Writers of Pro Football Prospectus 2008

20 Oct 2014

Week 7 Quick Reads

by Vincent Verhei

Jerick McKinnon's collegiate career was a bit of a mess, as he spent time a cornerback, running back, and quarterback at Georgia Southern. That made it hard to project McKinnon at the NFL level but there was no doubt he was a very special athlete. He shined at the combine, leading all running backs in the bench press (with six more reps than any other runner) and finishing second at the position in 40-yard dash, vertical leap, and broad jump. Though he stands just 5-foot-9, he weighs 209 pounds with a 40 time of 4.41 seconds. That works out to a Speed Score of 110.5, best of any drafted runner this year. Though he lacked experience and refinement, his athleticism was sure to garner interest from some NFL squad. Here's part of what Matt Waldman had to say about McKinnon in his predraft analysis of rooking running backs:

What I found most encouraging is that [McKinnon] ran with his eyes -- he could set up blockers and make defenders miss creases when asked to do so. There are still concerns about his footwork because he often loses balance when he attempts a dramatic cut or bend. McKinnon has everything a coach wants from a running back in terms of physical dimensions and talents. The question is if the back can develop a conceptual game commensurate with the NFL. ... Scouts love the idea of players like McKinnon. He offers enough to an organization as a contributor on special teams that he's the classic "safe" late-round pick even if he never pans out as a running back.

Though it appeared at the time that they had no immediate need for a running back, the Vikings took McKinnon in the third round in case anything bad happened to Adrian Peterson. As it turned out, Adrian Peterson happened to Adrian Peterson, and McKinnon became a key contributor far sooner than anyone in Minnesota had anticipated. Once Peterson was suspended, the Vikings tried third-year pro Matt Asiata as their starter, but he has gradually been losing time to McKinnon. That, if you're a Vikings fan, is for the best. McKinnon is already paying dividends for Minnesota, and he was the most valuable rusher of the week in the NFL.

That's partly because he was playing the Buffalo Bills, who entered the week ranked second in the league in run defense, giving McKinnon a big boost in opponent adjustments. Against a defense like that, success is going to come in small doses. McKinnon had only four first downs against Buffalo, and just two other successful runs. However, only two of his carries failed to gain yardage -- a read option where nobody blocked Jerry Hughes, and a dive play that ran right into the heart of the Buffalo defense.

Meanwhile, when he got a chance to explode, he made the most of it. Though two of McKinnon's runs are mysteriously missing from NFL Game Rewind (in fact, a whole Minnesota drive is omitted), we can still see that the Vikings did a lot with McKinnon: 22 personnel power runs, shotgun read options, "full house"-type plays with two tight ends essentially in the backfield, and pitch plays to go around, not through, the Bills. And whatever the Vikings tried, it usually worked out OK.

Jerick McKinnon's Carries vs. BUF, Week 7
Qtr Time Down To Go Yard line Yards Direction Type
1 14:55 1 10 MIN 9 3 middle Dive/inside zone
1 14:22 2 7 MIN 12 11 right end Pitch w/pulling linemen
1 13:56 1 10 MIN 23 4 left end Pitch w/pulling linemen
1 6:52 2 20 MIN 37 8 right guard Dive/inside zone
1 3:49 1 10 BUF 27 3 right end Missing
1 3:13 2 7 BUF 24 2 middle Missing
2 13:01 1 10 BUF 41 3 middle Read option
2 9:19 1 10 MIN 40 1 left end Pitch w/pulling linemen
2 8:43 2 9 MIN 41 7 middle "Full House" misdirection
3 13:00 1 10 MIN 6 29 middle Dive/inside zone
3 12:20 1 10 MIN 35 1 middle Inside zone
3 2:28 1 10 MIN 49 9 middle Read option
3 2:03 2 1 BUF 42 1 middle Read option
3 0:45 2 16 BUF 47 -2 right tackle Read option
4 14:52 1 10 BUF 21 0 middle "Full House" misdirection
4 12:19 3 23 BUF 23 8 middle Dive/inside zone
4 6:45 2 10 MIN 22 1 left guard Dive/inside zone
4 5:28 1 10 MIN 36 13 left guard Read option
4 4:41 1 15 MIN 44 1 left guard Dive/inside zone
Runs that gained a first down are shaded yellow.

Let's examine some of McKinnon's biggest runs against Buffalo, in the order they occurred in the game. Unfortunately, coaches' film for Sunday's games had not been uploaded as of the time of this writing, so we're limited to standard TV views.

First quarter, 14:22 to go, second-and-7 at the MIN 12: The Vikings line up in an I-formation with both receivers to the right, then pitch the ball to that side. Right tackle Phil Loadholt pulls to that side for some extra blocking power. The Bills blitz up the middle, a play-call that leaves them vulnerable to this kind of pitch, and McKinnon doesn't do anything special to make this a successful play. When he hits the corner, though, he makes his cut and zips up the field. It's rare acceleration, and a lot of NFL running backs might have gained 5 or 6 yards on the play instead of 11. (The Bills jumped offsides on the play; the Vikings denied the penalty, of course.)

Third quarter, 13:00 to go, first-and-10 at the MIN 6: This is McKinnon's longest run of the day, and he did the bulk of the work by himself. The Vikings start their first drive of the second half backed up deep in their own end. They go conservative, in an I-formation with two tight ends to the right. McKinnon starts to the left, then cuts back to the right. Mario Williams has a chance to tackle McKinnon from behind for a short gain, but McKinnon muscles through the arm tackle. Aaron Williams is then in perfect position to make the tackle, but McKinnon dismisses him with a Stiff-Arm of Complete Disdain. McKinnon's physicality is impressive enough, but what he does next might be my favorite part of the play. From a nearly stopped position, he turns on the jets and zips past several Bills defenders for about 15 more yards, gaining 29 in total on the play. This is some pretty special zero-to-60 time, and it produced the longest run Buffalo has allowed all year.

Third quarter, 2:28 to go, first-and-10 at the MIN 49: A simple read-option play that shows off McKinnon's power. McKinnon take the ball and immediately disappears into a pile of bodies. Time freezes for a moment, then McKinnon pops through the other side, and by the time the Bills realize it, McKinnon has a 9-yard gain. Here is where I'd really like to have access to the end zone cam. It almost appears that McKinnon barrels over his own offensive tackle and uses Loadholt's body as a path downfield.

Fourth quarter, 5:28 to go, first-and-10 and the MIN 36: Another read option, this one catches the Bills off guard. McKinnon finds the gaping hole on the right side of the line and again shows off his one-step burst. As the old saying goes, McKinnon isn't quick, he's sudden.

It's hard to find much not to like about McKinnon's game on Sunday. On one play it appeared he hesitated to hit a big hole in the line, and by the time he got there the hole was closed, resulting in a 1-yard gain that seemed like it should have been longer. Without the end zone camera angle, though, it's hard to judge how big the hole truly was, or to tell just what McKinnon saw on the play. It also appears the Vikings don't trust McKinnon in short yardage. The Vikings had five third-down plays with 5 yards or fewer to go for a first down: two runs by Asiata, one by Jerome Felton, and two passes by Teddy Bridgewater. They only converted one of those plays. Given McKinnon's steady production, there's no reason to think he couldn't have extended one of those drives.

This is McKinnon's second big game this season; he gained 135 yards on just 18 carries against Atlanta in Week 4. Two games is far too small a sample from which to draw any meaningful conclusions, but it's clear that Bridgewater is not the only exciting rookie in the Minnesota backfield.

Denver's Big Trio

You'll notice the Denver Broncos dominating this week's tables, with Peyton Manning the league's leading passer, Ronnie Hillman the top running back (including receiving value), and Demaryius Thomas the most valuable receiver. Is that the first time this has happened? By the time we noticed, it was too late to check for sure, but it seems safe to assume that somebody has done it at some point. Green Bay nearly did it in Week 2 last season, and the Saints just missed in Week 10.

Quarterbacks
Rk
Player
Team
CP/AT
Yds
TD
INT
Total
DYAR
Pass
DYAR
Rush
DYAR
1.
Peyton Manning DEN
22/26
318
4
0
232
232
0
Manning went 8-for-8 in the first quarter, and it's not like he was just checking down. Those eight completions netted 116 yards, two touchdowns, and four other first downs. He wasn't perfect, though; his last play of the quarter was a sack, proving that Manning is human ... -ish.
2.
Tony Romo DAL
17/23
279
3
1
156
153
3
Romo, meanwhile, was best late in the game. In the second half, he went 9-of-9 for 166 yards with one touchdown and four other first downs.
3.
Eli Manning NYG
21/33
248
3
0
145
140
5
As Aaron Schatz noted in Audibles, Manning made great use of his tight ends against Dallas. He threw nine passes to Larry Donnell and Daniel Fells, completing eight of them for 117 yards, one touchdown, and three other first downs. The ninth pass was an 11-yard DPI and another first down. (If you're looking for Donnell in the receiving tables, you won't find him, thanks mainly to a pair of fumbles.)
4.
Russell Wilson SEA
23/36
313
2
0
132
79
53
As you've probably heard by now, Wilson became the first player in NFL history to gain 300 yards passing and 100 yards rushing in a single game. In a case of too little, too late, Wilson was dominant on Seattle's last three drives, going 11-of-14 for 142 yards with two touchdowns and seven other first downs, while also rushing six times for 105 yard with another touchdown and three other first downs.
5.
Ryan Tannehill MIA
26/32
277
2
0
130
113
16
This, apparently, was the weak of streaky quarterbacks. In the first half, Tannehill went 14-of-15 for 176 yards with one touchdown and seven other first downs, though he was also sacked four times.
6.
Drew Brees NO
28/44
342
2
1
129
121
8
Speaking of streaky passers: On the Saints' first two drives, Brees went 8-of-11, but only gained 60 yards and two first downs, and he was also sacked once. On their next five drives he went 8-of-8 for 143 yards with a touchdown and six other first downs. And then on their last six drives, he went 7-of-19 for 82 yards with three first downs and an interception. Shockingly, his last 14 plays were all failures, a stretch where he went 3-of-14 for 17 yards and a pick.
7.
Andrew Luck IND
27/42
344
2
0
117
146
-29
Luck, like Eli Manning, was quite effective when throwing to his tight ends, going 8-of-12 for 136 yards with a touchdown and five other first downs when throwing to Coby Fleener, Dwayne Allen, and Jack Doyle.
8.
Aaron Rodgers GB
19/22
255
3
0
107
101
7
Rodgers likely would have ranked higher, but he was pulled after three quarters with the Packers ahead 35-3. On first downs, he went 9-of-10 for 170 yards with one sack, but five first downs, including all three touchdowns.
9.
Austin Davis STL
18/21
152
2
0
94
94
0
Only five of Davis' passes were thrown to receivers more than 10 yards past the line of scrimmage. Those five passes resulted in four completions for 76 yards, plus a 16-yard DPI.
10.
Tom Brady NE
20/37
261
3
0
74
74
0
On third downs, Brady went 7-of-10 for 74 yards with two touchdowns and three other first downs, plus a 32-yard DPI.
11.
Alex Smith KC
19/28
221
1
0
60
50
9
Smith's average pass attempt only traveled 4.61 yards beyond the line of scrimmage. The only starting quarterback who threw shorter passes was Aaron Rodgers (4.55).
12.
Ryan Fitzpatrick HOU
21/32
262
2
1
59
64
-4
Rk
Player
Team
CP/AT
Yds
TD
INT
Total
DYAR
Pass
DYAR
Rush
DYAR
13.
Ben Roethlisberger PIT
23/33
265
2
0
52
50
1
14.
Joe Flacco BAL
16/25
258
2
2
50
48
3
Feast or famine for Flacco on deep passes. He threw seven passes against Atlanta that traveled 15 yards or more downfield. None of those passes resulted in an incompletion. Four were complete for 100 total yards. A fifth resulted in a 36-yard DPI. The other two were intercepted.
15.
Colt McCoy WAS
11/12
128
1
0
47
49
-2
McCoy's completions gained 22 yards in the air and 106 yards after the catch, 63 of the on his very first throw.
16.
Derek Anderson CAR
5/8
43
1
0
39
39
0
All of Anderson's pass attempts came with Carolina down by 28 points in the fourth quarter.
17.
Philip Rivers SD
17/31
205
2
1
35
35
0
The KC-SD game was truly a contrast in styles. We already talked about Alex Smith's tendency to throw short passes. Meanwhile, Rivers focused on the deep ball, with an average pass attempt traveling 15.1 yards beyond the line of scrimmage, most of any starter this week. On ten deep passes, he went 3-of-9 for 78 yards and an interception, plus a 22-yard DPI.
18.
Carson Palmer ARI
22/31
253
2
1
30
30
0
On third and fourth downs, Palmer went 10-of-13 for 131 yards with eight conversions, including two touchdowns, plus one sack.
19.
Geno Smith NYJ
20/33
226
1
0
20
13
7
Smith was very bad throwing to most areas of the field, but he tore up the short middle of the Patriots' defense, going 8-of-12 for 117 yards with seven first downs, including a touchdown.
20.
Charlie Whitehurst TEN
17/26
160
2
1
-5
-4
-1
Whitehurst threw five deep passes in the first half, four incomplete, one intercepted. He only threw one deep pass in the second, half, a go-ahead 38-yard touchdown to Derek Hagan.
21.
Derek Carr OAK
16/28
173
0
0
-6
-1
-5
On Arizona's half of the field, Carr went 1-of-5 for 7 yards and no first downs.
22.
Colin Kaepernick SF
24/39
263
1
1
-11
-16
4
Passes to running backs and wide receivers are supposed to be high-percentage plays. Not for Kaepernick, who went 4-of-8 for 29 yards and only one first down on passes to Vernon Davis, Carlos Hyde, and Frank Gore.
Rk
Player
Team
CP/AT
Yds
TD
INT
Total
DYAR
Pass
DYAR
Rush
DYAR
23.
Cam Newton CAR
17/31
205
1
1
-31
-31
0
Newton did not complete a pass on Green Bay's side of the field until Carolina was down by 35 points in the fourth quarter. On the Packers' side of the 50, he went 2-of-7 for 20 yards with one interception, one sack, and one very meaningless touchdown.
24.
Kyle Orton BUF
32/42
283
2
1
-33
-33
0
On Minnesota's half of the field, Orton went 12-of-16, but for just 93 yards and five first downs (though that included the game-winning touchdown). Meanwhile, three of those completions lost yardage, he wa sacked three times, and he also committed an intentional grounding penalty.
25.
Jay Cutler CHI
21/34
190
1
1
-48
-47
-1
Long-yardage situations were death for Cutler against Miami. With more than 10 yards to go for a first down, he went 4-of-6 for 41 yards, with one first down, two sacks, and one fumble.
26.
Matt Ryan ATL
29/44
228
1
0
-61
-66
5
Ryan moved the ball at times, but had big problems on either end of the field. Inside his own 20, he went 5-of-9 for 36 yards with one first down and one sack for a safety. Inside the Baltimore 40, he went 6-of-11 for 47 yards with two first downs (including a touchdown), two sacks, and one fumble.
27.
Kirk Cousins WAS
11/16
139
0
1
-77
-77
0
Throwing to the middle of the field, Cousins went 2-of-5 for 4 yards with no first downs and one interception.
28.
Matthew Stafford DET
27/40
299
2
2
-90
-87
-4
Stafford failed to do much in short-yardage. With 6 yards or less to go for a first down, he went just 3-of-9 for 15 yards with one touchdown, one other first down, one interception, and two DPIs for 34 total yards.
29.
Andy Dalton CIN
18/38
126
0
0
-115
-112
-4
The Bengals started this game with eight straight three-and-outs, which is very, very bad. Over that stretch, Dalton went 6-of-17 for 12 yards (not a typo) with a long gain of just 8 yards, no first downs and, even further, no successful plays. Good Lord.
30.
Teddy Bridgewater MIN
15/26
157
1
2
-131
-131
-1
No Kyle Rudolph and no Adrian Peterson means that Bridgewater went 3-of-10 for 11 yards with one first down and an interception when throwing to his backs and tight ends.
31.
Brian Hoyer CLE
16/41
215
0
1
-149
-149
0
Inside the Jacksonville 40, Hoyer went 2-of-7 for 9 yards with one first down and one sack.
32.
Blake Bortles JAC
17/31
159
1
3
-160
-176
15
What a shootout this was. On third downs, Bortles went 4-of-11 for 40 yards, with three first downs, with two sacks and two interceptions.


Five most valuable running backs (Total)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
1.
Ronnie Hillman DEN
14
74
2
4/4
29
0
43
23
20
Hillman had touchdown runs of 1 and 37 yards, and added an 11-yard gain, with only two carries for no gain or a loss. As a receiver, he twice gained 11 yards on first-and-10.
2.
Matt Forte CHI
12
49
1
6/7
60
1
42
20
22
Forte's longest run covered 16 yards. He added a 1-yard touchdown on third-and-goal, while being stuffed only twice. As a receiver, he had a 10-yard touchdown, plus catches of 24 and 10 yards, and a 6-yard gain on fourth-and-2.
3.
Le'veon Bell PIT
12
57
0
8/8
88
1
41
2
39
4.
Ahmad Bradshaw IND
10
52
1
3/3
36
1
26
-6
32
Bradshaw makes this list despite fumbling on a running play. In addition to his goal-line touchdown, he had runs of 15 and 11 yards, plus a 3-yard gain on third-and-1. His three catches included an 18-yard gain on third-and-1 and a 10-yard touchdown.
5.
Trent Richardson IND
14
77
0
4/5
41
0
25
10
16
Richardson had catches of 20 and 13 yards, plus runs of 10 and 11 yards and a 5-yard gain on second-and-1.


Five most valuable running backs (Rushing)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
1.
Jerick McKinnon MIN
19
103
0
2/4
-2
0
7
30
-24
McKinnon's receptions were a 6-yard loss on first-and-10 and a 4-yard gain on second-and-16. So that's bad.
2.
Denard Robinson JAC
22
127
1
0/1
0
0
21
25
-4
Toby Gerhart currently has the worst rushing DYAR in the league, so maybe Robinson should stick in the starting lineup. He had five runs of 10 or more yards, tied with DeMarco Murray for most in the league.
3.
Jamaal Charles KC
22
95
1
2/4
12
0
8
24
-17
Charles makes this list mainly because he was so successful in short yardage, converting five of seven carries with 6 yards or less to go for a first down.
4.
Ronnie Hillman DEN
14
74
2
4/4
29
0
43
23
20
5.
Chris Ivory NYJ
21
107
1
4/4
18
0
22
22
0
Ivory was stuffed for a loss just once, but he had eight total first downs, converting seven of nine carries with 6 yards or less to go for a first down.


Least valuable running back (Total)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
1.
Giovani Bernard CIN
7
17
0
2/5
-1
0
-49
-17
-32
Bernard's longest carry, and only successful run, was a 6-yard gain on first-and-10. He twice failed to convert and second-and-4. His receiving numbers, well, they speak for themselves.


Least valuable running back (Rushing)
Rk
Player
Team
Runs
Rush
Yds
Rush
TD
Rec
Rec
Yds
Rec
TD
Total
DYAR
Rush
DYAR
Rec
DYAR
1.
Andre Williams NYG
18
51
0
0/3
0
0
-47
-31
-16
Williams was stuffed six times for no gain or a loss, and had four other runs for exactly 1 yard.


Five most valuable wide receivers and tight ends
Rk
Player
Team
Rec
Att
Yds
Avg
TD
Total
DYAR
1.
Demaryius Thomas DEN
8
9
171
21.4
2
83
Thomas had three catches of 32 or more yards. He had two failed completions: a 15-yard gain on third-and-26 and a 14-yard gain on third-and-17.
2.
Kenny Stills NO
5
5
103
20.6
1
64
Each of Still's catches gained at least 10 yards and a first down. He also drew an 11-yard DPI to convert a second-and-24.
3.
Dez Bryant DAL
9
12
151
16.8
0
55
Bryant produced five first downs, but no touchdowns. That's mostly an issue of opportunity. He had no targets in the red zone, and just one inside the New York 40.
4.
Sammy Watkins BUF
9
14
122
13.6
2
53
Watkins led the league with nine receiving first downs this week.
5.
Torrey Smith BAL
3
5
81
27.0
1
51
Each of Smith's catches gained at least 22 yards, and he also drew a 36-yard DPI.


Least valuable wide receiver or tight end
Rk
Player
Team
Rec
Att
Yds
Avg
TD
Total
DYAR
1.
Jermaine Gresham CIN
10
12
48
4.8
0
-50
The Bengals only passed for five first downs, and Gresham had two of them. He also had a 5-yard loss, plus incompletions on third-and-3 and fourth-and-4, and a 6-yard gain on third-and-7.

Posted by: Vincent Verhei on 20 Oct 2014

95 comments, Last at 24 Oct 2014, 9:59am by Karl Cuba

Comments

1
by theslothook :: Tue, 10/21/2014 - 3:24am

I feel like next year - assuming Bridgewater isn't a christian ponder redux, the vikings will be a much improved team both record wise and through dvoa. The defense has pieces scattered across the roster and the offensive line is bound to improve from this year's nosedive. Maybe a free agent or two will be enough to help "bridge" the gap. Sorry, couldn't resist :P

8
by Will Allen :: Tue, 10/21/2014 - 8:07am

They've just had so many weird things happen that it makes it hard to evaluate. They need to find out once and for all if Kalil is serviceable at left tackle; it really is confounding to see a guy, who played like he did as a rookie, play like he has from about halfway through his 2nd season and now halfway into his third. So much will depend on what Bridgewater turns out to be (I'll be shocked if Peterson ever plays for the Vikings again), and I wouldn't bet one way or another.

24
by Noah Arkadia :: Tue, 10/21/2014 - 11:56am

I'd bet on Bridgewater, even if he hasn't looked very good lately.

------
Who, me?

25
by Will Allen :: Tue, 10/21/2014 - 12:18pm

Well, I'm certainly not a pessimist, but he really is going to have to be mentally exceptional by NFL starting qb standards, and have some talent around him, because his physical tools seem pretty average, although that's hard to judge at this point, given the poor pass protection he's mostly received. Maybe he'll be really accurate, but he doesn't seem to be a guy who is going to drive the ball into small windows. I've not seen any All-22 tape with him, so I can't say how well or poorly he's been so far, at quickly identifying the right guy to throw to. That quality, and how good he is pre-snap, are going to be critical, because he isn't like a Kaepernick or some other guys who can overcome deficiencies with blatant physical superiority.

40
by Jianfu :: Tue, 10/21/2014 - 4:26pm

One thing about Bridgewater is he's remarkably young, even by rookie standards (he's still 21, turns 22 in November). Ignorant as I am about any definitive studies on age for rookie quarterbacks, a quick review shows most of the rookie starters of late came into the league during their age-24 seasons (defined by pff-reference as their age as of 12-31 of that year) when they were 2-3 years older than Bridgewater is: Russell Wilson, dalton, tannehill, Kaepernick (who didn't start until his age 25 season regularly). The closest recent rookie qbs to Bridgewater's age were Cam Newton and Andrew Luck (two guys who might have been created in a top secret lab), and an athletic marvel in RG3. (And I believe Bridgewater is younger now than these three were.)

Now, I'm not suggesting Bridgewater's going to experience massive physical development in the next couple years, but he might develop a little more physicality and strength. And if you consider smaller qbs with his profile (more or less), the successful ones have all come of age as quality qbs during their age 24 seasons (Brees in his third year, Russell Wilson and Dalton as rookies). Of course who knows if that's what we'll see with Bridgewater.

Edit: to be a little more comprehensive (but still not really), Jake Locker, geno smith, ej manuel, Ryan, flacco, Matt Leinert, and Mark Sanchez were all rookies and starters during their age 23 seasons; Blaine Gabbert might be the closest of all to Bridgewater as he too was a rookie starter in his age 22 season (but started the season at 21), as would Manziel if he becomes the starter this year.

2
by mushin2003 :: Tue, 10/21/2014 - 4:41am

Vincent thanks for the cliff notes on these players. Does the DYAR calculated take into account the SOS for the defenses each player faced, such as the rushingD or passingD DVOA ranking? At times it read as if you use the fact of a DPI as a positive and other times as either a negative fact or ignored fact.
While the DYAR rankings give another perspective on things the QB DYAR seems almost looks like a randomly generated list. I believe you would be hard press to find anyone who could or would have ranked them in this manner. The algo for the QBs DYAR is very unique.

3
by Vincent Verhei :: Tue, 10/21/2014 - 4:46am

The "D" in "DYAR" stands for Defense-Adjusted. So yes, all players' numbers are adjusted for the defenses they faced.

DPIs are pretty much always a positive for the offense, aren't they? I can't think of how that could be read any differently.

4
by JoeyHarringtonsPiano :: Tue, 10/21/2014 - 5:05am

Stop the presses! Trent Richardson made the most valuable running back list.

6
by coboney :: Tue, 10/21/2014 - 7:18am

but... but thats impossible!

7
by Kevin from Philly :: Tue, 10/21/2014 - 7:56am

Search your feelings...you know it to be true!

5
by Tomlin_Is_Infallible :: Tue, 10/21/2014 - 6:05am

hard to believe Foster didn't register.
oh wait, that was the Steelers run D he just smashed.

--------------------------------------
The standard is the standard!

10
by JoeyHarringtonsPiano :: Tue, 10/21/2014 - 8:44am

I also did a double take when I didn't see Golden Tate on the most valuable receivers...then I realized he played against the Saints pass defense.

11
by RickD :: Tue, 10/21/2014 - 9:13am

The fumble was bad. In his defense, I'm sure he thought he was down already.

9
by herewegobrownie... :: Tue, 10/21/2014 - 8:42am

"Adrian Peterson means that Peterson went 3-of-10 for 11 yards"

Typo alert.

12
by RickD :: Tue, 10/21/2014 - 9:16am

That's a monster rushing DYAR for Wilson. It would put him ahead of all the RBs by quite a bit.

16
by Alternator :: Tue, 10/21/2014 - 10:06am

When the baseline is dominated by guys like Henne and Carr, rather than RG3 and Vick, there's not much you have to do in the "Above Replacement" part.

17
by RickD :: Tue, 10/21/2014 - 10:25am

Is rushing DYAR computed differently for QBs than for RBs? I don't see why it should be.

26
by Alternator :: Tue, 10/21/2014 - 12:42pm

I'm assuming that it's comparing QBs to QBs, and RBs to RBs, which I think I've seen mentioned before. The average QB isn't a good rusher, so when you have one that is, it's fairly easy to look good compared to 'replacement level'.

32
by RickD :: Tue, 10/21/2014 - 2:31pm

Yes, the explanatory text suggests that QBs are compared to QBs, so you may be right. I'd be curious to know what rushing DYAR Wilson would be credited with if he had been been an RB.

43
by anotherpatsfan :: Tue, 10/21/2014 - 4:50pm

I wonder how meaningful that calculation would be. It is apples and oranges. The running that QBs do usually happens in a different context than running back carries -- RBs don't get to hang out in the pocket and decide to go if it looks good because the linebackers dropped. 3rd and 9 scrambles are fundamentally different animals then 3rd and 9 shotgun hand-offs.

13
by Ben :: Tue, 10/21/2014 - 9:27am

Wow, that Browns-Jags game must have been eye-bleedingly bad (and since it didn't involve the Cowboys, it wasn't because of Ebola. Too soon?) I watched all of the Colts-Bengals game, and Dalton was just useless. I guess he didn't have any turnovers, which must be why he isn't at the bottom of the list.

76
by Theo :: Wed, 10/22/2014 - 4:45am

As a game it was quite watchable. Bad, but better than the snoozefest of the Vikings - Lions that I saw last week.

14
by Paul R :: Tue, 10/21/2014 - 9:46am

Just as I was scrolling down to the Most Valueable Running Backs and finding two Colts players in the top five, and one of them Trent Richardson, there was a clap of thunder outside. Might be a good day to stay indoors, under the bed.

15
by Paul R :: Tue, 10/21/2014 - 9:54am

..

18
by Anger...rising :: Tue, 10/21/2014 - 10:26am

Rodgers likely would have ranked higher, but he was pulled after three quarters with the Packers ahead 35-3

Yeah, looks like an extra 2-11 for -7 yards could've gotten him up to third...

19
by dmstorm22 :: Tue, 10/21/2014 - 10:31am

You do realize that this is 2014, where the Panthers defense is playing like what we all thought Dallas would be, and Dallas is playing average?

20
by armchair journe... :: Tue, 10/21/2014 - 10:55am

Stiff Arm of Complete Disdain is of course a reference to some obscure late '90s emo ballad on Vince's fav mixtape.

//AJMQB

21
by serutan :: Tue, 10/21/2014 - 11:19am

That should have read "Stiff Arm of Sweetness", as Payton was the senseii of
stiffarming.

______
Was wr

22
by Paul R :: Tue, 10/21/2014 - 11:25am

Actually, I think it's a movie I saw at a bachelor party once. Can't quite remember, but it was a similar title.

23
by Rick_and_Roll :: Tue, 10/21/2014 - 11:41am

Is there a quick reads for defensive players? It would be great to see how guys like Watt, Miller and Sherman are affecting the game.

27
by Toner :: Tue, 10/21/2014 - 1:17pm

"Third quarter, 2:28 to go, first-and-10 at the MIN 49: A simple read-option play"

That's not a read option. Look at it again - no one is unblocked and the linebackers and corner run towards the A-gaps for the RB. If it's a read-option, who is Bridgewater reading on this play? And what about the defense indicates he should hand off since everyone moves inside?

Compare it with the last gif which is actually a read option: DE 94 is unblocked and runs straight at the mesh, so Bridgewater hands off. Not blocking the DE allows the RT to block down on the DT and then move to the linebacker, opening up a huge hole for McKinnon.

28
by Spoon :: Tue, 10/21/2014 - 1:33pm

I'd be curious to hear how Andrew Luck recorded -29 rushing. That's a number that would put him in contention for the least valuable running back on the week, when his passing DYAR was third-best. I assume the fumbled handoff with Trent Richardson plays a big part, but what else?

30
by Kevin from Philly :: Tue, 10/21/2014 - 1:53pm

Sacks, maybe?

31
by Hurt Bones :: Tue, 10/21/2014 - 2:26pm

1st Q(1:51) 2 and 7(CIN 7) Andrew Luck left end for 2 yards (tackle by Vontaze Burfict) 0 0 4.76 4.26 71.1
2nd Q(8:21) 2 and 4(CIN 33) Andrew Luck for no gain. Andrew Luck fumbles, recovered by Carlos Dunlap at CIN-36 (tackle by Trent Richardson)
3rd Q(8:01) 1 and 11(IND 1) Andrew Luck up the middle for 1 yard (tackle by Wallace Gilberry and Jayson DiManche)
4th Q(12:57) 3 and 1(CIN 12) Andrew Luck right guard for 2 yards (tackle by Vincent Rey)

Looks like it's the fumble. I'm guessing if a RB went 4 rushes for 5 yds with a fumble in those four situations they might have even more negative DYAR.

29
by TomC :: Tue, 10/21/2014 - 1:35pm

In the first half, Tannehill went 14-of-15 for 176 yards with one touchdown and seven other first downs, though he was also sacked four times.

This has a lot to do with the fact that the Bears' D-line is playing pretty well this year, but the back seven is a hot mess. Tannehill did play well, but most QBs this year against the Bears have had wide-open guys to throw to if they can stay upright. Rodgers, Newton (in the 2nd half), and Tannehill have taken advantage; G. Smith and Matt Ryan did not, partly due to their receivers' mistaking Crisco for stickum.

35
by theslothook :: Tue, 10/21/2014 - 3:07pm

Robert Mays says the linebacking core is the worst in football. Exaggeration or spot on?

61
by tuluse :: Tue, 10/21/2014 - 6:19pm

I have not seen all 32 teams' linebackers in enough detail to be sure, but they are plenty awful.

33
by intel_chris :: Tue, 10/21/2014 - 2:55pm

I was wondering how those failed completions rate in terms of DYAR. They stood out in my mind as "conventional stat padding plays with no impact on the actual game". Moreover, I assume that for DVOA they are simply failures, but given the down and yards-to-go, they were more "we-don't-have-anything-better-to-do" plays, and perhaps against "replacement" they aren't actually negative.

Actually, I'm curious about those "need-a-near-miracle" type plays, i.e third-down and 15 or more yards to go. It seemed like Kaepernick pulled more than a couple of them off early in the game. What is the success rate for such plays, 5%, 10%, 25% (I'm assuming not that much), but it does seem like that there is often one or more per game, so it isn't a complete long short. Are some defenses more susceptible to them than others? Are some teams more successful at getting them that others?

34
by oaktoon :: Tue, 10/21/2014 - 2:58pm

Rodgers' placement is Exhibit A as to why I don't feel DYAR is anywhere near an ultimate metric... All he did was get his team ahead by 5 TDs... and thus didn't play the entire 4th quarter. So he's 8th in the league for the week... Yeah, right....

36
by theslothook :: Tue, 10/21/2014 - 3:09pm

It's a cumulative stat, hence why a guy throwing less will not accumulate as much. I don't like it for other reasons, but the logic was to provide a balance between players who were highly efficient but threw very few times to those forced to throw to carry their teams more and had more negative plays sprinkled in.

41
by Duff Soviet Union :: Tue, 10/21/2014 - 4:38pm

"All he did was get his team ahead by 5 TDs."

Yeah, the defense had absolutely nothing to do with that whatsoever, it was all no. 12.

Just looking around, I can see that Brian Burke and Advanced NFL Stats had him 8th in WPA and 4th in Estimated Points Added (behind Manning, Wilson and Romo), and those numbers are not opponent adjusted.

I can't find a weekly QBR listing, but I can see that Rodgers is behind Manning, Romo and Austin Davis and about level with Eli, Wilson and Tannehill.

So, either all three of these methods are "laughable" or your criticism is. I know which way I'm guessing.

77
by ammek :: Wed, 10/22/2014 - 5:51am

The quarterback with the highest QBR in the Packers-Panthers game was … Derek Anderson!

37
by Vincent Verhei :: Tue, 10/21/2014 - 3:28pm

hard to believe Foster didn't register.
oh wait, that was the Steelers run D he just smashed.

He actually had negative DYAR on the day, thanks to the fumble and four stuffs, one of which came on third-and-1.

I also did a double take when I didn't see Golden Tate on the most valuable receivers...then I realized he played against the Saints pass defense.

He finished 14th. He might have made the tables were it not for incomplete passes on third-and-5 and second-and-2.

"Adrian Peterson means that Peterson went 3-of-10 for 11 yards"

Fixed.

Is rushing DYAR computed differently for QBs than for RBs? I don't see why it should be.

All players, in all stats, are compared to other players at their position. WRs are compared to WRs, RBs to RBs, QBs to QBs. In this case, that actually raises the baseline. RBs this year are averaging 4.1 yards per carry with a 47 percent Success Rate. QBs are averaging 6.2 yards per carry with a 62 percent Success Rate.

I'd be curious to hear how Andrew Luck recorded -29 rushing. That's a number that would put him in contention for the least valuable running back on the week, when his passing DYAR was third-best. I assume the fumbled handoff with Trent Richardson plays a big part, but what else?

There’s something weird here with a fumbled snap. It’s listed as a passing play in our spreadsheets, but it looks like it was somehow counted as a running play in QR calculations. So that’s most of the -29, though a 2-yard gain on second-and-7 and 1-yard gain on first-and-11 are also bad plays.

I was wondering how those failed completions rate in terms of DYAR. They stood out in my mind as "conventional stat padding plays with no impact on the actual game". Moreover, I assume that for DVOA they are simply failures, but given the down and yards-to-go, they were more "we-don't-have-anything-better-to-do" plays, and perhaps against "replacement" they aren't actually negative

This is something that is too complicated to get into here, but a quick data dump:

The average incompletion (not counting interceptions) in 2014 has been worth -7 DYAR.
The average failed completion has been worth -3 DYAR.
The average successful completion has been worth 12 DYAR.

Rodgers' placement is Exhibit A as to why I don't feel DYAR is anywhere near an ultimate metric... All he did was get his team ahead by 5 TDs... and thus didn't play the entire 4th quarter. So he's 8th in the league for the week... Yeah, right...

Rodgers had kind of a weird day. Yes, he went 19-of-22, but he had seven failed completions and was also sacked twice. So his Success Rate was only 50 percent. I mean, yes, Green Bay had five touchdowns in three quarters, but they also had three punts and a 33-yard field goal drive. And that’s with James Starks and Eddie Lacy both averaging better than 5 yards per carry, with Success Rates higher than 70 percent. Rodgers is dinged fairly highly for failing to convert on three plays with 4 yards or less to go for a first down.

38
by oaktoon :: Tue, 10/21/2014 - 4:03pm

"Rodgers had kind of a weird day" Oh, boy, this is gonna be fun...

First Drive: 80 Yards TD.. 59 Yard TD pass to nelson; Rodgers is sacked, but gets Packers another 1st down in large part because his hard count drew the Panthers offsides twice.... (does that count for him? Does the INT he "threw" knowing he had the penalty already count as a "failed completion"? Still trying to understand that term as he only threw three incompletions in this game)

Second Drive: 56 Yards TD... 5 yard lacy Run... Rodgers threw an INC pass on 3rd down but he hurried the snap because he knew Panthers had 12 men on field--- which they were called for. (Does that count for him.. or was the "INCOMPLETE PASS" counted against him?) there was one legit INC pass on this drive.

Third Drive: 62 Yards TD... Starks 13 yd run... Pretty standard stuff

Fourth Drive-- a 3 and out with an INC pass on 3rd down

Fifth Drive: 94 Yards TD... Again a hard count led to a neutral zone infraction... TD pass of 3 yards to Cobb

Sixth Drive-- 3 and out-- first down completed pass-- two failed running plays on 2nd and 3rd down..

Seventh Drive-- 58 Yards TD-- in 4 plays, the final one being 1 21 yard pass to Adams...

Eighth Drive-- FG... One first down and then a 3 and out with his final INC PASS...

So the Packers got the ball 8 times...In 5 of their possessions, Rodgers led his team 80, 56, 62, 94 and 58 yards for TDs... In another he led them to a FG.... they punted twice (not three times, I am sorry to inform you) with him in game....

So gaining 38 pts in 8 possessions in 3 quarters with only 3 incompleted passes, 3 TD passes and no turnovers is "weird"????

What is weird is DYAR's assessment of this performance... laughable, actually...

39
by theslothook :: Tue, 10/21/2014 - 4:21pm

There's really no mystery to this whatsoever. You're bringing up drive success - but that will be reflected in the TEAM DVOA page.

As to Rodgers for DYAR. Its a counting stat. Rodgers didn't throw that much, hence he didn't vault that high in the rankings. He was also playing a bad defense and took sacks. Since those are baked into the formula - it hurts his rankings.

44
by oaktoon :: Tue, 10/21/2014 - 4:52pm

"there's really no mystery..." true--- A great QB had a great 3 Quarters-- and a metric penalized him for two sacks-- one of which was in a drive that culminated with a 59 YD TD Pass-- almost certainly gave him no credit for causing 3 neutral zone infractions and exploiting a 12 man on field penalty-- and stopped him from "vaulting high" because he used "only" 22 passes to produce 38 pts... In other words, if he had been throwing the ball all over the place--- lots of INCs/lots of 6-8 Yd completions but none of those dreaded sacks, he would "vault higher".

It isn't mysterious, I agree...Bogus is a better word... Listen I applaud you guys for trying... but when a guy gets 38 pts in 45 minutes-- against North Carolina or Carolina-- he ain't the 8th best QB of the day.... Did Dalton have a higher DYAR for scoring 37 pts in 60 minutes against this same bad defense? Just curious.... (I looked. not much worse. dalton threw 2 TDs and 2 Picks-- scored 37 pts in 75 mins; compared to rodgers 38 pts in 45 mins with 3 TDS and 0 Picks... Dalton got 94 DYAR to Rodgers 107-- not far apart... I rest my case)

49
by theslothook :: Tue, 10/21/2014 - 5:15pm

First, you're committing the egregious crime of attributing team offensive success all to the qb. If you want to do that, go for it, but time and time again its shown to be a lazy way of doing things.

Let's get to the actual issues here. DYAR tracks certain metrics for a reason - they are predictive. Getting neutral zone infractions is indeed good, but its not predictive(you can disagree, but that's what the data suggests). That goes for everything working in dvoa and dyar - giving added weight to things that are predictive and lower weights to things that are not.

That said, I don't go here to pretend like dyar is going to tell me who the best qb was this week. The closest stat that does that is qbr but even that is flawed. Instead, I and probably many like me, use it along with a number of other metrics to get a rough sense of how teams and the players did. DYAR will tell us something about how efficient they were relative to the opponent along with what sort of performances they had converting third downs and being efficient on 1st down.

62
by tuluse :: Tue, 10/21/2014 - 6:21pm

Rodgers also did not have the most passing yard this week. Bogus indeed.

42
by carljm :: Tue, 10/21/2014 - 4:44pm

A "failed completion" means a completed pass that wasn't actually a successful play in context - like a 7-yard completion on 3rd and 10.

I don't know whether DVOA/DYAR give a QB any credit for a neutral zone infraction, though I kind of doubt it. It would be interesting to see how much statistically significant correlation there is season-to-season in which QBs are successful with hard counts, and whether adding that to DVOA (presuming it's not already there) improves its predictive correlation at all.

45
by oaktoon :: Tue, 10/21/2014 - 5:04pm

Well they credit him for 7 "failed completions"-- but somehow the Packers scored 5TDs and a FG in 6 out of 8 drives with Rodgers at the helm... So what precisely were the failures?? I suspect the penalties counted against him, which of course is a ridiculous conclusion given that Rodgers, among others, nearly always gambles on a big play when he knows he has a flag in his pocket....

All I am saying is that when a team manages to score TDs on 5 drives, a FG on another-- all of them from their side of the field, 2 of them 80 yds or more, mostly from passing yards, with a QB who is 19-22 and throws 3 TDS and no INTs-- i just don't see "failure" or "weirdness"-- their words, not mine...

"His "success rate" was only 50 percent"-- i am trying to get my arms around that little gem also... Gee-- i thought success in football was measured by Touchdowns... 38 pts in 8 possessions is a "success rate" of 50%??? What would 100% be-- 76 pts in 3 quarters/8 possessions? I'd love to see that game!!

"they were baked into the formula"-- get a new oven...

46
by dmstorm22 :: Tue, 10/21/2014 - 5:05pm

Read what this site is about, what they are tracking, what they are calculating, and what a failed vs. successful completion is.

If you still don't like it, then I suggest that this probably isn't really the site for you.

48
by Will Allen :: Tue, 10/21/2014 - 5:12pm

Nothing like a Packer hot streak to bring out a certain, er, aggressiveness in the commentary. It may get like 2011 in here before too long. Where's Big10Freak to lend some reason?

56
by DisplacedPackerFan :: Tue, 10/21/2014 - 5:36pm

I hope it doesn't get to 2011 levels. The ... aggressiveness (I like that term) of some of my fellow fans makes me want to just go and hide for awhile. DVOA has pegged the Packers and Rodgers pretty well over the years.

57
by Will Allen :: Tue, 10/21/2014 - 5:49pm

Another rational cheesehead! Hey, that was pretty entertaining in November and December 2011!

59
by Hurt Bones :: Tue, 10/21/2014 - 6:05pm

In this instance I believe it's Turoencephalitis. Too bad because I'm pretty sure there is a vaccine.

63
by tuluse :: Tue, 10/21/2014 - 6:24pm

I think I find strangest (and this doesn't apply only to Packers fans, but this is a great example). When fans are enjoying a QB as incredible as Rodgers, isn't that enough in it's own? I wouldn't care a wit what DYAR said about him. You can see with your eyes it's a top 3 QB in the league.

I think you guys need to spend a couple decades with the likes of Rex Grossman and Jim Miller to reset your expectations. :)

50
by oaktoon :: Tue, 10/21/2014 - 5:15pm

i want a site that tells me who is valuable... I presumed that's what DYAR does... So the one time I see a game with immense value from one particular player, it isn't registered... So I should leave the site because DYAR "failed this completion"? Or maybe poke around a bit instead-- in hopes that the metric might be improved? if the answer is Rodgers is punished for throwing incompletions/interceptions that never count in real life because he had drawn the opposition offside, then the metric needs to be improved... If the answer is Rodgers is punished because he didn't do something optimally on a drive that culminated with him producing a TD, then the metric needs to be improved... If the answer of why Andy Dalton can face the same defense and produce 1 pt every 2 minutes with two INTs, and Rodgers faces them and produces 1 pt every 1 minute 10 seconds with no INTS and yet their rating is more similar than different, then the metric needs to be improved... I don't believe in blind loyalty to any metric... because FO does good work, my issues ought to be respected, not turned aside...

51
by theslothook :: Tue, 10/21/2014 - 5:17pm

You really need to read how the stat is calculated...

53
by oaktoon :: Tue, 10/21/2014 - 5:29pm

Maybe not just me.... That is my point, after all...

54
by Will Allen :: Tue, 10/21/2014 - 5:32pm

Why on earth would you want a metric measuring individual performance that doesn't count it against a player when he fails to meet the optimal outcome, because the player later threw a td pass on that drive? Huh? You really think that a guy, starting a drive on the fifty, who had two incompletions, then an 10 yard completion, then two incompletions, followed by a 20 yard completion, then two incompletions, followed by 20 yd td pass, played as well as a guy who had a 20 yard completion, then a 20 yard completion, then a td pass? Really?

55
by dbostedo :: Tue, 10/21/2014 - 5:34pm

"...If the answer is Rodgers is punished because he didn't do something optimally on a drive that culminated with him producing a TD."

First, yes, he is "punished" when he doesn't do things optimally, regardless of whether or not the TEAM scored a TD.You need to really understand that to understand how the ratings work. Two incompletes and a 60 yard touchdown are NOT worth as much as three 20 yard completions ending in a TD. That makes sense because a QB who has 3 20 yard completions is more likely to be successful the next time than the QB with 2 failed plays and 1 long bomb. That's a fact that's been shown time and again in developing this system. So unsuccessful plays DO matter for making the assessment better, despite the TD outcome on the drive being the same.

If you're not willing to understand how that works, then there's no point in trying to understand the rankings or complain about them.

Second, DVOA is a rate stat, as opposed to DYAR. So DVOA will help filter out the amount of usage a QB gets. The weekly rankings are based on DYAR though, so QBs with a ton of attempts sometimes rank highly even if their scoring isn't as high, provided they have a lot of successful plays. I'm not sure what the DVOA for Rodgers is for that game or the season, but that could perhaps change your view a bit.

70
by intel_chris :: Tue, 10/21/2014 - 9:22pm

I don't think your issues are being turned aside. However, the site offers a variety of metrics that measure different aspects of the play of the game and how players, teams, etc. have influenced the outcome. It also offers other analysis to cover points that aren't in its measurements.

One of the things FO is noted for is concentrating on metrics that are reproducible (I'm not sure that's the right word, maybe predictive or some other word is better), i.e. when a player (team) scores well on a stat in one game, it is likely that that player (team) will perform at a similar level in future games (on the same stat).

The canonical counter-example is fumble recoveries, FO has measured fumble recoveries numerous times and has found that while there are correlations between fumble recoveries and where the fumble occurs, their isn't a significant correlation between a team's rate of fumble recoveries in one part of the season with that in a different part. Instead teams "regress to the mean", in other words they will tend to have an average number of fumble recoveries going forward (the gambler's fallacy is to assume that number will be below average to even out the total to average; which is just as much a fallacy as assuming that the fumble rate will continue indefinitely to be above or below average).

Anyway, similar things are used to determine what events are counted as part of they various statistics collected here. The events that are counted are ones that have been measured to have a predictive effect. Some events are sufficiently rare to be non-predictive. For example, if only a small percentage of QBs can do hard counts which draw the other team offside, and even less get to run a free play due to the flag, the number of plays where that occurs may be small enough to discount the effect. Moreover, such information is often not listed in the play-by-play. You will not see, "incomplete pass on play nullified by offside penalty caused by QB hard count" in the play-by-play. You will only see "offside penalty (charged to player XY)". Thus, metrics like DYAR and DVOA which are collected from the play-by-play cannot measure those "hidden" effects. As a viewer, you can see them, but they aren't events being measured, so the stats cannot include them. Moreover, it is a judgment call as to what factors influenced a player jumping offsides, and those judgments are not included in the play-by-play.

So, DYAR does not measure all of the factors which makes a QB (or any player or team) good. It is however, an objective measurement based upon evidence that gives an approximation to the player's performance. It is in particular a precise metric calculated from events recorded in the play-by-play that have been found to be reproducible. You can find it useful or not. You may find that it conflicts with your subjective judgment. However, it isn't clear how to improve it without ruining some of its key attributes--its objectivity and its reproducibility. If you have suggestions that do that, you may be surprised how receptive people here are.

79
by Noah Arkadia :: Wed, 10/22/2014 - 12:57pm

If you want a site to tell you who's valuable but you expect value to be derived from scoring, why don't you just look at the official NFL page? They have all the scores and it should be all downhill from there.

------
Who, me?

47
by Vincent Verhei :: Tue, 10/21/2014 - 5:09pm

Here's the play-by-play. Masthay had two punts in the second quarter, one in the third.

DYAR totals do not include penalties, except for defensive pass interference. Plays that are wiped out by penalties, like the interception you've described, do not count at all.

DYAR examines the results of each play, not of each drive or each series. If a QB is sacked on second down and then converts on third-and-long, he had one good play and one bad play (although the good play will outweigh the bad play). If a different quarterback in the same situation just converts on second down and never gets to third down, he will have a higher DYAR, even though they have both produced a first down and have thus had similar results on their teams' drives.

On first down, a play is considered a success if it gains 45 percent of needed yards; on second down, a play needs to gain 60 percent of needed yards; on third or fourth down, only gaining a new first down is considered success. Here is a list of Rodgers' failed completions against Carolina:

8-yard gain on second-and-20
1-yard gain on second-and-4
4-yard gain on first-and-10
Zero-yard gain on second-and-6
5-yard gain on second-and-12
5-yard gain on third-and-12
2-yard gain on second-and-6

52
by oaktoon :: Tue, 10/21/2014 - 5:27pm

i stand corrected on the punts... Thanks for the detail.. My overall point still stands-- Rodgers was infinitely more "valuable" vs. Carolina than Andy Dalton was the previous week.... 38 pts in 3 Quarters without a turnover and only helped once by a defensive turnover stands on its own merits... listen, I had DYAR thrown in my face because Brady and Manning have had many higherDYAR seasons than Rodgers... Funny thing, though-- Rodgers has been winning at a .750 clip since his first 6-10 season... That's 12-4 seasons over and over again.... Which is about what Brady and Manning do (Brady a bit better, Manning a bit worse).... So I will stop by saying that DYAR doesn't get the full measure of this particular player-- and that's a weakness...

58
by theslothook :: Tue, 10/21/2014 - 5:53pm

oh boy...now you're bringing up qb wins as a supporting argument.

you know, ur dyar criticisms might be taken more seriously if your arguments weren't so agenda driven.

64
by dbostedo :: Tue, 10/21/2014 - 6:33pm

Team wins does not equal better QB. It's, in fact, a pretty poor way to compare quarterbacks. DYAR and DVOA are not perfect, but are much, much better.

65
by Bernie :: Tue, 10/21/2014 - 6:44pm

I've never understood why people get so butt hurt if other people don't agree with their assessment of their favourite player's performances. If you think Player X is the best player in the division / conference / nfl / of all time, then congratulations - he is!!!! If someone else disagrees and thinks it is someone else, then he's right too!!!! The correct answer of course, is who gives a shit. Enjoy the players you enjoy, hate the ones you hate, and who cares if no one else agrees with you. Don't be so insecure about your team or player loyalties, that you feel the irrational need to drag other people onto your bandwagon.

66
by Duff Soviet Union :: Tue, 10/21/2014 - 6:46pm

"Rodgers has been winning at a .750 clip since his first 6-10 season"

LOL. Not counting the games he's lost, Rodgers is undefeated!!!!111!

71
by oaktoon :: Tue, 10/21/2014 - 9:45pm

Oh please.... 2009-2014 vs the first season he had a chance to play..... hardly cherry-picking....

Listen i get a lot of this discussion-- trust me.. team vs. individual.. whether a player can repeat certain outcomes... even those who think i am being agenda-driven... This ain't my first rodeo...

but here's the thing... I think the NFL has, more than almost any sport save the NBA and its seemingly infinite garbage time, a lot of opportunities for players to amass "value" in situations where it doesn't really affect the most important outcome-- the team's 16 decisions each year-- and conversely, a lot of possibilities for a player to do things that achieve success in the latter but somehow don't stand up in the former... Any lifelong GB fan (I am not one) has the best example maybe of all time: Bart Starr. It couldn't have all been Lombardi, the line, the great backs-- right? I mean he made the QB sneak decision all by his lonesome to win the Ice Bowl...

So I think for a lot of reasons Rodgers is undervalued (and maybe the incomprehensible QB rating overvalues him-- he does have the best mark of all time). Some people attack him for being too cautious when the Packers are behind or bemoan his 4th quarter or comeback skills-- even though he actually is one of the best 4th quarter QBs in history and is, with Manning, the best in the business at bringing his team from behind if the deficit is modest. And they don't examine whether Rodgers' play produces 3rd quarter leads more often than a less cautious style thus making comebacks unnecessary. (And how cautious can one be who is generally at or near the top in yards/attempt?) Others criticize him for holding on to the ball too long, and, as was said here, bang him for taking sacks when a) he hasn't exactly had the best pass protection for the last several years and b) the by-product of holding the ball and taking some sacks is the lowest INT rate in the modern history of the game.... (and yes, maybe the injury last year was the product of him holding the ball a bit too long)

And yet.. his teams win (don't tell me the QB doesn't have an undue influence on that metric).. his offenses score... his play is smart (Carolina was a freakish game with four different times Rodgers caused or exploited a penalty from the opposition-- but believe me, he does it all the time)... and plenty of his peers or former peers believe he is, to quote John Lennon, at the toppermost of the poppermost.

So I therefore have a problem with a metric that treats games like Sunday's as just a fairly run-of-the-mill good game, when in fact it was sheer brilliance from beginning to end... He can and does recover from a sack or "failed completion" with a 25 yard play or other success more than most-- so why punish him in a metric for the failings of the other guys? And you know what?-- the guys that throw the 12 yard completion, followed by the 10 yard completion, followed by the 12 yard completion are usually going to get picked more often, too... It isn't stupid, heretical, or not worthwhile for someone to challenge these metrics-- take it as a compliment-- FO does things that go beyond the normal fare... But maybe like the search for the best baseball metric that measures defense... or the unified field theory (!)-- it can be even better.

72
by Will Allen :: Tue, 10/21/2014 - 10:02pm

Do the work required to flesh out the supposition you've made here, and I'm sure it will be extremely well received. Until that time, it is a bit, er, aggressive, to be flinging words like "bogus" at people who have done a lot more work than you. A little humility, from those who haven't spent the time to establish their arguments, goes a long ways. I've had some issues with individual player metrics through the years (although they work better for qbs than the other positions) , but I try to keep in mind that the perfect metric will never exist, that even imperfect metrics can tell us useful things, and having instituted one of the better imperfect metrics is something to respect.

73
by The Hypno-Toad :: Tue, 10/21/2014 - 10:28pm

For what it's worth, I don't read 107 DYAR in three quarters against a pretty weak defense as "just a fairly run-of-the-mill good game." That seems like a pretty thorough overcoming of two of the things that tend to negatively impact DYAR, limited opportunities and large, unhelpful opponent adjustments.

75
by dbostedo :: Wed, 10/22/2014 - 12:03am

First, thanks for a reasonable response. It seems like you're a thoughful fan, so I'd suggest you do the same thing you're asking FO to do - put a little more thought/work into things. If you don't understand what's being measured and why, then you're just arguing broadly that Rodgers isn't rated high enough, without picking out what's wrong with the way things are currently done.

Some other points/comments/rebuttals...

I think the NFL has, more than almost any sport save the NBA and its seemingly infinite garbage time, a lot of opportunities for players to amass "value" in situations where it doesn't really affect the most important outcome

It's been shown, and written about here many times, that "garbage time" contrary to popular wisdom, is predictive of future performance. Things get compared against similar positions and game situations and adjusted for defense, meaning it still matters how play goes when the game isn't close.

Some people attack him for being too cautious when the Packers are behind or bemoan his 4th quarter or comeback skills

I've actually never heard this. Any references? Could it be a bit of a straw-man setup?

...even though he actually is one of the best 4th quarter QBs in history and is, with Manning, the best in the business at bringing his team from behind if the deficit is modest

Well, he's one of the best period! Not just in the 4th. Trying to measure the best at comebacks has been done several times, and is fraught with problems of not having a good control and sample size issues. This kind of statement needs a lot of support - any references?

And they don't examine whether Rodgers' play produces 3rd quarter leads more often than a less cautious style thus making comebacks unnecessary.

Actually, around here over the years, I believe I've seen lots of comments talking about how comebacks lead to over-rating QBs, because it doesn't credit QBs who don't fall behind in the first place enough. That was a big critcism of QBRs clutch factors originally.

...as was said here, bang him for taking sacks when a) he hasn't exactly had the best pass protection for the last several years

Again, many times they've shown that the QB is typically more responsible for sacks than most people would expect. Is it possible Rodgers is an outlier? Sure, but it's going to need more than an eye test or assumptions.

And yet.. his teams win (don't tell me the QB doesn't have an undue influence on that metric)...

QBs are acknowledged as the most important position. But as you said, his TEAM wins. Should he get credit? Sure. But a win shouldn't blind you to trying to measure how the QB actually played adjusted for the defense he played.

So I therefore have a problem with a metric that treats games like Sunday's as just a fairly run-of-the-mill good game, when in fact it was sheer brilliance from beginning to end...

You're assuming what you saw was an amazing performance. Perhaps the metric here is telling you that it was excellent, but had its faults as well. The error is as likely to be with your assessment as it is to be with the metric.

He can and does recover from a sack or "failed completion" with a 25 yard play or other success more than most

THAT is a huge assumption that needs some real numbers to back it up.

And you know what?-- the guys that throw the 12 yard completion, followed by the 10 yard completion, followed by the 12 yard completion are usually going to get picked more often, too...

This one doesn't make sense. Are you just saying that most QBs get picked more than Rodgers? If so, he already gets credit for it. The 12/10/12 QB - all else equal - is still more valuable in the long run.

It isn't stupid, heretical, or not worthwhile for someone to challenge these metrics

I, and probably most others here including the site owners, agree wholeheartedly. They constantly try to add new things to make the metrics better and more predictive. There have been many articles and mentions here over the years of things that have been tried. The problem is, you're arguing for things that might make some intuitive sense, but have no statistical backing. "My visual assessment doesn't match the metric, therefore the metric is wrong" is not a good argument. That's the same issue that gives rise to "common wisdom" types of arguments that are sometimes shown to be wrong. As was stated before, the work needs to be put in. I'm sure the site owners would love a contributed article actually looking at some of this in a statistical or formal analysis way.

74
by theslothook :: Tue, 10/21/2014 - 10:44pm

Have you done any real statistical analysis to backup your claims? You seem to making assumptions about things like A) A qb sack is 100 percent not the fault of the qb. B) a qb who manages to recover from a sack and an incomplete pass is somehow a better predictor of future success than one who completes passes and avoids incompletions and sacks.

Dvoa and dyar have both been very kind to Rodgers in the past, so your griping about the stat somehow underrating Arod is very strange.

60
by coremill :: Tue, 10/21/2014 - 6:10pm

"In this case, that actually raises the baseline. RBs this year are averaging 4.1 yards per carry with a 47 percent Success Rate. QBs are averaging 6.2 yards per carry with a 62 percent Success Rate."

Isn't there a huge selection bias for QBs in terms of who runs? A disproportionately large proportion of QB rushing attempts are going to come from the good rushing QBs. The top 3 in QB rushing attempts are, not surprisingly, Kaepernick, Newton, and Wilson, with between them 125 attempts for 777 yards (6.2 y/a); the three of them account for 19% of all QB runs.

Also, for DYAR purposes shouldn't the average be irrelevant? Average and replacement level aren't the same thing. The average is affected by the presence of high-performing outliers and the replacement level should not be.

67
by Vincent Verhei :: Tue, 10/21/2014 - 7:29pm

Isn't there a huge selection bias for QBs in terms of who runs? A disproportionately large proportion of QB rushing attempts are going to come from the good rushing QBs. The top 3 in QB rushing attempts are, not surprisingly, Kaepernick, Newton, and Wilson, with between them 125 attempts for 777 yards (6.2 y/a); the three of them account for 19% of all QB runs.

Well, yes, but why is that a bad thing? If we want to know what happens when quarterbacks run, shouldn't we start with the quarterbacks who actually do run? Shouldn't Kaepernick and Wilson count more than the Mannings?

Also, for DYAR purposes shouldn't the average be irrelevant? Average and replacement level aren't the same thing. The average is affected by the presence of high-performing outliers and the replacement level should not be.

Replacement level is determined by finding a percentage of the average. So yes, if the average changes, replacement level changes too.

78
by coremill :: Wed, 10/22/2014 - 11:02am

Consider a hypo in which no QB EVER rushed except for the Kaepernick/Wilson/Newton trio. The average rush would be extremely high. But in this world the replacement level rushing value baseline should be zero, because anyone could step in and provide the same rushing value as the 32nd best (or 4th best) QB. Taking replacement level as a percentage of the average rush would give you a distorted view of the baseline.

Nate Silver made a similar observation about shortstop replacement levels in 2006, when he pointed that using a comparison to league average would be skewed when the performance in the data set exhibits a feast-or-famine distribution. http://www.baseballprospectus.com/article.php?articleid=4891

80
by Vincent Verhei :: Wed, 10/22/2014 - 2:36pm

OK, that makes sense. Just for fun, let's see if that's going on with rushing quarterbacks in the NFL. Not counting kneeldowns, there have been 458 QB rushes this year. About half of those (50.7%, to be precise) are by the nine guys with at least 15 rushes: Kaepernick, Newton, Wilson, Geno Smith, Luck, Alex Smith, Bortles (!), Tannehill, and Fitzpatrick. Those nine guys have a combined 1,538 yards on 232 carries, an average of 6.6 yards per rush.

The rest of the league -- guys with 14 carries or fewer -- have a combined 1,300 yards on 226 carries, an average of 5.8 yards per rush.

So yes, the guys who run a lot do seem to be skewing the data a little bit, but I doubt it's making a huge difference.

81
by coremill :: Wed, 10/22/2014 - 4:20pm

It would of course be more helpful to use success points/attempt (especially since QB sneaks are a meaningful subset of QB runs, and Y/A doesn't tell you much about those), but that almost half a yard per attempt seems like a meaningful effect to me.

There's more going on here, too. Not only do the poor runners run less often, but they are likely to run only in the most obviously optimal situations, which is going to artificially boost their results. The best runners will run in more marginal situations, which will drag down their per play averages.

There's probably also an effect where the good runners convert some sacks into unsuccessful runs because they manage to scramble back to the line of scrimmage and gain a yard or two, while the bad runners just get sacked. This would drag down their rushing results but boost their passing results, since they wouldn't have as many sacks counting against their passing.

82
by LionInAZ :: Wed, 10/22/2014 - 6:06pm

Your whole argument is so convoluted as to fail Occam's Razor. Isn't it smart for poor runners to avoid running in situations where they're likely to fail? Isn't it bad decision-making when good runners run when they can make a more successful play?

This site is about value to team performance, not about whether one player is better than another.

84
by coremill :: Wed, 10/22/2014 - 6:27pm

I don't see what your objection is. What I'm talking about is how calculate value to team performance. My issue is about using average performance as the baseline for determining replacement. My point is that, for various reasons, the average QB rush is not reflective of how well an average QB actually runs, because there are various biases in the data. As a result, if you don't attempt to control for those biases, you'll end up with a QB-rush replacement level that's too high, and you'll understate the actual value of good running QBs.

92
by LionInAZ :: Thu, 10/23/2014 - 7:29pm

It was your statement about 'poor QB runners not running as much, and only when optimal'. That is actually a positive. A bad QB would be running when he shouldn't. A poor running QB who only runs when good opportunities present themselves should not be punished. DYAR is not about QB running ability, it's about the value they add when they actually do it.

83
by intel_chris :: Wed, 10/22/2014 - 6:10pm

How are sacks counted? Are they failed runs or failed passes (or both)? I doubt that the play-by-play tells you whether the QB was attempting to "run" when they got sacked.

85
by Vincent Verhei :: Wed, 10/22/2014 - 7:45pm

Sacks are passing plays. There are some plays where it's hard to tell whether a quarterback was sacked, or tackled for a loss on a designed run; we're pretty much at the mercy of the play-by-play data on those.

86
by Karl Cuba :: Thu, 10/23/2014 - 11:37am

Is there a problem caused by lumping all qb runs together? The 'running' qbs are often called to carry the ball on designed runs/read options nowadays, while most traditional passers mainly carry the ball on scrambles from a called passing play. Would things look different if called and uncalled runs were treated as distinct?

(Plus, I like the write up on McKinnon but the gifs are crippling my tablet thing. Is there a way of putting them up so they don't auto-play?)

87
by intel_chris :: Thu, 10/23/2014 - 2:46pm

I think it is the opposite. If sacks are not counted as runs, then QBs don't suffer "stuffs" the way RBs do, which I suspect would bring down the average gain by quite a bit, since even (especially) a non-running QB can suffer a sack. The question is whether this better states the value of a running QB. The argument being that a running QB might have run (scrambled) for a gain instead of getting sacked, while the non-running QB has to take the sack.

It shouldn't matter to total DYAR whether we account the sacks as running or passing plays, but it would mean that the 5 least valuable "running QBs" would probably be non-runners who got sacked for the most lost yardage.

89
by Karl Cuba :: Thu, 10/23/2014 - 5:39pm

I can kind of see where you're coming from but that would be an even larger shift then I was suggesting.

90
by coremill :: Thu, 10/23/2014 - 6:12pm

It might matter to total DYAR if the baselines change depending on whether a play is categorized as a "run" or a "pass."

88
by Vincent Verhei :: Thu, 10/23/2014 - 3:33pm

(Plus, I like the write up on McKinnon but the gifs are crippling my tablet thing. Is there a way of putting them up so they don't auto-play?)

I'll see if there's something I can do about that. I'm kind of learning as I go here.

91
by LionInAZ :: Thu, 10/23/2014 - 7:19pm

I'll second Karl's comment about the animated gifs. I like them, but I don't like them automatically running, all at the same time.

94
by Jerry :: Fri, 10/24/2014 - 4:41am

I'll take it a step further. Although GIFs often tax underpowered machines, they make sense in Word of Muth or Film Room. In a column like Quick Reads, though, where I'm coming to look at numbers for the whole league, it gets frustrating waiting for GIFs to load, especially if I'm coming back to look at comments.

93
by The Hypno-Toad :: Fri, 10/24/2014 - 1:31am

I don't know what kind of tablet you're running, but "Enable GIF Animation" is a checkbox option in Android's default browser. Settings>Advanced inside the app, if I remember correctly.

95
by Karl Cuba :: Fri, 10/24/2014 - 9:59am

It's a nexus but I can't find that option, thanks anyway though.

68
by intel_chris :: Tue, 10/21/2014 - 8:37pm

I asked how those failed completions rate in terms of DYAR...

This is something that is too complicated to get into here, but a quick data dump:

The average incompletion (not counting interceptions) in 2014 has been worth -7 DYAR.
The average failed completion has been worth -3 DYAR.
The average successful completion has been worth 12 DYAR.

So, can I infer that because Thomas was about 3 yards short of a successful completion on those plays, he was about 3 yards short of replacement level (and the -3 DYAR is a fair assessment of his performance)?

Now, that would suggest to me that about 50% of all 3rd down plays are successful. However, thinking about that a little harder, I suspect I'm neglecting the down-and-distance part, since the average 3rd down play is probably not 15 yards-to-go. Hmmm, so much I don't know....

69
by Vincent Verhei :: Tue, 10/21/2014 - 8:40pm

It's way, way, way more complicated than that. You can't just assume that a play would have been worth 1 more DYAR if it was 1 yard longer.

A zero-yard run on third-and-1 is a bad play. If you add 1 yard to that run, you're going to add a hell of a lot more than 1 DYAR.