
Sammy Watkins BUF 37 19.2% 11 429 20 71 312 46
John Brown ARI 46 17.0% 14 413 21 1 352 60
Pierre Garcon WAS 48 20.0% 9 376 22 -15 128 263
Steve Smith BAL 37 18.6% 13 361 23 79 125 157
Allen Robinson JAC 42 -3.7% 26 344 24 10 318 16
Jarvis Landry MIA 48 13.6% 19 344 25 102 72 170
Julian Edelman NE 39 46.1% 1 329 26 137 144 48
Jordan Matthews PHI 46 4.3% 24 305 27 194 112 -1
Allen Hurns JAC 42 -11.8% 27 151 28 -13 236 -72

Antonio Brown and Julio Jones really distance themselves from the pack with more than 1,100 DYAR each. A.J. Green (820 DYAR) would likely be right there if he had enjoyed better health last season. It probably comes as little surprise that the highest DVOA averages are for wide receivers who played in offenses with Tom Brady (Patriots), Ben Roethlisberger (Steelers), Aaron Rodgers (Packers), and Drew Brees (Saints) at quarterback, the cream of the crop. It's also probably no surprise that the opposite can be said about Kenny Britt (Rams), Allen Robinson and Allen Hurns (Jaguars), and DeAndre Hopkins (Texans), stuck with some of the worst poor quarterback play.

Miami's pass offense with Landry in the lineup ranks 19th in DVOA (13.6%), so it's a little below average in this sample, but not too bad. He still only ranks 25th in DYAR (344), but it is also true that the bottom four receivers on the table (Landry, Julian Edelman, Jordan Matthews, and Hurns) are all primarily slot receivers. They serve a different role than some of these other receivers, though just being a slot receiver alone does not make one incapable of being highly valuable. Seattle's Doug Baldwin ranks fourth in DYAR, and he is one of the league's best slot receivers. He still makes plays down the field, and he caught 14 touchdowns in 2015, or one more than Landry has in his career. Not every slot receiver has to be in that Wes Welker-New England mold. Victor Cruz used to be a vertical slot threat for the Giants when he was healthy. T.Y. Hilton now serves a similar role in Indianapolis with Andrew Luck, and he ranks seventh in DYAR here with an average team passing DVOA that's actually ranked one spot lower than Landry's Dolphins.

Still, there is no denying that Landry is used differently than practically every wide receiver on this list. When we look at our advanced stats on these players since 2014, the difference is clear.

WRs: Advanced Metrics, 2014-2016
Player Team Passes REC +/- Rk C%+ Rk ALEX Rk PYD YAC YAC+ Rk
Sammy Watkins BUF 262 1.7 21 0.6% 21 6.4 1 14.9 3.8 -0.4 17
DeSean Jackson WAS 222 10.4 11 4.7% 7 6.4 2 15.5 6.5 0.9 3
Mike Evans TB 420 -0.1 23 0.0% 23 6.1 3 15.0 2.5 -1.5 28
John Brown ARI 261 -1.3 24 -0.5% 24 5.7 4 14.7 3.9 -0.9 27
Kenny Britt LARM 173 1.7 22 1.0% 18 5.0 5 13.5 4.0 -0.3 15
Allen Robinson JAC 361 -9.2 27 -2.5% 27 5.0 6 13.6 3.6 -0.7 22
Emmanuel Sanders DEN 400 17.9 7 4.5% 9 4.9 7 13.5 3.7 -0.8 24
DeAndre Hopkins HOU 442 2.8 19 0.6% 22 4.7 8 13.2 3.2 -0.9 26
Dez Bryant DAL 289 -3.8 25 -1.3% 26 4.6 9 13.2 4.0 -0.1 12
T.Y. Hilton IND 391 10.3 12 2.6% 14 4.5 10 13.3 4.5 0.1 9
Alshon Jeffery CHI 310 4.2 17 1.4% 17 4.3 11 12.9 4.4 -0.4 18
A.J. Green CIN 326 24.1 5 7.4% 4 4.1 12 13.0 4.1 -0.4 16
Eric Decker NYJ 252 7.5 15 3.0% 13 3.5 13 11.8 3.5 -0.7 21
Odell Beckham NYG 437 16.0 9 3.7% 11 3.0 14 11.3 5.6 1.1 1
Player Team Passes REC +/- Rk C%+ Rk ALEX Rk PYD YAC YAC+ Rk
Julio Jones ATL 473 25.6 3 5.4% 5 2.8 15 12.0 4.9 0.1 7
Brandin Cooks NO 307 24.2 4 7.9% 3 2.7 16 11.9 4.4 -0.8 23
Allen Hurns JAC 255 -9.3 28 -3.6% 28 2.3 17 11.3 5.1 0.6 5
Antonio Brown PIT 511 42.9 1 8.4% 2 2.1 18 10.5 4.3 -0.5 19
Demaryius Thomas DEN 487 -5.5 26 -1.1% 25 1.4 19 10.4 4.8 0.3 6
Pierre Garcon WAS 315 12.8 10 4.1% 10 1.1 20 10.0 3.9 -0.8 25
Steve Smith BAL 294 1.9 20 0.7% 20 1.0 21 9.8 4.7 0.0 11
Jordan Matthews PHI 334 3.1 18 0.9% 19 0.4 22 9.0 4.6 -0.2 14
Doug Baldwin SEA 312 32.3 2 10.4% 1 0.0 23 9.0 5.2 0.1 10
Julian Edelman NE 359 5.8 16 1.6% 16 -0.6 24 8.1 4.9 0.1 8
Larry Fitzgerald ARI 386 20.8 6 5.4% 6 -0.6 25 8.4 4.2 -0.6 20
Randall Cobb GB 322 9.6 14 3.0% 12 -0.8 26 7.8 6.0 1.0 2
Golden Tate DET 386 9.6 13 2.5% 15 -1.6 27 7.2 6.6 0.9 4
Jarvis Landry MIA 382 17.2 8 4.5% 8 -2.7 28 6.5 5.5 -0.1 13

Notes: REC +/- is receiving plus-minus, explained here. C%+ is looking at receiving plus-minus as a rate stat (catch rate adjusted for where the ball was thrown). ALEX is the difference between how far the pass was thrown relative to yards needed for a first down. PYD is the average depth of target. YAC+ is an adjustment similar to plus-minus, and is explained here.

This table is sorted by descending ALEX, and that column alone really explains why Landry is at a disadvantage in generating efficient plays for his offense. His ALEX is minus-2.7, meaning his average target is thrown 2.7 yards short of the first-down marker. The average ALEX for the other 27 receivers is plus-2.9, or 2.9 yards beyond the first-down marker. You can see that slot receivers tend to have the lowest ALEX, though Golden Tate (minus-1.6) is the only other receiver below minus-1.0. Some of the deep threats such as Sammy Watkins (plus-6.4), DeSean Jackson (plus-6.4), and Mike Evans (plus-6.1) are getting their targets more than 6 yards past the marker. Landry's average target is thrown only 6.5 yards beyond the line of scrimmage, the shortest of any wideout studied.

This makes it easier for Landry to rack up receptions, but it's quite difficult to turn those plays into successful gains, or to generate YAC that exceeds expectations. That's why his YAC+ is rather average here at minus-0.1. While you won't find strong correlation between ALEX and DYAR, there is no denying that the location of Landry's targets is a detriment to his success.

That "throw it to him short and see if he can break tackles" style is what makes Landry unique for a No. 1 wideout, but uniqueness does not always transfer into value-added performance. Kyle Juszczyk is one of the league's best pass-receiving fullbacks, but it shouldn't take much effort to find another person willing to block, never carry the ball, and stand 3 yards behind the line of scrimmage so that Joe Flacco can mercifully dump the ball off. Let the 49ers make Juszczyk the highest-paid fullback in NFL history. In fact, the Ravens added Danny Woodhead this offseason, so problem solved.

Landry's 2016 Play Design

We want to believe that teams know the best roles to put their players into based on their skill sets. We know this doesn't always work out. But now on his second full-time head coach in Adam Gase, Landry still played the slot role for the Dolphins last season while Kenny Stills was the flanker and DeVante Parker was the split end.

Thanks to our friends at Sports Info Solutions, we now have play design data for the 2016 season. We can look at the type of routes the receiver ran on his targets. This data was referenced a few times throughout FOA 2017, and if there's enough interest, we could certainly do some formal articles on the site to look at the numbers league wide. For now, we're going to look at Landry in particular. The sample sizes are of course small for the route types, but it's still interesting to see a player's DYAR broken down this way.

Jarvis Landry: 2016 Play Design
Play Design DYAR DVOA Passes Yards PYD YAC
Slant 72 54.0% 14 167 5.4 8.4
Corner 68 176.2% 5 123 25.8 6.0
Dig 64 45.6% 14 178 11.9 4.3
Flat 29 87.1% 4 71 1.3 22.0
Wheel 19 188.0% 1 33 16.0 17.0
Post 19 81.4% 3 39 15.0 5.5
Post Corner 10 51.8% 2 32 15.0 13.0
Fade 8 42.6% 2 25 22.5 6.0
Comeback 7 -4.7% 11 66 7.6 2.3
Quick 7 85.9% 1 14 1.0 13.0
Drag 5 -4.7% 9 110 3.2 15.7
Swing 4 37.6% 1 8 -3.0 11.0
No Video -6 -92.1% 1 0 5.0 -
Hitch & Go -7 -107.0% 1 0 17.0 -
Out & Up -7 -105.9% 1 0 20.0 -
Curl -13 -23.6% 15 69 5.3 3.1
Out -13 -25.3% 13 88 6.0 2.4
Whip -13 -55.2% 4 8 3.0 1.0
Broken Play -14 -47.5% 5 8 8.2 4.0
WR Screen -64 -48.4% 24 97 -2.0 6.5
Totals 174 4.8% 131 1136 6.3 6.6
Source: Sports Info Solutions

(Ed. Note: Note that "broken play" here refers to a scramble drill where the protection and routes break down, not a play with an aborted snap or bad handoff or something along those lines.)

We can spend plenty of time talking about the general ineffectiveness of wide receiver screens at a league-wide level, but that is certainly Landry's most common route with 24 of them. Landry’s minus-64 DYAR on wide receiver screens was the second-worst total in the league in 2016, ahead of only Tavon Austin's screens to nowhere (minus-74 DYAR). Landry did some of his best work on slants (72 DYAR, ranked sixth in NFL), corners (68 DYAR led the NFL), and dig routes (64 DYAR ranked fourth). However, the Dolphins still insisted on using him more on short curls (minus-13 DYAR), out routes (minus-13 DYAR), and of course those unproductive screens.

We were able to look at similar data on Stills and Parker, but only wanted to show the 12 routes where each of the trio had at least one target last season.

2016 Dolphins: Wide Receiver Play Design
Play Design Player DYAR DVOA Passes Yards PYD YAC
Comeback D.Parker 14 78.2% 2 32 12.0 4.0
J.Landry 7 -4.7% 11 66 7.6 2.3
K.Stills -23 -46.9% 9 33 11.3 0.7
Corner J.Landry 68 176.2% 5 123 25.8 6.0
D.Parker 21 40.6% 5 50 23.0 1.0
K.Stills 6 25.2% 2 29 28.5 3.0
Curl K.Stills 25 14.3% 13 119 10.3 3.3
D.Parker -9 -23.0% 12 67 6.9 1.0
J.Landry -13 -23.6% 15 69 5.3 3.1
Dig J.Landry 64 45.6% 14 178 11.9 4.3
D.Parker 39 25.4% 13 120 8.8 2.9
K.Stills -9 -26.7% 8 53 12.0 0.5
Drag D.Parker 20 29.1% 7 79 2.9 16.8
J.Landry 5 -4.7% 9 110 3.2 15.7
K.Stills -1 -17.2% 3 17 1.3 4.3
Flat J.Landry 29 87.1% 4 71 1.3 22.0
D.Parker 5 17.5% 2 18 5.5 17.0
K.Stills -1 -16.2% 2 9 3.0 8.0
Play Design Player DYAR DVOA Passes Yards PYD YAC
Hitch & Go D.Parker 23 270.5% 1 56 30.0 26.0
J.Landry -7 -107.0% 1 0 17.0 -
K.Stills -7 -102.9% 1 0 32.0 -
Out K.Stills 8 -5.5% 16 85 6.8 1.8
D.Parker 6 -5.6% 11 59 8.3 0.4
J.Landry -13 -25.3% 13 88 6.0 2.4
Out & Up K.Stills 14 74.6% 2 24 24.0 0.0
D.Parker -7 -99.2% 1 0 29.0 -
J.Landry -7 -105.9% 1 0 20.0 -
Post J.Landry 19 81.4% 3 39 15.0 5.5
K.Stills 12 12.7% 6 68 26.7 7.5
D.Parker 3 7.5% 2 18 21.5 1.0
Slant J.Landry 72 54.0% 14 167 5.4 8.4
D.Parker 23 16.6% 10 86 9.4 3.0
K.Stills 7 74.6% 1 15 10.0 5.0
WR Screen D.Parker -1 -14.8% 7 31 0.1 4.3
K.Stills -26 -103.8% 4 4 -0.8 3.5
J.Landry -64 -48.4% 24 97 -2.0 6.5
Source: Sports Info Solutions

Again, it's a small sample size and we only have one season of data to go on, but maybe what Landry needs are more corners, digs, and slants than those pesky little bubble screens and short comebacks/curls. So Gase could either "unleash" Landry more, or look Parker in the eye and tell him, "Yo, you got the juice now, man." Either way, something has to change with this passing game if the Dolphins are going to outscore the likes of New England, Pittsburgh, and Oakland. Death by 1,000 Landry paper cuts is not the formula.

Conclusion

In an era where passing is at an all-time high, a record like most catches in a player's first three seasons is almost meaningless, and sure to have an early expiration date. If that becomes a bargaining chip in Landry's contract negotiations, then it's not a good one, especially with the way he has accumulated his stats compared to former college teammate Odell Beckham. Beckham is the type of receiver who deserves a monster second contract based on his overall value and play-making ability.

If the Dolphins are set on Landry's role as the slot receiver, then they would be wise to offer him a deal that's in the $11 million range, or "Keenan Allen money." Paying Landry like a top-flight, do-everything receiver would be a mistake, as would continuing to feature him as the No. 1 receiver in such a limited capacity.

For the Dolphins to take the next step offensively, Parker is the Miami receiver who needs to make the biggest strides, which wouldn't be uncommon to see in a talented player's third season. As long as Landry is far and away the leading receiver in Miami’s offense, it is unlikely that this passing game will rise above mediocrity. If Gase is the head coach we think he can be, then he should realize that a formula of less Landry, more Parker, and more Jay Ajayi runs is the right one for Miami to continue having success in 2017 and beyond.

Posted by: Scott Kacsmar on 27 Jul 2017

8 comments, Last at 31 Jul 2017, 3:22pm by Scott Kacsmar

Comments

1
by BywaterBrat :: Thu, 07/27/2017 - 6:27pm

absolutely love that last chart

2
by Noah Arkadia :: Fri, 07/28/2017 - 11:22am

Yeah, great write up and it matches what I believe is happening with that Miami offense.

3
by Laufy :: Fri, 07/28/2017 - 12:58pm

Awesome analysis - definitely a case where the numbers back up what I've been watching the past 3 years. Love Landry, hate WR screens (looking at you Mack Brown era Texas), and really want to see him utilized in a more versatile way to take advantage of his talents.

4
by johonny :: Fri, 07/28/2017 - 1:03pm

The good news is Parker does look good in camp, and Miami may have a TE (although it's far from certain). An effective TE should allow more options than Landry screen pass that goes nowhere.

5
by Raiderjoe :: Fri, 07/28/2017 - 1:09pm

wr scren is like for baby and crappy quaretrbacks if doing it more than once a game. completely garbage oplay call in my opnion. if gonna resort to that crpa, might as well put fullback in game and run ball from pro set or I-formation

6
by lightsout85 :: Fri, 07/28/2017 - 4:56pm

How many missed/broken tackles (on receiving plays) does FO has for Landry? (over w/e span). Looking at their twitter, PFF as 76 in his 3-year career, had 26 in 2016 (tie 2nd among WR), and 48 since 2015 (2nd to Golden Tate's 53 - and among the top 5 in the tweet-graphic, his rate of MT was also 2nd highest (assuming this includes PO catches & MTs)).

Perhaps this is why PFF & his peers rate him higher - the "talent" is there, even if the production isn't (or rather, isn't by measure of DYAR/DVOA, because his rate of 1st downs by routes-run was pretty good. So even if he wasn't getting bigger plays, he was still picking up 1st downs sufficiently).

I'd love to see a "debate" of sorts between "pro/anti" Landry (/his current & past value).

7
by Aaron Brooks Go... :: Mon, 07/31/2017 - 1:56pm

What's the deal with drag routes? That has to be a huge situational difference, because Landry and Parker have basically identical number of passes, yards, average pass distance, and YAC.


Drag D.Parker 20 29.1% 7 79 2.9 16.8
null J.Landry 5 -4.7% 9 110 3.2 15.7

8
by Scott Kacsmar :: Mon, 07/31/2017 - 3:22pm

Yeah, Parker turned one of his into a long TD on 3rd down, while Landry had several targets on 3rd-and-long.