Week 8 Quick Reads
by Vincent Verhei
"Jameis Winston is part of arguably the most frustrating groups of quarterbacks. His talent is undeniable; he's not a Beathard or a Peterman or something, where everything is terrible. He's also not an Alex Smith-type, a steady, unspectacular game manager. Winston's highs are so high; he can do fantastic things ... but his lows are among the lowest in the league. And at this point, I don't know there's much you can do to fix that."
That's part of what FO writer Bryan Knowles had to say about Tampa Bay Buccaneers quarterback Jameis Winston in Audibles at the Line this weekend. It came on a day when Winston's lows were very low, and worse, all too common. He finished the day with four interceptions and five sacks against Cincinnati, and might well have had more if he had not been benched late in the third quarter for Ryan Fitzpatrick. But even on one of the worst days of his career, we still saw the highs Winston is capable of reaching -- he had four completions of 20 yards or more, including a 60-yard touchdown to DeSean Jackson.
This was not just a one-game anomaly. Winston has now thrown 54 interceptions in 1,692 pass attempts in his career. That's an interception rate of 3.19 percent, the worst such rate of any quarterback with at least a thousand pass attempts in the NFL since he was drafted first overall in 2015. However, he has also completed 1,035 passes for 12,817 yards. That works out to 12.4 yards per completion, and that's the second-best of any passer in that same timeframe -- and since Carson Palmer, the only player ahead of Winston, has now retired, it's fair to say that Winston is the reigning mad bomber of the NFL.
This season, at least, that dichotomy has shown up clearly in Winston's advanced stats too. At Football Outsiders, the first step in all of our analysis is to define each play as a success or a failure. As a reminder, a play is considered a success if it gains 45 percent of needed yards on first down; 60 percent of needed yards on second down; or a first-down conversion on third or fourth down. With this methodology, we can separate Winston's successes from his failures, and compare those results to the successes and failures of the league's other quarterbacks. This will help us answer the question of whether Winston's highs and lows are really as extreme as they appear on the surface.
Winston has a success rate of 49.7 percent on 163 pass plays this season. That might sound low, but it's actually 11th-best of the 35 quarterbacks with at least 100 pass plays. (There are a LOT of checkdowns and dumpoffs in the NFL in 2018.) On his 81 successes, he has a DVOA of 157.3%; on his 82 failures, he has a DVOA of -206.7%. That's a difference of 364.1%. That's not the highest gap between success and failure, this season, but it's close.
| Largest Gap in DVOA on Successful and Failed Pass Plays, 2018 | ||||||
| Player | Plays | Suc% | DVOA | DVOA, Successes |
DVOA, Failures |
Gap |
| 14-R.Fitzpatrick | 153 | 51.6% | 36.0% | 207.3% | -176.0% | 383.3% |
| 10-J.Garoppolo | 106 | 46.2% | -21.2% | 157.8% | -213.3% | 371.0% |
| 3-J.Winston | 163 | 49.7% | -14.9% | 157.3% | -206.7% | 364.1% |
| 3-R.Wilson | 201 | 43.3% | 10.3% | 201.4% | -158.8% | 360.3% |
| 3-C.Beathard | 189 | 42.9% | -25.3% | 164.5% | -192.5% | 357.0% |
| 4-D.Watson | 288 | 45.1% | 8.4% | 191.5% | -162.8% | 354.4% |
| 16-J.Goff | 275 | 51.6% | 30.4% | 194.1% | -156.2% | 350.3% |
| 17-J.Allen | 163 | 31.9% | -63.7% | 165.8% | -179.8% | 345.6% |
| 17-R.Tannehill | 141 | 41.1% | -20.3% | 165.1% | -178.5% | 343.5% |
| 14-S.Darnold | 269 | 39.8% | -20.0% | 180.4% | -162.0% | 342.4% |
| 4-C.Keenum | 315 | 43.8% | -8.7% | 175.2% | -161.1% | 336.3% |
| 15-P.Mahomes | 300 | 53.0% | 46.2% | 193.4% | -140.7% | 334.2% |
| 11-C.Wentz | 248 | 49.6% | 7.1% | 167.9% | -165.5% | 333.4% |
| 8-M.Mariota | 173 | 46.2% | -25.8% | 148.5% | -183.5% | 332.0% |
| 3-J.Rosen | 188 | 40.4% | -29.8% | 162.5% | -168.1% | 330.6% |
| 6-B.Mayfield | 246 | 39.0% | -27.9% | 161.6% | -164.9% | 326.5% |
| 14-A.Dalton | 312 | 48.1% | 3.8% | 164.1% | -162.0% | 326.1% |
| 9-M.Stafford | 270 | 50.0% | 7.8% | 166.3% | -158.6% | 324.9% |
| 17-P.Rivers | 229 | 52.8% | 41.5% | 188.9% | -135.1% | 324.0% |
| 10-E.Manning | 350 | 42.0% | -11.9% | 164.0% | -158.8% | 322.8% |
| 2-M.Ryan | 287 | 53.0% | 27.7% | 166.6% | -155.1% | 321.7% |
| 10-M.Trubisky | 260 | 47.7% | 2.5% | 162.9% | -156.5% | 319.4% |
| 12-T.Brady | 262 | 50.8% | 6.8% | 158.6% | -158.3% | 316.9% |
| 12-A.Rodgers | 306 | 43.5% | 10.2% | 175.3% | -139.7% | 315.0% |
| 12-A.Luck | 358 | 48.9% | 2.4% | 155.3% | -157.4% | 312.7% |
| 4-D.Prescott | 231 | 38.5% | -18.6% | 157.6% | -151.7% | 309.3% |
| 4-D.Carr | 277 | 53.4% | 9.8% | 145.8% | -160.7% | 306.4% |
| 7-B.Roethlisberger | 311 | 47.9% | 10.7% | 158.8% | -145.1% | 303.9% |
| 8-B.Osweiler | 123 | 43.9% | 6.4% | 168.2% | -135.0% | 303.2% |
| 1-C.Newton | 247 | 50.2% | 10.6% | 150.6% | -152.2% | 302.8% |
| 5-B.Bortles | 310 | 44.5% | -10.8% | 147.3% | -146.2% | 293.5% |
| 11-A.Smith | 245 | 44.9% | -5.8% | 144.0% | -149.2% | 293.2% |
| 5-J.Flacco | 359 | 45.1% | 4.7% | 159.8% | -131.2% | 291.0% |
| 8-K.Cousins | 369 | 48.5% | 6.0% | 142.7% | -143.1% | 285.7% |
| 9-D.Brees | 257 | 57.2% | 36.1% | 137.2% | -126.7% | 263.9% |
| Minimum 100 pass plays | ||||||
As it turns out, Winston's DVOA on successes isn't very high at all; it's actually in the bottom ten in this group. That's largely because he has thrown only six touchdowns, a rate of 4.1 percent of his pass attempts that would be the lowest of his career. However, Winston's failures have been devastating, with a worse DVOA than anyone except Jimmy Garoppolo. This is mostly due to his 10 interceptions, tied for the most in the league with Sam Darnold and Case Keenum, even though Winston has only started three games. Garoppolo's problem was sacks -- he took 13 of them in less than three games. There are seven quarterbacks with at least seven starts this year who have taken fewer sacks than that.
Thanks to all those sacks, Garoppolo is one of the two players with a larger gap between success and failure than Winston. The one player with a higher gap than either of them is -- of all people -- Fitzpatrick, Winston's backup/replacement. Fitzpatrick currently leads the league with 15.8 yards per completion, and his touchdown rate of 9.0 percent is just a few decimals behind that of Patrick Mahomes'. As a result, he has a 207.3% DVOA on successful plays, the highest in the league. Meanwhile, his -176.0% DVOA on failures is in the bottom ten; his interception rate of 3.5 percent is better than Winston's, but still significantly higher than average.
The remarkable thing about the Winston/Fitzpatrick pairing is that in recent history, they have been the same player. Remember when we said Winston had the worst interception rate but second-best yards per completion in the last four seasons? Well, Fitzpatrick has the second-worst interception rate, but the third-best yards per completion. Both players are in the bottom five in completion rate too. By accident or by design, the Buccaneers have armed themselves with the NFL's two most prominent low-percentage, turnover-prone, explosive quarterbacks. Fitzpatrick will start for Tampa Bay against Carolina this weekend, but despite his torrid start to the season during Winston's suspension, the results for the Buccaneers will likely be similar no matter who's taking snaps.
At the other end of the spectrum, we have Drew Brees. As usual, Brees has been one of the league's best quarterbacks this year, trailing only Mahomes and Philip Rivers in overall DVOA, but he has been doing it with jabs, not haymakers. Brees' success rate of 57.2 percent is best in the league by a substantial margin; the gap between Brees and second-place Derek Carr (!!!) is as big as the gap between Carr and 11th-place Winston. Brees' -126.7% DVOA on failures is also the best in the league, because he has only one interception, nine sacks, and two fumbles this year. However, Brees' 137.2% DVOA on successes is the worst in the league. Brees is only averaging 10.6 yards per completion, which would be his lowest such figure since 2010. None of his top four receivers is averaging even 12.0 yards per catch. Brees sits at the bottom of one column and the top of the other, so obviously he has the smallest gap between highs and lows in the NFL.
The splits of players like Winston, Fitzpatrick, and Brees are not typical. The correlation between DVOA on successes and DVOA on failures is just -0.024. Just because your highs are high doesn't guarantee that your lows will be low; at the same time, a lack of punching power has nothing to do with an ability to avoid catastrophe.
A few other big-picture notes on this data:
- After Fitzpatrick, Garoppolo, and Wilson, other players with large gaps between ceiling and floor include Russell Wilson, C.J. Beathard, Deshaun Watson, and Jared Goff. Fitzpatrick is the only one of those players over 30, which suggests young quarterbacks might be more prone to erratic performance.
- As noted, Brees has the smallest gap between his highs and lows. He's followed in that category by Kirk Cousins, Joe Flacco, Alex Smith, and Blake Bortles. Flacco and Smith have been two of the NFL's most conservative passers for years now, so their presence here makes sense. Cousins' name is more of a surprise, but he's averaging just 10.5 yards per completion in his first year in Minnesota. Bortles has thrown for just 10 touchdowns, which is limiting his DVOA on successes.
- Passing distance definitely plays a part in all this. Fitzpatrick and Winston are the only players in the league whose average pass has traveled more than 11 yards beyond the line of scrimmage (nobody else is even at double digits), and we would expect deep throwers to have more boom-and-bust results. Carr's average pass has traveled a league-low 6.3 yards downfield; Brees, Bortles, and Cousins are also in the bottom five. Each of those players is on the low end of gap between success and failure. On the other hand, Beathard is second behind Carr in lowest pass distance, and the margin between his highs and lows has been huge. He's the exception, though. The correlation between passing distance and DVOA on failed plays is -0.324 (as passing distance goes up, DVOA on failed plays tends to go down). DVOA on successful plays has an even stronger correlation with pass distance at 0.599. And the correlation between pass distance and the gap between highs and lows is 0.631.
- Speaking of correlations: the correlation between DVOA on failures and overall DVOA (0.561) is stronger than the correlation between DVOA on successes and overall DVOA (0.355). This suggests that a high DVOA (and theoretically a better football team) has more to do with limiting the damage of bad plays than it does with maximizing the impact of big plays. However, the correlation between success rate and DVOA is an overwhelming 0.839. No matter the extremes, as long as you get a lot of good plays and few bad plays, you'll be fine (duh).
| Quarterbacks | |||||||||||
Rk |
Player |
Team |
CP/AT |
Yds |
TD |
INT |
Sacks |
Total DYAR |
Pass DYAR |
Rush DYAR |
Opp |
1. |
Deshaun Watson | HOU | 16/20 |
239 |
5 |
0 |
0 |
172 |
166 |
7 |
MIA |
In the first 40 minutes of this game, Watson only threw for three first downs. (Two of those were touchdowns, but still.) Then the Dolphins scored a touchdown that left Houston clinging to a precarious 21-17 lead. From that point to the end of the game, here is the result of every Watson dropback: 73-yard touchdown to Will Fuller; 49-yard touchdown to DeAndre Hopkins; 19-yard completion to Hopkins; 38-yard DPI on a throw to Fuller; 2-yard touchdown to Hopkins; 4-yard gain to Hopkins for a third-down conversion. Houston might want to consider mixing up its play calling though. Watson only threw two passes on first down all night. One of those was on first-and-20. |
2. |
Patrick Mahomes | KC | 24/34 |
303 |
4 |
1 |
2 |
163 |
175 |
-12 |
DEN |
Mahomes would have been the top quarterback this week if he had skipped the fourth quarter. His last pass of the third was a 23-yard touchdown to Kareem Hunt, but in the fourth he went 4-of-7 for 39 yards with an interception and two sacks. His hot streak came on the last two drives of the first half (not counting a kneeldown) and the first two drives of the second. He threw touchdowns on all four of those drives, going 15-of-18 for 213 yards in the process. |
3. |
Russell Wilson | SEA | 14/17 |
248 |
3 |
0 |
2 |
132 |
129 |
3 |
DET |
Every pass Wilson threw on third down resulted in a completed pass for a conversion, as he went 6-of-6 for 117 yards and a touchdown. (He was also sacked once on third down, and his one pass on fourth down was incomplete at the goal line.) He threw 11 passes that traveled at least 7 yards downfield and every one was completed for a first down, for a total of 238 yards. Meanwhile, each of Wilson's three incompletions were thrown to receivers within 1 yard of the line of scrimmage. Weird. |
4. |
Cam Newton | CAR | 21/29 |
219 |
2 |
0 |
0 |
126 |
115 |
11 |
BAL |
Throwing to the middle of the Baltimore defense, Newton went 4-of-5 for 57 yards. Each of those completions produced a first down, including a 6-yard touchdown to Christian McCaffrey. |
5. |
Derek Carr | OAK | 21/28 |
244 |
3 |
0 |
0 |
122 |
115 |
7 |
IND |
Carr's first pass of the second quarter was incomplete. He then completed 17 passes in a row, a stretch that lasted into the fourth quarter and produced 208 yards and 11 first downs, including three touchdowns. |
6. |
Tom Brady | NE | 29/44 |
324 |
0 |
0 |
2 |
106 |
102 |
4 |
BUF |
7. |
Ben Roethlisberger | PIT | 24/36 |
257 |
2 |
1 |
1 |
97 |
93 |
4 |
CLE |
Roethlisberger threw six passes that traveled 12 to 18 yards downfield. Each was completed for a first down, for a total of 101 yards. |
8. |
Ryan Fitzpatrick | TB | 11/15 |
194 |
2 |
0 |
1 |
96 |
95 |
2 |
CIN |
On throws that traveled at least 12 yards downfield, Fitzpatrick went 6-of-8 for 166 yards. Each of those completions resulted in a first down, including two touchdowns. |
9. |
Mitchell Trubisky | CHI | 16/29 |
220 |
2 |
0 |
2 |
95 |
73 |
22 |
NYJ |
Trubisky and the Bears were good at making plays at their own end of the field, but struggled beyond that. Outside the Chicago 40, Trubisky went 8-of-16 for 59 yards with four first downs and two sacks. |
10. |
Matthew Stafford | DET | 27/40 |
310 |
2 |
1 |
3 |
93 |
89 |
4 |
SEA |
Detroit was playing with a deficit for most of this game, and Seattle's soft zones left plenty of space for Lions running backs to make plays. Throwing to his backs, Stafford went 8-of-10 for 98 yards and six first downs. |
11. |
Carson Wentz | PHI | 21/30 |
286 |
3 |
1 |
4 |
90 |
71 |
19 |
JAX |
It was feast or famine for Wentz in scoring range. Inside the Jacksonville 40, he went 5-of-7 for 92 yards and three touchdowns -- but he also gave up two sacks, a fumble, and an interception. |
12. |
Andrew Luck | IND | 22/31 |
239 |
3 |
0 |
0 |
78 |
78 |
0 |
OAK |
Luck was nearly perfect in the fourth quarter, going 6-of-7 for 95 yards. All six of those completions resulted in first downs, incuding a touchdown. |
Rk |
Player |
Team |
CP/AT |
Yds |
TD |
INT |
Sacks |
Total DYAR |
Pass DYAR |
Rush DYAR |
Opp |
13. |
Andy Dalton | CIN | 21/33 |
280 |
2 |
0 |
2 |
67 |
66 |
0 |
TB |
Dalton didn't do much to help when the Bengals were pinned deep. Inside his own 25, he went 3-of-7 for 25 yards with one sack and no first downs. Those completions included a 9-yard gain on second-and-21 and a 7-yard gain on second-and-15. |
14. |
Aaron Rodgers | GB | 18/30 |
286 |
1 |
0 |
3 |
64 |
64 |
0 |
LAR |
Deep-ball passing: 5-of-6 for 173 yards and a touchdown. |
15. |
Jared Goff | LAR | 19/35 |
295 |
3 |
0 |
5 |
63 |
77 |
-13 |
GB |
Goff had his struggles in the middle of the field, but was very good in scoring range. Inside the Green Bay 40, he went 6-of-7 for 103 yards with three touchdowns and one sack. |
16. |
Alex Smith | WAS | 20/32 |
178 |
1 |
0 |
0 |
62 |
59 |
4 |
NYG |
Smith and Washington did a fine job of avoiding negative plays. Only two of his passes came with more than 10 yards to go for a first down; his average yards to go was just 7.7, second-lowest this week behind Aaron Rodgers' 7.6. With 6 yards or less to go for a first down, Smith went 8-of-14 for 92 yards and seven of his 11 first downs. |
17. |
Sam Darnold | NYJ | 14/29 |
153 |
1 |
0 |
1 |
37 |
30 |
7 |
CHI |
Third-down passing: 3-of-10 for 37 yards and only two conversions. |
18. |
Kirk Cousins | MIN | 31/40 |
359 |
2 |
1 |
4 |
33 |
41 |
-8 |
NO |
The Vikings didn't get the win, but their wide receivers dominated the Saints' defensive backs. Throwing to his wideouts, Cousins went 21-of-25 for 266 yards with three touchdowns and one interception. And these weren't failed completions either -- 15 went for first downs, including two scores. And three other throws resulted in DPIs for 30 total yards. |
19. |
Blake Bortles | JAX | 24/41 |
288 |
1 |
0 |
4 |
15 |
5 |
10 |
PHI |
Red zone passing: 2-of-8 for 14 yards and just one touchdown. Keep in mind that Jacksonville lost this game by six points. |
20. |
Baker Mayfield | CLE | 22/36 |
180 |
2 |
1 |
2 |
10 |
6 |
4 |
PIT |
Seven of Mayfield's ten first downs came with 6 yards or less to go for a first down. Any more than that, he went 16-of-28 for 107 yards with an interception and a sack. |
21. |
Josh Rosen | ARI | 24/40 |
252 |
2 |
1 |
3 |
-2 |
-8 |
6 |
SF |
The Cardinals had a lot of success on screens. On throws to receivers behind the line of scrimmage, Rosen went 8-of-9 for 66 yards and five first downs. |
22. |
C.J. Beathard | SF | 14/28 |
190 |
1 |
0 |
4 |
-4 |
8 |
-13 |
ARI |
In Arizona territory, Beathard went 3-of-9 for 22 yards with two first downs and two sacks. |
Rk |
Player |
Team |
CP/AT |
Yds |
TD |
INT |
Sacks |
Total DYAR |
Pass DYAR |
Rush DYAR |
Opp |
23. |
Drew Brees | NO | 19/23 |
120 |
1 |
1 |
0 |
-8 |
-8 |
0 |
MIN |
Brees only threw two passes that traveled 10 or more yards beyond the line of scrimmage. One was complete for a 20-yard gain; the other was intercepted. It's not just that Brees threw for only 120 yards, but 78 of those yards came after the catch. That's 65 percent of Brees' production that came after the pass was completed, most of any starter this week. He only had 42 air yards on completions. There were four different quarterbacks this week who gained more air yards than that on just one play. |
24. |
Brock Osweiler | MIA | 21/37 |
241 |
1 |
1 |
2 |
-33 |
-33 |
0 |
HOU |
Inside Houston territory, Osweiler went 3-of-10 for 24 yards with one sack and no first downs. |
25. |
Derek Anderson | BUF | 22/39 |
290 |
0 |
1 |
3 |
-42 |
-42 |
0 |
NE |
26. |
Case Keenum | DEN | 23/34 |
266 |
2 |
1 |
5 |
-53 |
-56 |
3 |
KC |
Here's an amazing four-play stretch in the fourth quarter: sack-fumble, interception, sack, 2-yard loss on third-and-2. On third downs, he had one 24-yard touchdown pass; one incompletion; one completion for a loss of 2; three sacks; and two fumbles. (He also had a 4-yard touchdown pass on fourth down.) |
27. |
Eli Manning | NYG | 30/46 |
316 |
1 |
2 |
7 |
-56 |
-56 |
0 |
WAS |
Manning's first pass of the third quarter was a 16-yard gain on third-and-4. From that point to the start of the fourth quarter, he went 2-of-6 for -1 yard (not a typo) with an interception and three sacks. |
28. |
Joe Flacco | BAL | 22/39 |
192 |
1 |
2 |
2 |
-77 |
-85 |
7 |
CAR |
On passes that traveled at least 15 yards downfield, Flacco went 0-for-9 with two interceptions. |
29. |
Jameis Winston | TB | 18/35 |
276 |
1 |
4 |
5 |
-158 |
-162 |
4 |
CIN |
In keeping with the theme of our essay, Winston had a lot of good plays in this game. He finished with 15 successful plays; only Ben Roethlisberger and Patrick Mahomes had more in the first three quarters this week. He had 14 first downs; only Jared Goff and Kirk Cousins had more through three quarters. If you take out everyone's sacks and interceptions, Winston was seventh in DYAR through three quarters. On the other hand, if you look only at sacks and interceptions, he wasn't just the worst quarterback of the week, he had nearly twice as much negative DYAR as anyone else. |
| 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. |
Kenyan Drake | MIA | 12 |
58 |
1 |
2/2 |
37 |
1 |
65 |
37 |
28 |
HOU |
Drake would have made the top five running backs this week without them, but it's opponent adjustments that really shoot him ahead of everyone else. He averaged 4.83 yards per carry, best of any player with at least 10 carries in a game against Houston this year. He was stuffed twice, but half of his carries gained 5 yards or more, and he had a 12-yard touchdown and two other first downs. His two catches were a 9-yard gain on first-and-10 and a 28-yard touchdown. |
2. |
Latavius Murray | MIN | 13 |
56 |
1 |
5/6 |
39 |
0 |
46 |
38 |
8 |
NO |
Murray ran for three first downs against New Orleans and also had a 12-yard gain on first-and-20 while getting stuffed twice. All his pass targets came on first-and-10, and he gained first downs on two of them. |
3. |
James Conner | PIT | 24 |
146 |
2 |
5/6 |
66 |
0 |
42 |
9 |
32 |
CLE |
Conner had six runs of 10 yards or more and one shorter first down, while being stuffed four times. Each of his catches gained at least 12 yards, and four produced first downs. |
4. |
Todd Gurley | LAR | 25 |
114 |
0 |
6/7 |
81 |
1 |
38 |
-2 |
40 |
GB |
Gurley had three runs of 10 yards or more and five first downs, but he was stuffed six times, and converted only one of his six carries with 3 yards or less to go for a first down. Four of his catches produced first downs, including a 30-yard touchdown and a gain of 32. |
5. |
Tarik Cohen | CHI | 5 |
40 |
0 |
1/3 |
70 |
1 |
36 |
14 |
23 |
NYJ |
Each of Cohen's carries gained at least 3 yards and two produced first downs, including a gain of 21. His 70-yard touchdown reception included 65 yards after the catch. |
| 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. |
Latavius Murray | MIN | 13 |
56 |
1 |
5/6 |
39 |
0 |
46 |
38 |
8 |
NO |
2. |
Kenyan Drake | MIA | 12 |
58 |
1 |
2/2 |
37 |
1 |
65 |
37 |
28 |
HOU |
3. |
Marlon Mack | IND | 25 |
132 |
2 |
2/4 |
17 |
0 |
34 |
34 |
0 |
OAK |
Mack only had one run longer than 9 yard, but that run was a 49-yarder. He ran for eight first downs while being stuffed just twice. |
4. |
Lamar Miller | HOU | 18 |
133 |
1 |
0/2 |
0 |
0 |
24 |
34 |
-10 |
MIA |
Miller had runs of 15, 19, and 58 yards, and one shorter first down. He was stuffed three times. |
5. |
Frank Gore | MIA | 12 |
53 |
0 |
1/2 |
5 |
0 |
17 |
29 |
-12 |
HOU |
Again, opponent adjustments are a big part of Gore's ranking, more than doubling his rushing DYAR. He only had two first downs, his longest run gained only 9 yards, and he was stuffed three times, but nine of his carries gained 4 yards or more. |
| 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. |
Carlos Hyde | JAX | 6 |
11 |
0 |
0/2 |
0 |
0 |
-28 |
-15 |
-13 |
PHI |
Hyde failed to run for a first down against Philadelphia, even on second-and-3 and third-and-1. His longest carry gained just 7 yards, and he was stuffed twice. His incomplete targets came on second-and-6 and third-and-2. |
| 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. |
LeSean McCoy | BUF | 12 |
13 |
0 |
6/8 |
82 |
0 |
-11 |
-36 |
24 |
NE |
McCoy's first run was a 12-yard gain on first-and-10. His next 11 runs produced a total of 1 (one) yard and no first downs. He was stuffed five times, including four runs that lost 3 yards or more. |
| Five Best Wide Receivers and Tight Ends by DYAR | ||||||||||
Rk |
Player |
Team |
Rec |
Att |
Yds |
Avg |
TD |
Total DYAR |
Opp |
|
1. |
Marvin Jones | DET | 7 |
10 |
117 |
16.7 |
2 |
74 |
SEA |
Only one of Jones' catches failed to pick up a first down; that was a 5-yard gain on second-and-10. He had touchdowns of 19 and 39 yards, and also gained 56 yards on a DPI. |
2. |
Sammy Watkins | KC | 8 |
9 |
107 |
13.4 |
2 |
71 |
DEN |
Six of Watkins' catches produced first downs, including touchdowns of 10 and 13 yards. He also had a 17-yard DPI. |
3. |
D.J. Moore | CAR | 5 |
6 |
90 |
18.0 |
0 |
66 |
BAL |
Moore's totals include 36 DYAR rushing, 31 DYAR rushing for his two carries for 39 yards. Four of his catches gained at least 13 yards and a first down. |
4. |
Tyler Boyd | CIN | 9 |
10 |
138 |
15.3 |
1 |
55 |
TB |
Boyd's totals include 61 DYAR receiving, -6 DYAR rushing for his one carry for a 2-yard loss. Boyd's first six targets each resulted in a first down, totaling 112 yards. His next target was an incompletion, and his last three targets failed to pick up a first down, though two were still successful plays. |
5. |
Dave Moore | SEA | 4 |
4 |
97 |
24.2 |
1 |
54 |
DET |
It was a big week for receivers names "D.Moore." Moore's four targets for Seattle: 10-yard gain on third-and-7; 15-yard touchdown; 27-yard gain on third-and-7; 45-yard gain on first-and-10. |
| Worst Wide Receiver or Tight End by DYAR | ||||||||||
Rk |
Player |
Team |
Rec |
Att |
Yds |
Avg |
TD |
Total DYAR |
Opp |
|
1. |
Jarvis Landry | CLE | 8 |
12 |
39 |
4.9 |
0 |
-41 |
PIT |
Landry had more catches that lost yardage (three) than gained first downs (two). Only one of his catches gained more than 7 yards, a 21-yarder that made up more than half his yardage on the day. He converted just one of his four third-down targets. |
Comments
70 comments, Last at 31 Oct 2018, 6:25pm
#2 by Yu Narukami // Oct 30, 2018 - 8:13am
TB12 that high is one of the case where eye-test doesn't match analytics. Red area offense was sputtering all night long and some throws were really erratic (Oline play had one of the worse outing in the season, both in running (inexistent, thanks also to no power back available) and passing game).
Defense-adjustment can do a lot.
#22 by Lost Ti-Cats Fan // Oct 30, 2018 - 1:08pm
Which again begs the question of what could they possibly be thinking during the offseason? You have a top tier defence so you only need a mid level offense to be a contender. You're in good shape at running back, so its OL, receivers, and always of course QB that you need to be concerned about.
You decide Taylor's not good enough to deliver you a mid tier offense. Fine. You want to draft a rookie sensation QB who could be your guy in the future. No problem, you can't just think about 2018, even if that D is unlikely to stay this good for long. You want to give that rookie some time to develop. I suppose that makes sense, even if it means your top draft pick isn't going to contribute in 2018. So we'll keep Peterman and start him. Wait, what???
I guess I just partially answered my own question: apparently they thought Peterman was better/more cost effective than Taylor and good enough to let them be competitive given the strength of their D. That only begs the question of how could they have thought that.
Hope springs eternal.
#46 by RobotBoy // Oct 30, 2018 - 3:30pm
You would think that benching Taylor last year, and the result, would have given them all the information they needed not to make that exact same mistake again.
Apparently not.
Arrogance and stupidity makes for an intoxicating, dangerous brew.
My friend the former boxing manager, professional gambler, Deadspin writer, and composer/pianist Charles Farrell put together a recording of looped phone messages from his boxing days that includes threats from mafiosi and pleas from fighters, including the legendary Mitch 'Blood' Green. It's called (wait for it) 'Hope Springs Eternal'. Worth a listen if you can track it down. Here's an interview with Charles that includes excerpts from the recording. http://loveandradio.org/2014/07/sesquipedalian/
#49 by RobotBoy // Oct 30, 2018 - 3:43pm
I definitely think the Pats game-plan coming in didn't involve having Brady throw it 50-odd times against a good secondary. Being without their best o-lineman and best power runner forced the issue.
Teams are still sitting on Gronk, who also is injured, and Gordon is uneven, but Edelman seems all the way back.
#56 by Lost Ti-Cats Fan // Oct 30, 2018 - 5:35pm
Agreed.
Actually, I think the Pats pretty much recycled the same game plan they used earlier against Jacksonville. Low risk, conservative offensive plays. Yes, they threw a lot, but mostly it was low risk tosses, until late in Q3 when they realized the Bills could steal this one with one good play. On defence, stop the opposition run game and wait for their QB to self-implode.
It didn't work against Bortles (which in itself is surprising in light of the rest of the season). It almost didn't work against Anderson until the Q4 interception.
#4 by Will Allen // Oct 30, 2018 - 8:33am
It turns out that the pick 6 Cousins threw was due to Diggs not finishing his route. Presumably that int has a pretty large negative DYAR, even if the return is not considered predictive, and is thus discounted.
Now, Diggs is an upper tier player so Cousins certainly derives more statistical benefit, than harm, over the course of a year by throwing to Diggs. However, given the outsized negative value of ints in DYAR ( and rightly so), I wonder if we really appreciate how DYAR, even if it is much better than other metrics, is not really measuring qb performance, as much as it is capturing the performance of the offense as a whole. How confident can we be that in the course of a year, the stuff that the qb doesn't control balances out? And if we can't have that confidence that a year's sample size is large enough, how long does a career have to be in order to be confident?
I don't have time early this morning to fully flesh this out, but once again I am struck by how difficult a project FO has taken on.
.
#63 by Jerry // Oct 30, 2018 - 6:01pm
As you know, and as Vince reiterates below, FO has always acknowledged that individual numbers are very team-dependent. From the Our New Stats Explained page, with a missing quotation mark:
Football is a game in which nearly every action requires the work of two or more teammates -- in fact, usually 11 teammates all working in unison. Unfortunately, when it comes to individual player ratings, we are still far from the point at which we can determine the value of a player independent from the performance of his teammates. That means that when we say, "In 2014, Marshawn Lynch had a DVOA of 23.1%, what we are really saying is “In 2014, Marshawn Lynch, playing in Darrell Bevell’s offensive system with the Seattle offensive line blocking for him and Russell Wilson selling the keeper when necessary, had a DVOA of 23.1%."
#6 by Aaron Brooks G… // Oct 30, 2018 - 8:40am
Speaking of correlations: the correlation between DVOA on failures and overall DVOA (0.561) is stronger than the correlation between DVOA on successes and overall DVOA (0.355). This suggests that a high DVOA (and theoretically a better football team) has more to do with limiting the damage of bad plays than it does with maximizing the impact of big plays.
I would argue it shows that good teams have better bad plays than bad teams do. This is arguably a subset of "limiting their damage", but I think of that more in terms of having fewer of them or having a defense that can bail them out (a la the 80s Bears or early 2000s Ravens -- teams that were often shitty on offense but could survive it anyway).
That's probably as much a function of overall talent as of QB-specific play. If you can throw a bailout pass and have a receiver make two guys miss and turn 3rd-12 into 3rd-2, everyone comes out looking better.
#23 by ChrisS // Oct 30, 2018 - 1:09pm
For the top QB's (ex Fitzmagic) it looks like you need to be a little better than average on successes and a lot better than average on failures. For the worst QB's it seems to be the opposite, they are a litlle below average on successes and much worse than average on failures.
#55 by dbt // Oct 30, 2018 - 5:35pm
I still think of the mid-2000s Bears defense is the epitome of that trait. Most defenses need to rest. Those guys used to watch Kyle Orton go 3 and out, or Rex Grossman throw a stupid pick, and be happy that they got to go back out there and smash some shitty offense and try to score.
(Granted, all that time contributed to Brown and Harris getting hurt, but damn. So much fun to watch.)
#7 by RoninX // Oct 30, 2018 - 9:21am
Chris Carson must have been pretty punished by the Lions pre-Snacks run defense ratings to not make either RB list. This brings up something I've wondered about, is DYAR adjusted retrospectively (like if the Lions' run defense actually improves down the road) or is final season dyar simply an aggregate of weekly DYAR with adjustments capturing a snapshot of the competition as that were rated at the time?
#8 by ammek // Oct 30, 2018 - 10:11am
Winston has thrown 4 or more picks in a game three times in 48 starts. John Elway managed it in 41 starts; Dave Krieg got there in 33; Dan Fouts had three four-pick games in a single season in 1980. (Brett Favre, perhaps surprisingly, needed 173 starts, although he celebrated in style by having six passes intercepted in a playoff game.)
In the gunslinger era, interceptions weren't so rare that a QB would lose his job for chucking a bunch of them every now and then, provided he could do other stuff. Frustrating as it could be for a fan (and I rooted for the Lynn Dickey Packers), it was a lot more entertaining than watching Alex Smith and Sam Bradford throwing no-hope screens on third-and-long.
See also: Payton, Dickerson, Marcus Allen and their fumbles.
Also, is it me, or are Shady McCoy and Jarvis Landry in the Frame of Futlity (aka, Worst Rusher/Receiver By DYAR) as often as not this season?
#45 by ammek // Oct 30, 2018 - 2:49pm
Sure, but there were already QB 'styles' in the 1980s. At one end we had the low-interception-rate, high-completion-percentage guys like Bernie Kosar, Ken O'Brien and, above all, Joe Montana; at the other, the gunslingers with big yards-per-completion numbers: Dickey, Krieg, Jay Schroeder, and above all Fouts. For a few seasons it wasn't clear which, if either, of the styles would be more productive; gradually the 'west-coast' influence won out and offenses started to become more homogeneous, with a focus on lower-risk passes and on avoiding turnovers. That homogeneity was only intensified by the rules changes later on.
I've argued before on here that the nadir for excitement in the sixteen-game NFL was 1990-93: yards per completion plummeted by a full yard in four seasons, completion percentage rose by more than two percentage points over the same period, interceptions continued to decrease, and the game plodded along at a dreary pace. (Ironically the early 1990s are usually associated with the run-and-shoot, but comparatively few teams ran that system, and the ones that did rarely had exceptional y/c numbers.) In other words, the transition toward the present QBing era had already begun before the rules were changed, and the first tweaks (in 1994) actually, albeit briefly, reversed the trends: completion percentage in 1997-99 was lower than at any time since 1990; the downward trend in yards per completion was paused for a handful of seasons; and the decline in interception rate slowed down, not taking off again until the early 2000s.
#62 by Jerry // Oct 30, 2018 - 5:57pm
The rules started changing with the Mel Blount Rule in 1978. Once DBs could no longer reroute receivers, innovations like the West Coast offense were practical. Obviously, we've seen a lot more offense-friendly rules changes since.
#12 by Mike B. In Va // Oct 30, 2018 - 11:00am
I can't speak to Landry, but Shady looks like crap this year. Admittedly, Buffalo's O-Line is terrible, but Ivory looks much better since he just picks his gap and goes. Shady's looking for the home run every play, it seems.
#17 by Boots Day // Oct 30, 2018 - 11:54am
It's too bad we didn't get an Audibles for last night's game. There was a lot going on that would have been comment-worthy, starting with the Bills coming out in the Wildcat, which seems like a good idea when you have no quarterback.
#39 by BJR // Oct 30, 2018 - 2:17pm
Which basically limits the outcomes to nothing, on every play.
As the poster above suggests, those carries should be given to Ivory, who at least seems willing to run straight ahead for a couple of yards when there is no hole (i.e. every running play). Or even better, find creative ways to use McCoy as a receiver, or get the ball to him in space. My interest piqued when the Bills used assorted trickery on their first drive last night, but needless to say they had nothing else up their sleeves, and ran their historically terrible regular offense for the rest of the night.
#47 by RobotBoy // Oct 30, 2018 - 3:33pm
I also think Shady is suffering from a nagging injury/physical decline or some combination thereof. He can't make something out of nothing, at least between the tackles, anymore. Just not as supernaturally quick in his cuts.
#58 by Lost Ti-Cats Fan // Oct 30, 2018 - 5:42pm
The only Bills game I've seen this season is the Pats game on Monday. So my observations are based solely on that game.
McCoy can't teleport through a line whose only holes are filled with NE defenders crowding the line because they don't need to respect the Bills' passing game. Yes, his dipsy-doodling lost more yards than if he'd plowed straight ahead, but running north-south was going nowhere. He kept looking for opportunities to bounce it outside, but apparently in addition to not being able to get open, his WR's can't block either, and going east-west just resulted in a bad play getting worse.
#10 by Scott P. // Oct 30, 2018 - 10:32am
I am confused by the table. If you are comparing successful plays against successful plays, shouldn't you recalibrate 'average' to be the average successful play? And the same for unsuccessful plays? That would give a DVOA for each centered around approximately 0%.
#16 by Travis // Oct 30, 2018 - 11:50am
Under the current scoring rules, it counts as an incomplete pass.
From the Guide for Statisticians: "Charge a player with a pass attempt when he throws ... an incomplete pass, including when ... the pass is illegally touched by an eligible offensive player returning from out of bounds (and is not thereafter intercepted). If the illegal pass is caught, score as an incomplete pass."
#26 by Joe Pancake // Oct 30, 2018 - 1:13pm
Thanks for the correction. I even went through and counted the incompletions from the play-by-play data to make sure I was getting this right. But I accidentally counted the dumb challenged-completion-that-then-became-a-penalty as an incompletion instead of the illegal touch. D'oh!
#15 by Joe Pancake // Oct 30, 2018 - 11:41am
I wonder if the powers-that-be at FO have considered changing the definition of "success" on 3rd down in 4-down situations. If it's 3rd-and-20 and a team picks up 18 yards knowing they have to go for it on 4th down, I think everybody would call that a successful play.
Maybe in the past such plays have been too infrequent to matter much to the bottom line, but maybe they will start to matter as more coaches get more aggressive. If teams actually start following a more optimal strategy on fourth down (e.g., almost always going for it between the 40s), as seems to be the trend, then all-of-nothing success on third down might stop being a reasonable assumption.
Food for thought.
#35 by RoninX // Oct 30, 2018 - 1:53pm
You'd need to build in some more logic, 4th quarter, down by X, Y time remaining, field position. Seems like a lot of work and supposition to round off some edge cases. In the first quarter getting 80% of needed yards on 3rd down is not a success.
#41 by Lost Ti-Cats Fan // Oct 30, 2018 - 2:18pm
Could be as simple as 80% of the required yards on 3rd down and the team goes for it on 4th.
That covers off getting stuffed on a short 3rd down (didn't achieve 80% of required yards) and covers getting a bunch of yards but not enough in the circumstances (decided not to go for it), both of which still count as fails.
Obviously, that's still not as much of a success as getting a first down, but getting 60% of the required yards on 2nd down isn't as much of a success as getting a first down either.
#25 by Will Allen // Oct 30, 2018 - 1:12pm
As should failure. A throw made into very tight coverage, on 3rd and 16, while trailing by 35-17, with12 minutes left in the game, resulting in a int, is not nearly the failure that doing so while trailing 10-3 in the 2nd quarter.
#28 by Lost Ti-Cats Fan // Oct 30, 2018 - 1:18pm
I had that same thought when Peterman entered the game last night on 3rd and a gazillion and completed his first pass to make it 4th and a chance:
(a) that's a successful play in this situation; and,
(b) it's not an interception, so it's definitely a success in relative terms.
Of course, the defence was also playing soft to allow the big but not quite to the sticks gain, so that mitigates (a) somewhat. However, there were still 11 NE players on the field, so Peterman's ability to avoid all of them with his pass still makes (b) relevant.
Then to prove it was no fluke, he also avoided throwing his second pass to any of the 11 NE players. Unfortunately, that pass didn't convert the fourth down, so was objectively a fail, except to the extent that it allowed Peterman to retreat to the sideline without having to risk scrutiny on where a third pass would go.
#18 by Will Allen // Oct 30, 2018 - 12:03pm
I know the charting project measures dropped ints, and ints which are clearly the result of receiver tips, but I can't remember if those metrics are ever worked into the DYAR for qbs in the offseason. Same with ints where the qb is trailing by a lot at the end of games, and clearly has to force throws, because incompletions are no longer an acceptable outcome, relative to ints.
Now also factor ints due to receivers not running routes correctly, or when a qb gets hit while throwing, following an instant or near instant pass rush (football will never be able to have the equivalent of baseball scorekeeping, since no such official could have such immediate knowledge of each players' assignments). Now imagine two qbs, one with 20 ints at year's end, and one with 10. I could easily imagine the one with 20 ints actually being responsible for pretty close to the same number of ints as the other qb. Now factor how much negative DYAR an extra 10 ints represents. How many years of a qb's career would we need before having confidence that these factors not attributable to qb performance were coming reasonably close to evening out? Can we envision situations where they never have? I think I can.
I swear, the more I watch this game, the more convinced I become that all of us are actually ignorant about stuff that we think we have knowledge of.
#29 by Jose21crisis // Oct 30, 2018 - 1:18pm
There should be an extra point dedicated to all the trades that are prone to happen about now. Demaryius Thomas is going to the Texans for a "Mid" Round pick. The Texans probably hope he can be a bandage over Will Fuller. And Broncos are totally committed to Courtland Sutton now.
#30 by Biebs // Oct 30, 2018 - 1:18pm
Based on quick reads week to week: I have this week as the Bills best passing week at -42 DYAR. Curious if any team in the Football Outsiders database has started with 8 consecutive passing DYARs worse than this.
* Week 1: -175 DYAR (Peterman -160, Allen -15)
* Week 2: -78 DYAR (Allen)
* Week 3: -65 DYAR (Allen)
* Week 4: -200 DYAR (Allen)
* Week 5: -42 DYAR (Allen)
* Week 6: -151 DYAR (Allen -59, Peterman -92)
* Week 7: -160 DYAR (Anderson)
* Week 8: -42 DYAR (Anderson)
#40 by Will Allen // Oct 30, 2018 - 2:17pm
Really weird that the Lions would trade a 5th round pick, for a 30 year old nose tackle last week, with his 7 million cap number next year, or 3.2 million cap hit if they cut him, and then trade away Golden Tate for draft value back this week. Makes me think they are confused.
#52 by RoninX // Oct 30, 2018 - 4:08pm
Harrison is under contract for two more years and Nose Tackles historically age relatively well. Tate's contract is expiring and the Lions have a couple of good young (cheap) receivers. If they knew they weren't going to resign him in the offseason and accept that while the division is still in reach they are very (very) unlikely to win a superbowl this year, then both these moves make a ton of sense.
#53 by Will Allen // Oct 30, 2018 - 4:24pm
Yeah, I'm no expert on Giants' nose tackles. Nose tackles do tend to age better than some other guys, but 30 is still 30. Just mildly surprising.
Next year will be Stafford's 11th season. Man, it seems like they have wasted a quality 1st round qb pick.
#61 by Mountain Time … // Oct 30, 2018 - 5:55pm
I don't remember where I saw the data, but I'm not sure NTs really do age better than any other position. Are we sure this perception isn't skewed thanks to Ted Washington and a few others that played up to their 40s?
#64 by Lost Ti-Cats Fan // Oct 30, 2018 - 6:26pm
I don't have any data to back this up, but potentially it's possible to be an effective NT longer (i.e. older) than most other positions, but still not any more likely that you will have a longer career, as you're still susceptible to being replaced by a younger/larger/cheaper kid on a rookie contract.
#68 by LionInAZ // Oct 30, 2018 - 8:06pm
Yes, it will be Stafford's 11th season, but he's only 30 now. Tate is 30 too. I love me some Golden Tate, and I'm sad to see him go away. But a 3rd round pick in exchange sounds like a great deal. What the Lions need now is a pass rush and a decent TE.
#43 by BJR // Oct 30, 2018 - 2:25pm
Nelson Agholor is ranked second to bottom in DYAR for qualifying receivers this season. He was solid last season, but was terrible prior to that, and has seemingly reverted to type this year. I'm guessing Tate is his direct replacement.
#48 by TomC // Oct 30, 2018 - 3:42pm
Darnold gets a DYAR/DVOA gift this week by playing the Mack-less Bears. He and the rest of the team were brutal, despite basically no pass rush except when the Bears brought multiple blitzers. One should really compare his stat line to Brady's last week and Osweiler's the week before.
Overall, the Jets looked completely overmatched, by a team that really isn't all that good, especially without Mack. Even when the Bears were regularly shooting themselves in the foot, the outcome of the game never seemed in serious doubt.
#51 by Biebs // Oct 30, 2018 - 4:04pm
That's sort of true. But as bad as the Jets offense is (or comically inconsistent), THey were also missing 3 of their top 4 WRs and their 1A RB.
This is not to say the DYAR is skewed, but more to say that the Bears defensive stats are also skewed because the Jets were running Deontae Burnett, Andre Roberts, and Charone Peake as their WRs after Kearse.
#50 by Arkaein // Oct 30, 2018 - 3:52pm
Bit surprised to not see Davante Adams on the top WRs list. No TDs, but 133 yards, and unlike his previous high target volume weeks this was with 5 catches on only 7 targets, against the Rams who FO has as a top-10 D against both the pass and #1 WRs.
#54 by Vincent Verhei // Oct 30, 2018 - 5:03pm
Thanks for posting this so soon after MNF.
I've been working to manage my time better this season. Halfway through, so far, so good. Typos and mistakes are also way down. Glad you appreciate it.
Could be because of opponent adjustments. Not sure if the Bills have a good pass defense but that will definitely impact TB12 numbers.
Brady had 105 DYAR with opponent adjustments, 62 without them. Take away opponent adjustments from everyone, and he falls from sixth to 14th. Even after last night's game, the Bills 10th in completion percentage allowed, third in yards per attempt allowed, and third in net yards per pass play allowed.
I wonder if we really appreciate how DYAR, even if it is much better than other metrics, is not really measuring qb performance, as much as it is capturing the performance of the offense as a whole.
As I like to say, there is really no such thing as an individual stat in football. Every player in the league is a product of his environment to some degree.
Chris Carson must have been pretty punished by the Lions pre-Snacks run defense ratings to not make either RB list. This brings up something I've wondered about, is DYAR adjusted retrospectively (like if the Lions' run defense actually improves down the road) or is final season dyar simply an aggregate of weekly DYAR with adjustments capturing a snapshot of the competition as that were rated at the time?
1) He wouldn't have made either list anyway (long run of just 12, only five first downs, 12 runs for 2 yards or less, two failures to convert on third-and-1), but yes, he was torpedoe'd by opponent adjustments, going from 13 rushing YAR o -19 rushing DYAR.
2) We don't go back and update the tables in the Quick Reads columns, but yes, the DYAR you see on our positional pages retroactively adjusts for opponents as the year goes along. Drew Brees had 298 passing YAR in Quick Reads in Week 1, but that game is now only worth 187 DYAR because we now know the Bucs defense is shamefully bad.
I am confused by the table. If you are comparing successful plays against successful plays, shouldn't you recalibrate 'average' to be the average successful play? And the same for unsuccessful plays? That would give a DVOA for each centered around approximately 0%.
If I had more than 24 hours to produce this I might have done that. It probably would have made the numbers smaller and easier to comprehend. But I don't think it would have made much difference in the results.
I wonder if the powers-that-be at FO have considered changing the definition of "success" on 3rd down in 4-down situations. If it's 3rd-and-20 and a team picks up 18 yards knowing they have to go for it on 4th down, I think everybody would call that a successful play.
As others have said, this might be something we do down the road if fourth-down attempts continue to become more common. For now, I will point out that there have been 81 third-down completions this year that gained at least 10 yards but did not get a first down. None of those plays had negative DYAR; the average was about 4 DYAR.
How much DYAR did Gurley "lose" by not scoring that last touchdown?
About 8. A 17-yard run on third-and-10 was already a pretty big play.
There should be an extra point dedicated to all the trades that are prone to happen about now. Demaryius Thomas is going to the Texans for a "Mid" Round pick. The Texans probably hope he can be a bandage over Will Fuller. And Broncos are totally committed to Courtland Sutton now.
I'm assuming you've seen this by now, but if not:
https://www.footballoutsiders.com/extra-points/2018/nfl-trade-deadline-thread
Bit surprised to not see Davante Adams on the top WRs list. No TDs, but 133 yards, and unlike his previous high target volume weeks this was with 5 catches on only 7 targets, against the Rams who FO has as a top-10 D against both the pass and #1 WRs.
He was seventh. One more big catch or touchdown would probably have put him in the top five. Remember, there are about 150 receivers and tight ends playing every weekend. It takes a very special day to make the top five.