Quick Reads
The best and worst players of the week according to Football Outsiders stats.

Week 1 Quick Reads

by Vincent Verhei

The Kansas City Chiefs defeated the Jacksonville Jaguars on Sunday, and as you could tell from the 40-26 final score, there are many, many differences between the two teams. Beyond history, geography, color schemes, mascot, coaches, and players, though, I want to focus on how each squad distributed the ball in the passing game. Even with Tyreek Hill leaving the game due to injury, the Chiefs tended to force the ball to a small handful of players. Sammy Watkins led the team with 11 targets, while Travis Kelce had eight and Damian Williams had six; nobody else had more than two. The Jaguars spread things around much more evenly; six players had at least four targets, but none had more than seven.

What's the best way to measure each team's target distribution? For more than a century now, economists have been using Gini coefficients (developed by Italian statistician Corrado Gini) to measure the wealth distribution of various nations, among other things. In a nutshell, Gini coefficients measure the difference between actual financial distribution a perfectly even distribution. If every citizen in a given nation had exactly equal net worth, that nation's Gini coefficient would be 0.000; if one citizen had 100% of that nation's worth, the Gini coefficient would be 1.000. No, we are not going to discuss income inequality in this space (or, please, in the comments), but we can use these same principles to analyze each team's approach to the passing game.

Before we begin, I should note that I am a writer who knows how to use Excel. I am neither a mathematician nor an economist. Everything I know about Gini coefficients I learned from this Wikipedia page and this YouTube video. Gini coefficients go beyond cocktail napkin math, and it's entirely possible I'm calculating them incorrectly. However, I have checked the most extreme results with each team's target list, and they make intuitive sense.

Take that Chiefs-Jaguars game, for example. The Chiefs came out of that game with a Gini coefficient of 0.494, highest in the league (not counting the Monday night games). The Jaguars had the lowest Gini coefficient at 0.208. No team concentrated its targets around a small group of players as tightly as Kansas City, while no team spread its targets around more evenly than Jacksonville.

Other teams that used Kansas City's focused approach include Detroit, Chicago, and Carolina. For the Lions, Danny Amendola had 13 targets and Kenny Golladay and T.J. Hockenson had nine each, while nobody else had more than four. The Bears threw 13 passes to Allen Robinson, 10 to Tarik Cohen and seven to Mike Davis, but not more than five to anyone else. The Panthers had a trio of players (Christian McCaffrey, D.J. Moore, and Greg Olsen) with at least nine targets, but nobody else had more than four.

On the other hand, even target circulation appears to be a Florida thing -- the Jaguars were followed in this category by their in-state rivals, the Buccaneers (six players with at least four targets, none more than six) and Dolphins (seven players with at least three targets, none more than seven). The Steelers also spread things around quite a bit: Donte Moncrief had 10 targets and Juju Smith-Schuster had nine, while five of their teammates had four to six targets each.

These are the results from just one weekend, however, and perhaps random noise. A full season's worth of information should give us more meaningful info. The following table shows each team's Gini coefficient in 2018.

NFL Gini Coefficient Target Distributions, 2018
Team Gini
Coefficient
Total
Targets
Leader Targets Share
MIN 0.659 599 19-A.Thielen 154 25.7%
PHI 0.622 600 86-Z.Ertz 157 26.2%
ATL 0.614 613 11-J.Jones 173 28.2%
CHI 0.614 516 12-A.Robinson 102 19.8%
NO 0.611 524 13-M.Thomas 155 29.6%
LAR 0.610 552 17-R.Woods 134 24.3%
PIT 0.605 686 84-A.Brown 170 24.8%
KC 0.597 571 87-T.Kelce 152 26.6%
SEA 0.585 410 89-D.Baldwin 76 18.5%
BUF 0.583 487 11-Z.Jones 104 21.4%
DAL 0.582 521 21-E.Elliott 95 18.2%
CLE 0.574 563 80-J.Landry 150 26.6%
OAK 0.568 535 87-J.Cook 103 19.3%
GB 0.567 623 17-D.Adams 171 27.4%
NE 0.558 564 28-J.White 123 21.8%
CAR 0.558 552 22-C.McCaffrey 124 22.5%
TEN 0.550 432 84-C.Davis 117 27.1%
TB 0.548 627 13-M.Evans 140 22.3%
DET 0.547 571 19-K.Golladay 124 21.7%
HOU 0.541 500 10-D.Hopkins 165 33.0%
IND 0.538 644 13-T.Hilton 126 19.6%
SF 0.534 524 85-G.Kittle 139 26.5%
BAL 0.533 551 15-M.Crabtree 104 18.9%
NYG 0.531 583 13-O.Beckham 129 22.1%
MIA 0.529 441 80-D.Amendola 79 17.9%
ARI 0.528 478 11-L.Fitzgerald 113 23.6%
WAS 0.515 495 86-J.Reed 84 17.0%
LAC 0.510 508 13-K.Allen 137 27.0%
JAX 0.505 527 12-D.Westbrook 102 19.4%
CIN 0.503 539 83-T.Boyd 108 20.0%
NYJ 0.469 506 11-R.Anderson 94 18.6%
DEN 0.414 572 10-E.Sanders 99 17.3%

Minnesota leads the way here -- no team leaned on its top guys more than the Vikings. Adam Thielen and Stefon Diggs had over 150 targets each. Tight end Kyle Rudolph was a distant third with 83, and nobody else had even 60. The Eagles are next in line, led by the 157 targets of Zach Ertz. Alshon Jeffery and Nelson Agholor had more than 90 targets apiece, but nobody else in Philadelphia had 50. Julio Jones led the NFL with 173 targets, nearly double any of his teammates, so it's no surprise to see Atlanta third here. Nobody else in Atlanta had even 100 targets; Calvin Ridley and Mohamed Sanu were both in the 90s, while Austin Hooper just missed with 89.

It's clear to see that injuries and other personnel moves can affect a team's Gini coefficient when looking at the teams at the bottom of the table. Denver had the lowest Gini coefficient in the NFL; Emmanuel Sanders led the team with just 99 targets, making them one of six teams with no 100-target players. However, eight Broncos had 40 targets or more. Things looked a lot different in Denver in the middle of the year. Through Week 8, Sanders had 65 targets and Demaryius Thomas had 56, but nobody else had even 40. Then Thomas was traded and Sanders got hurt and everything changed. Nobody on the Jets had 100 targets either, in part because Robby Anderson, Jermaine Kearse, and Quincy Enunwa each missed a handful of games. Running back Elijah McGuire missed half the season, but still had 31 targets. For Cincinnati, Tyler Boyd had 108 targets, while A.J. Green was second with 77 (in only nine games) and four other players had 48 or more.

The most surprising team in this table is the Houston Texans. DeAndre Hopkins led the NFL with 33% of his team's targets -- nobody else was even at 30% -- yet the Texans had a lower Gini coefficient than most teams. This is not an error on our part, just a bit of a statistical fluke; the targets that did not go to Hopkins were scattered all over the place. Nobody else on the Texans had even 50 targets, but three were in the 40s, two were in the 30s, and four were in the 20s. This is another reflection of the Demaryius Thomas trade, as well as injuries -- in this case, to Will Fuller and Keke Coutee.

The correlation between Gini coefficient and pass offense DVOA last year was .366 -- generally, teams that spread the ball around did not do as well as those who concentrated on smaller groups of players. Correlation does not mean causation, however. Teams that have superstars are going to feed them as often as possible, and those superstars will raise both the Gini coefficient and the DVOA. Teams that lose all their receivers to injury, meanwhile, will have the opposite results.

Quarterbacks
Rk
Player
Team
CP/AT
Yds
TD
INT
Sacks
Total
DYAR
Pass
DYAR
Rush
DYAR
Opp
1.
Dak Prescott DAL
25/32
405
4
0
0
273
265
8
NYG
Yes, Prescott threw four touchdowns, each within New York's 30-yard line ... but he was even better at the other end of the field. He threw 10 passes within the Dallas 40, and eight of them resulted in first downs. The other two were a 6-yard gain on first down and an incompletion on second-and-4. Those nine completions produced a total of 216 yards.
2.
Lamar Jackson BAL
17/20
324
5
0
1
201
204
-3
MIA
Jackson would have made the top 10 in this table based on his first quarter alone: 6-of-6 for 163 yards and two touchdowns. And then his first pass of the second quarter was his third touchdown, a 33-yarder. He's only below Prescott in these tables because Prescott threw almost 50 percent more passes.
3.
Patrick Mahomes KC
25/33
378
3
0
0
168
174
-6
JAX
Mahomes was also a monster in the first quarter: 8-of-9 for 211 yards and two touchdowns. He was much more quiet in the second half, when only two of his 13 dropbacks gained 10 or more yards, none more than 15 yards.
4.
Case Keenum WAS
30/43
380
3
0
1
151
151
0
PHI
Big first quarters were the theme this week, apparently. Keenum's first 15 minutes: 9-of-11 for 132 yards and a touchdown. His first pass of the second quarter was a 69-yard touchdown, but then he too cooled off. As Washington turned a 10-point lead into a 9-point deficit, Keenum went nine straight dropbacks without picking up a first down, padding his stats with 11-yard gains on second-and-25 and third-and-20 in the process.
5.
Derek Carr OAK
22/26
259
1
0
0
139
141
-2
DEN
6.
Tom Brady NE
24/36
341
3
0
1
139
139
0
PIT
In his first game post-Gronkowski, Brady only threw two passes to tight ends, both to somebody identified as "85-R.Izzo" in the play-by-play. One was incomplete; the other was caught for a 3-yard gain on second-and-13. So, way to go, Steelers. You totally shut that dude down.
7.
Carson Wentz PHI
28/39
313
3
0
1
123
119
5
WAS
Wentz's first third-down pass was incomplete. He nailed every third-down throw after that, completing all 12 of them for 197 yards and nine conversions, including seven in a row at one point, with three touchdowns.
8.
Drew Brees NO
32/41
370
2
1
1
105
105
0
HOU
9.
Matthew Stafford DET
27/43
385
3
0
3
96
85
11
ARI
Stafford's DYAR is worse than you might think because he fumbled twice. The Lions recovered one, but the other was lost to Arizona on first down just outside the red zone. That happened in the first quarter and might have gone overlooked in all the excitement at the end of the game, but it's a big reason Detroit ended with a tie and not a win.
10.
Gardner Minshew JAX
22/25
275
2
1
1
83
82
2
KC
Minshew's debut was fantastic, better in relief than most starters this week, but we must point out that his gaudy completion percentage was padded by a lot of dumpoffs. He had eight failed completions in three quarters and change; Ben Roethlisberger and Matt Ryan led the league with 10 failed completions each this week. Minshew completed all five of his third-down throws, but only two resulted in first downs.
11.
Deshaun Watson HOU
20/30
268
3
1
6
71
51
21
NO
12.
Philip Rivers LAC
25/34
333
3
1
4
46
40
5
IND
On deep passes, Rivers went 6-of-7 for 129 yards and a touchdown.
Rk
Player
Team
CP/AT
Yds
TD
INT
Sacks
Total
DYAR
Pass
DYAR
Rush
DYAR
Opp
13.
Marcus Mariota TEN
14/24
248
3
0
4
37
29
8
CLE
Mariota's three touchdowns came on three consecutive throws over the third and fourth quarters, producing a total of 89 DYAR between them. His other 25 dropbacks were pretty bad. Obviously, if you take away any quarterback's best plays, his numbers are going to look worse, but it's kind of stunning how useless Mariota was outside his explosion. His average completion gained a league-high 12.4 yards after the catch.
14.
Nick Foles JAX
5/8
75
1
0
0
33
33
0
KC
Foles' career in Jacksonville lasted less than 10 minutes of game time before he broke his clavicle. He'll be back in a few months, assuming Gardner Minshew hasn't Brady-Bledsoed him by then. We're a long way from that, but "lost his job to a sixth-round rookie with a moustache" would be a fitting end to the bizarre tale of Foles' career.
15.
Eli Manning NYG
31/44
306
1
0
1
31
30
1
DAL
Manning played well over most of the field, but struggled in the red zone: 4-of-6 for 22 yards, just one touchdown, one sack-fumble, recovered by the Cowboys.
16.
Joe Flacco DEN
21/31
268
1
0
3
29
33
-5
OAK
17.
Russell Wilson SEA
14/20
196
2
0
4
21
21
0
CIN
Here's a perfect summary of how hot-and-cold Wilson's day was: seven of his eight first downs (including both touchdowns) came in the second and fourth quarters. In the first and third quarters, he went 6-of-7 for 63 yards but just the one first down, while also being sacked three times. In the second and fourth, he went 8-of-13 for 133 yards with just one sack, plus a 4-yard DPI.
18.
Kirk Cousins MIN
8/10
98
1
0
1
19
30
-11
ATL
According to Over The Cap, Cousins is being paid $29 million this year in salary and bonuses. That's an average of $1.8 million over 16 game checks. That means Cousins earned more than $150,000 for each of his 12 dropbacks (ten passes, one sack, one DPI) in this game. The Vikings, of course, will be perfectly happy writing those checks as long as they win by multiple touchdowns every week.
19.
Jacoby Brissett IND
21/27
190
2
0
2
18
15
3
LAC
Brissett completed seven of eight passes thrown to his right. That sounds good, but those seven completions gained only 46 yards and one first down.
20.
Andy Dalton CIN
35/51
418
2
0
5
10
10
0
SEA
Dalton was best throwing down the middle against Seattle's defense, going 9-of-12 for 142 yards and a touchdown, with a 13th throw resulting in a 15-yard DPI. Where have you gone Earl Thomas? (Oh. Right. Baltimore.)
21.
Jimmy Garoppolo SF
18/27
166
1
1
1
-1
-1
0
TB
Red zone passing: 2-of-4 for 8 yards, no touchdowns, with a sack that gave those 8 yards back. A total of five plays that went absolutely nowhere.
22.
Aaron Rodgers GB
18/30
203
1
0
5
-6
-6
0
CHI
Rodgers had 13 dropbacks between the 40-yard lines, and only one of those plays resulted in a first down. All told, he went 3-of-11 for 21 yards with two sacks in the middle of the field.
Rk
Player
Team
CP/AT
Yds
TD
INT
Sacks
Total
DYAR
Pass
DYAR
Rush
DYAR
Opp
23.
Ryan Fitzpatrick MIA
14/29
185
1
1
2
-11
-16
5
BAL
A very bad day throwing to his right, where Fitzpatrick went 2-of-8 for 9 yards with an interception.
24.
Matt Ryan ATL
33/46
304
2
2
4
-32
-39
7
MIN
Ryan's splits were like Wilson's but even more extreme. In the first and third quarters, he went 11-of-19 for 84 yards (plus a 26-yard DPI) with two sacks and two interceptions, finishing last in DYAR. In the second and fourth quarters, he went 22-of-27 for 220 yards (plus a 9-yard DPI) with two sacks and two touchdowns, finishing third in DYAR behind Dak Prescott and Matthew Stafford.
25.
Ben Roethlisberger PIT
27/47
276
0
1
1
-46
-48
2
NE
In the first 59 minutes of Sunday night's loss to New England, Roethlisberger threw for exactly 1 yard on New England's half of the field. "Yard." Singular. All told, he went 4-of-17 for 36 yards with one sack and one interception after crossing the 50.
26.
Jared Goff LAR
23/39
186
1
1
1
-57
-55
-2
CAR
Goff did not complete a deep pass against Carolina, going 0-for-5 with an interception. He did gain 16 yards on a deep-ball DPI to Robert Woods.
27.
Sam Darnold NYJ
28/41
187
1
0
4
-81
-81
0
BUF
Darnold did not complete a deep pass either, going 0-for-4. His deepest completion was caught 12 yards downfield by Robby Anderson, who promptly ran backwards 4 yards, then fumbled the ball away. Darnold's average pass traveled 5.4 yards downfield, and his average completion was caught 2.5 yards downfield, both the worst marks for any starter this week.
28.
Josh Allen BUF
24/37
254
1
2
1
-81
-91
10
NYJ
Allen did not have a single dropback in the red zone. In the front zone -- the area between New York's 20- and 40-yard lines -- he went 4-of-9 for 64 yards with one touchdown, one sack-fumble, and one fumble on an aborted play.
29.
Cam Newton CAR
25/37
239
0
1
3
-83
-57
-26
LAR
Newton did not have a dropback in the red zone either. Throwing down the middle, he went 4-of-6 for all of 17 yards with no first downs and one interception.
30.
Kyler Murray ARI
29/54
308
2
1
5
-88
-85
-3
DET
By DYAR, Murray was the worst quarterback in the first quarter, third-worst in the second, and second-worst in the third ... but then the best in the fourth quarter and overtime. In those 25 minutes, he went 20-of-29 for 238 yards with two touchdowns and a sack. Only two of his dropbacks in this game came from under center. He was sacked on the first one and threw an incompletion on third-and-4 on the other.
31.
Mitchell Trubisky CHI
26/45
228
0
1
5
-93
-95
2
GB
On Green Bay's side of the 50, Trubisky went 6-of-14 for 29 yards with only one first down to go with one interception and one sack.
32.
Baker Mayfield CLE
25/38
285
1
3
5
-101
-101
0
TEN
Through three quarters, Mayfield was comfortably among the top 10 passers of the week, but he was far and away the worst passer in the fourth quarter/OT. In the last 15 minutes, he completed six of his nine official pass attempts: three to his own team (for 21 total yards) and three to Tennessee. He also gained 16 yards on a DPI but lost 16 yards on a pair of sacks.
33.
Jameis Winston TB
21/36
194
1
3
3
-222
-227
5
SF
It's not just that Winston threw three interceptions, it's where he threw them. At or inside the Tampa Bay 25 -- you know, where most drives begin -- he went 3-of-7 for 22 yards (plus a 16-yard DPI) with a fumbled snap and ALL THREE interceptions, including two pick-sixes. Winston put together the following four-play stretch just before and just after halftime: completion to O.J. Howard, who fumbled the ball away (not Winston's fault, but still); interception on another pass to Howard; fumbled snap; pick-six.

 

Five Best Running Backs by DYAR (Total)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
Opp
1.
Christian McCaffrey CAR
19
128
2
10/11
81
0
85
61
25
LAR
Five runs for 10-plus yards with three shorter first downs, including two red zone scores. He was stuffed for no gain or a loss just twice, and one of those was on third-and-12 so it barely mattered. Six of his 10 catches were successful (a good rate for a running back), including three first downs.
2.
Austin Ekeler LAC
12
58
1
5/8
96
2
79
19
60
IND
Only two stuffs, with four first downs, including gains of 19 and 13 yards. In addition to his 5-96-2 receiving statline, he also drew a 3-yard DPI to convert a first-and-15.
3.
Derrick Henry TEN
19
86
1
1/2
75
1
54
24
30
CLE
Four first downs on the ground, with gains of 18 and 16 yards, with only two stuffs. His 75-yard touchdown reception included 80 yards after the catch.
4.
Marlon Mack IND
25
174
1
0/0
0
0
54
54
0
LAC
Eleven big first downs on the ground, four more than anyone else this week, including a 63-yard touchdown and gains of 25 and 10 yards, with only two stuffs.
5.
Mark Ingram BAL
14
107
2
0/0
0
0
47
47
0
MIA
Each of Ingram's carries gained at least a yard, and his shortest runs both went for touchdowns or they would have gained more. He had two other first downs, the longest a 49-yard gain.

 

Five Best Running Backs by DYAR (Rushing)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
Opp
1.
Christian McCaffrey CAR
19
128
2
10/11
81
0
85
61
25
LAR
2.
Marlon Mack IND
25
174
1
0/0
0
0
54
54
0
LAC
3.
Mark Ingram BAL
14
107
2
0/0
0
0
47
47
0
MIA
4.
Carlos Hyde HOU
10
83
0
1/1
2
0
30
34
-4
NO
5.
Devin Singletary BUF
4
70
0
5/6
28
0
32
33
-2
NYJ
Usually when you see a guy gain 70 yards on four carries, you'd assume he had a 50- or 60-yarder in there, but no -- EACH of Singletary's carries gained at least 12 yards and a first down. Only Christian McCaffrey has more 12-plus-yard runs than Singletary (not counting Monday night), and it took McCaffrey 14 more total carries to do it.

 

Worst Running Back by DYAR (Total)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
Opp
1.
Miles Sanders PHI
11
25
0
1/2
2
0
-47
-32
-15
WAS
19 yards and a first down on one carry, 6 yards (and six stuffs!) on his other 10 runs. He lost yards on first- and second-and-goal from the 3, as well as second-and-5. His two targets were a 2-yard gain on first-and-10 and an incompletion on second-and-1.

 

Worst Running Back by DYAR (Rushing)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
Opp
1.
Miles Sanders PHI
11
25
0
1/2
2
0
-47
-32
-15
WAS

 

Five Best Wide Receivers and Tight Ends by DYAR
Rk
Player
Team
Rec
Att
Yds
Avg
TD
Total
DYAR
Opp
1.
Sammy Watkins KC
9
11
198
22.0
3
89
JAX
Watkins had more DYAR in the first quarter than any other receiver had in an entire game this week. His first five targets: 68-yard touchdown on third-and-3; 21-yard gain on first-and-10; 13-yard gain on first-and-15; 49-yard touchdown; 22-yard gain on second-and-7. His next six targets produced only 25 total yards, but two first downs, including another score.
2.
DJ Chark JAX
4
4
146
36.5
1
65
KC
His four targets: 7-yard gain on second-and-11; 35-yard touchdown on third-and-8; 69- and 35-yard gains on first-and-10.
3.
Michael Gallup DAL
7
7
158
22.6
0
64
NYG
Five of Gallup's catches gained at least 12 yards and a first down, the longest a 62-yarder. The others were both 6-yard gains on first-and-10.
4.
Marquise Brown BAL
4
5
147
36.8
2
59
MIA
Brown's first quarter was almost as explosive as Watkins': 47- and 83-yard touchdowns, plus a 14-yard gain on second-and-5. He didn't pick up another first down the rest of the game, but who cares?
5.
Phillip Dorsett NE
4
4
95
23.8
2
53
PIT
Four targets: 8-yard gain on third-and-5; 25-yard touchdown; 4-yard gain on first-and-10; 58-yard touchdown on third-and-10.

 

Worst Wide Receiver or Tight End by DYAR
Rk
Player
Team
Rec
Att
Yds
Avg
TD
Total
DYAR
Opp
1.
Donte Moncrief PIT
3
10
7
2.3
0
-75
NE
Moncrief's ten targets resulted in a total of (checks notes) zero first downs. In fact, he had zero successful plays. His three catches: a pair of 3-yard gains (on first-and-10 and second-and-18) and a 1-yard gain on third-and-10. He was targeted on incomplete passes on third- and fourth-and-1.

Comments

43 comments, Last at 11 Sep 2019, 2:19pm

1 A couple of typos:"Adam…

A couple of typos:

"Adam Thielen and Stefon Diggs had over 150 targets each."

Also, I assume Matt Ryan's comment was supposed to say:  "finishing third in DYAR behind Dak Prescott and Matthew Stafford."

2 Nifty Gini analysis. Well…

Nifty Gini analysis. Well done.

You could do a similar analysis of pass DYAR concentration among plays for QBs, within a game or for a season. A high Gini would indicate explosive or explosively bad play. A low one would be for consistent grinding, death by a thousand cuts QBs. Your fine Mariota observation would be quantified.

7 The problem with Gini is it…

The problem with Gini is it's just a distribution number. There's no indication what the platonic ideal value is. A population where all values are zero also has a totally even distribution.

It would be interesting to track, although I suspect you'll find dissimilar teams with similar Gini coefficients for disparate reasons. Embarrassments of wealth look a lot like the distribution of plain old embarrassments.

10 Poor Miles Sanders. He had a…

Poor Miles Sanders. He had a touchdown run called back on a very bad holding call by the officials. He'll bounce back. Looked good in pass protection too.

11 I'm really surprised DeSean…

I'm really surprised DeSean Jackson's not in the top 5, considering DYAR's a counting stat. Yeah, he had 3 negative plays (2 incompletions and a 3 yard loss) and probably 1 neutral play (5 yard gain on 3rd and 10) but the remaining 6 targets were 11-yard gain on 2nd and 10, 9 yard gain on 3rd and 3, 51 yard TD on 3rd and 10, 53 yard TD on 3rd and 10, 19 yard gain on 3rd and 9, and 9 yard gain on 3rd and 4. That collection seems objectively better (and higher volume) than #4 and #5 in that list, so it's kindof surprising that the negative plays would be enough to outweigh it.

14 Field Position

"1 neutral play (5 yard gain on 3rd and 10)" - in the DYAR formula this is a failed completion and results in -DYAR;

Connected Rant:

This kind of play to me is indicative of one of the most difficult things in watching and commenting on football right now - what is field position really worth. In some games I have the subjective perception that yards basically don't matter because the distance between your own 15 and the opponents 35 simply represents a clock requirement and an opportunity for the offense to make a bad mistake (the two NE/KC games last year had this perception) - in other games it's clear that recovering some field position makes a real difference to win% as yards & first downs are hard to get (the super bowl was like this, Jets-Bills was like this on Sunday). It's made worse because I think how a game is being called with respect to PI and other week-to-week variance mean as much to this as does opponent adjustment. It makes it hard to know what your watching in the first half of a game let alone when trying to objectively evaluate a players performance, though the opponent adjustments that come later in the year definitely help.

16 Field position matters

In reply to by sbond101

A team like KC moves the chains almost 80% of the time, by an average of around 18 yards each time.

To stand your formulation on its head: stops against an offense like KC's are hard to get. An extra 18 yards of field position to defend gives you another chance to stop them, which matters a lot when stops are so hard to get.

Five yards of field position isn't much. Think of it as being worth an extra set of downs about a third of the time. You'd much rather keep your own drive going. But it's not nothing.

21 using win probabilities to find this

In reply to by sbond101

If someone wanted to, they could calculate the value of any third down play that doesn't result in a first down using the old win probability calculator that's on another site somewhere (google it i guess). the thing is that it would vary a great deal based on field position, so i doubt there's a rule of thumb. picking up 5 yards on 3rd and 10 when you start at your own 1 yard line is huge. picking up 5 yards near midfield when you have a good punter is close to irrelevant. anyway, not that any of us are going to sit down and do it, but it is an answerable question, and the tools for answering it do exist and are free to use

28 Failed completions aren't…

In reply to by sbond101

Failed completions aren't necessarily negative DYAR. It depends on what the league average performance on 3rd and 10 is. In 2018, that was around 6 yards or so, so that means replacement level should be around 5 yards or so. Meaning it should be a pretty close to zero DYAR play.

12 I realize it's National Jump…

I realize it's National Jump to Conclusions week, but I feel pretty safe saying Case Keenum > Joe Flacco.

42 As a rams fan I never…

As a rams fan I never understood all the hate directed at Case Keenum. Sure, he's at the very low end of reasonable starters. He's also paid that way. His flaw is that his arm strength really sucks, but there's nothing else too bad about him. Okay, that's sort of a big deal, but if you surrounded him with talent, you'd have success. If you don't surround him with talent, then you won't have success. It's not like he's some boat anchor around the neck of your offense that will absolutely ruin a passing attack like KC.

13 Just from look-ins from Red…

Just from look-ins from Red Zone I had the impression that Dalton wasn't as efficient as you would normally think when you see 400+ passing yards, but even then it must have been the 2 fumbles that drove his DYAR down toward zero.

If Seattle is below-average on defense this year he could end up with negative DYAR for the game. I wonder how common it is for 400 yard passing games to result in negative DYAR? Maybe it's not that uncommon when you get into games with 50+ pass attempts and multiple interceptions or fumbles.

22 Bortles had one last year,…

Bortles had one last year, against the Chiefs.
430 yards and 1 TD, but 5 sacks, 4 INTs, and 1 lost fumble.
https://www.pro-football-reference.com/boxscores/201810070kan.htm
https://www.footballoutsiders.com/quick-reads/2018/week-5-quick-reads (the formatting is broken on this page, btw)
-136 DYAR!

Roethlisberger had one in 2013.
406 yards, but 2 INTs and 3 fumbles (2 lost).
https://www.footballoutsiders.com/quick-reads/2013/week-3-quick-reads
https://www.pro-football-reference.com/boxscores/201309220pit.htm
-11 DYAR, unadjusted.

 

2008 had a potential candidate, Warner in week 4.
472 yards and 2 TDs, but 5 sacks, 3 INTs, and 4 fumbles (3 lost).
https://www.pro-football-reference.com/boxscores/200809280nyj.htm

In 2000, Frerotte may have had one.
462 yards and 5 TDs, but 3 sacks, 4 INTs, and 1 lost fumble against the 1-15 Leaf Chargers.
https://www.pro-football-reference.com/boxscores/200011190den.htm

1996 had two candidates.
Brunell's 425 yards, with 2 sacks, 5 INTs, and 1 fumble (0 lost).
https://www.pro-football-reference.com/boxscores/199610200ram.htm

and Esiason's 522 yards and 3 TDs, but 2 sacks, 4 INTs, an 2 fumbles (0 lost).
https://www.pro-football-reference.com/boxscores/199611100was.htm

34 Nice finds. That…

Nice finds. That Roethlisberger game would almost certainly have been adjusted down given how the 2013 Bears defense turned out.

That Brunell game at the Rams is wild - they moved the ball up and down the field - 36 first downs!! - never punted, but he threw those five interceptions (including a pick-six), they had a bonus fumble on a kick return that set up a Rams FG, and had both drives at the end of each half stall out in Rams territory - it even looks like they got down to the 5 yard line on the last play of the game. Lost 17-14.

17 Defensive adjustments

These figures have no defensive adjustments, right? I have a feeling that once the defenses are accounted for, Lamar Jackson's day is going to be seriously downgraded, and Mahomes may well end up on top for Week One.

18 no RGIII? What's the cut-off?

Is there a minimum number of pass attempts or plays to be included? I know it was one quarter of work, but RGIII was 6/6 with a TD and rushed 4 times. Makes sense not to include someone who comes in to throw one pass or a trick play, but I figured backups who take over still made the list.

19 Was expecting TJ Hockenson…

Was expecting TJ Hockenson to show up in the list, but it looks like there some very good receiving games this week. Still like 6/9 for 131 yards (shortest completion was 7 yards) and a TD, something positive for a Lions fan to focus on in light of the 4th qtr collapse.

27 PFF gave very poor grades to…

PFF gave very poor grades to Elflein and Bradbury. I didn't watch the game so can't comment further, but would be keen to know how the team managed to have such success running the ball if the interior of their line was dominated. 

29 Their pass blocking was bad,…

Their pass blocking was bad, and between the tackles wasn't much, but those outside zone plays that hit big were competently executed, and Cook can turn competent blocking into explosive gains. Vikings were lucky to have him available in the 2nd round. Hope he stays healthy.

I think Kubiak was brought in to make that run game explosiveness enough of a constant threat to make pass protection, by guys, especially in the middle, who are a little physically overmatched, somewhat easier. If Cook can continue to hit a couple big runs a game, maybe Cousins will get more time. Here's hoping....

33 Thanks. FWIW, whilst I do…

Thanks. FWIW, whilst I do think they do valuable work overall, I certainly don't treat PFF grades as gospel. As you indicate, it's very possible for guys to do a mostly-competent job within a scheme, whilst appearing pedestrian on tape.

30 OL geades

PFF loved Brett Jones, gave him top marks in the preseason... and of course that was who the Vikings cut to bring in Doctson.  Reportedly they now want him back and that was their plan all along (didnt have him clean out his locker), so presumably they cut someone else assuming he comes back.

I can't think of any reason to risk that apart from not guaranteeing his contract or something like that.

edit - And they have Brett Jones signed again, and cut TE Brandon Dillon, who did play some (7 snaps) sunday.   They hope to sign him to the practice squad, enough so that they already released the K they signed when Vedvik was claimed by the Jets (they had hoped to put Vedvik on their practice squad).      Of course if they actually want Vedvik, he was released by the Jets....

24 re Gini for full season

I'm thinking averaging the game-by-game Gini numbers may give a better indication of an offense's tendencies than calculating based on full year results.

If I've understood your approach (and I may not have fully), then calculating Gini on full year results will tell you as much about how much turnover there was in the receiving corp as anything else. Or if there's a QB change and QBs have differing favourite targets, that'll tilt the numbers towards a low Gini. Neither necessarily tells you anything about how the team approaches (or at least executes) it's game plans.

Game-by-game Gini, on the other hand, tells you how frequently the game plan revolves around 1 or 2 receivers, even if who those receivers are changes from one game to another.

Just a thought.

25 PS

PS Good work on thinking about and reporting on this way of analyzing team tendencies!

26 Glad that DYAR backed up my…

Glad that DYAR backed up my impression of watching Baker Mayfield on Sunday; that he played just fine in the first half (whilst the offense were being completely undone by penalties), but hurt his arm taking a sack at some stage in the third quarter, after which his play deteriorated horribly as he was also forced to chase the game. Apparently an X-ray revealed only deep bruising, so hopefully it is nothing too much to be concerned about going forward.

Also, Dak Prescott's day is going to look a lot less stellar by season's end. That Giants pass defense is truly horrific.

43 WRT Dak, maybe not. I…

WRT Dak, maybe not. I remember seeing two or three touchdown passes where the defense just forgot to cover people. No idea how often this happened on regular passes which weren't shown on replay from different angles. That is something that might be worked out as early as next week. It's a lot easier to play decent defense when you're not shooting yourself in the foot.

37 "The Vikings, of course,…

"The Vikings, of course, will be perfectly happy writing those checks as long as they win by multiple touchdowns every week."

There's no way the Vikings continue to win games by eschewing the forward pass. Right?

38 It's definitely small sample…

It's definitely small sample size theater but Devin Singletary looked really, really good on the field both on Sunday and in the preseason. His lack of size and speed probably puts a ceiling on his production but the dude seems to make up for it with excellent body control, vision and general understanding of how to use his blockers.

The Bills have 3 starter(ish) players that were absolutely loathed by FO metrics with Singletary (70% of snaps on Sunday), Allen (100%, obvs) and Zay Jones (42%). None of them is a clear bust so far - Jones is probably the closest but he was productive enough last season to justify a depth position.

39 "In the front zone -- the…

"In the front zone -- the area between New York's 20- and 40-yard line." Is that the term used for that region of the field? I remember Gregg Easterbrook called it the Maroon Zone, which I think sounds a lot better.

It's always fun to see how random of a collection of players you see at the top of the QB charts. This week the top five includes four players who were mediocre or significantly worse in DVOA and DYAR last year.

40 It's looking like the FO…

It's looking like the FO Almanac was correct in thinking Gase would ruin Darnold. I didn't watch the game in real time, but saw most of the highlights, and they definitely didn't throw deep enough, although Darnold threw too short every time Anderson got open deep. And now Gase is throwing the receivers under the bus. I'm hoping it's a long year, and they send Gase packing at the end of it.

41 Q&A

I'm really surprised DeSean Jackson's not in the top 5, considering DYAR's a counting stat.

Was expecting TJ Hockenson to show up in the list, but it looks like there some very good receiving games this week.

Jackson finished eighth. As you noted, he only had four negative targets … but that's still more than anyone in the top five. Chark and Gallup had zero. Dorsett only had one.

Hockenson was 17th. Only four of his catches produced first downs.

Is there a minimum number of pass attempts or plays to be included? I know it was one quarter of work, but RGIII was 6/6 with a TD and rushed 4 times.

As Aaron noted, quarterbacks need eight passes to qualify. That keeps Griffin, Daniel Jones, Josh Rosen, and Matt Moore out of the tables.

I'm thinking averaging the game-by-game Gini numbers may give a better indication of an offense's tendencies than calculating based on full year results.

That's a good idea, and you're probably right, but actually measuring that would have multiplied my workload 16-fold, so that wasn't an option.

"In the front zone -- the area between New York's 20- and 40-yard line." Is that the term used for that region of the field?

That's what we've called it here since Day 1. https://www.footballoutsiders.com/info/glossary