Writers of Pro Football Prospectus 2008

Most Recent FO Features


» 2018 Free Agency Cost-Benefit Analysis

Is Kirk Cousins the best free-agent quarterback in recent memory? Should Trumaine Johnson or Malcolm Butler have gotten the larger contract? And what makes a free-agent contract good or bad, anyway?

20 Dec 2011

Week 15 Quick Reads

by Vince Verhei

With 7:47 to go in the fourth quarter on Sunday, Detroit Lions quarterback Matthew Stafford was sacked and fumbled. Oakland Raiders linebacker Aaron Curry recovered the ball for a touchdown, putting his team up 27-14 and seemingly knocking the Lions out of the playoff picture. With his team's season on the line, Stafford stepped up to save the day, throwing for 143 yards and a pair of touchdowns on the next two Lions drives to lead Detroit to a 28-27 victory. It was the third time this season Stafford has overcome a double-digit deficit to win. Has Stafford established himself as an elite quarterback? How does he stack up next to his NFC playoff peers?

For the season, Stafford now ranks ninth among quarterbacks in DYAR (Defense-adjusted Yards Above Replacement). That's the good news. The bad news is that five of the quarterbacks ranked above him are guys he'll likely have to beat to get to the Super Bowl. How does Stafford compare to those other quarterbacks in specific situations? We can answer that question using DVOA (Defense-adjusted Value Over Average), which measures value per play. More information on DYAR and DVOA can be found here, but in a nutshell: yardage matters, first downs matter, consistency matters, and avoiding turnovers matters a lot.

In alphabetical order, here are the six quarterbacks to whom we're comparing Stafford today:

  • Drew Brees, New Orleans Saints
  • Eli Manning, New York Giants
  • Aaron Rodgers, Green Bay Packers
  • Tony Romo, Dallas Cowboys
  • Matt Ryan, Atlanta Falcons
  • Alex Smith, San Francisco 49ers

While Tarvaris Jackson and the Seahawks or Caleb Hanie and the Bears still have mathematical hopes at reaching the playoffs, the NFC field will likely come from those seven teams (including Stafford and the Lions).

Stafford ranks sixth among that field in DVOA over the course of the season, but the beauty of DVOA is that we can separate it into a number of different categories. For starters, we can measure each passer's performance by quarter. Stafford is last among this group in first-quarter passing DVOA, fourth in the second quarter, and third in the third quarter. However, despite his late-game heroics, he ranks just six in the fourth quarters and overtime. Stafford's big finishes are not guaranteed, and the NFC's other top quarterbacks are also capable of clutch end-of-game play.

We can further break DVOA down by, er, down. Stafford ranks second among this group (behind Rodgers) in first down passing DVOA, but slips to sixth place on second downs and a distant seventh place on third downs. The lesson here is that the Lions are better off passing early in drives, rather than running to set up makeable third downs. (DVOA accounts for yards needed for a new set of downs, so Stafford's bad showing here doesn't mean that he was left in third-and-long situations.)

What about field position? At Football Outsiders, we break the field down into five 20-yard segments, starting with the red zone (you're probably familiar with this one), and then moving back to the front, mid, back, and deep zones. Stafford is first in this group in the front zone, which keeps Detroit from kicking too many long field goals or short punts. However, he's no better than fifth in any other zone. Stafford needs to play better by the goal-line, but also on his own side of the field.

And then there's the length of the passes. Stafford is third in DVOA on passes thrown behind the line of scrimmage. However, he's a distant seventh in passes thrown up to 15 yards downfield, and just fifth on passes thrown 16 yards or deeper. The Lions have a dangerous screen game, but Stafford falls short of his peers on medium and deep throws.

It's been a breakout year for Stafford, and a great year for Detroit. If the Lions are counting on their quarterback outplaying his opposition in the postseason, though, they're in for a rude awakening.

Drew Brees NO
Brees' game against the Vikings was the best game for any quarterback this season. Before opponent adjustments, it worked out to 346 YAR, but Brees is penalized for playing a flimsy Vikings defense that has now given up the most touchdown passes in the league (26) while tying for the fewest interceptions (six). Brees now leads the league with 2,159 passing DYAR on the season, ahead of both Tom Brady (2,022) and Aaron Rodgers (1,999). However, Brees also has more than 50 pass plays more than Brady, and nearly 100 more than Rodgers, and Rodgers is still the league's top quarterback in DVOA.
Philip Rivers SD
At halftime, San Diego was ahead 17-7 and the game was still in doubt. Rivers then did this in the third quarter: 5-of-7, 89 yards, with every completion gaining at least 11 yards and a first down, including a 28-yard touchdown. At the end of the third San Diego was ahead 31-7, and Rivers threw only one more pass after that.
Matt Stafford DET
Stafford won the game with two late touchdown drives, but his "clutch" performance (fourth quarter, within one score) wasn't as good as you'd expect: 7-of-12 for 96 yards, six first downs, one touchdown, plus a 17-yard DPI.
Carson Palmer OAK
Palmer's passing wasn't quite as great as his raw numbers look. He had 11 failed completions, tied with Matt Hasselbeck for most in the league. His third-down passing was particularly deceptive: he went 6-of-7 for 38 yards, but none of those completions actually picked up a first down. He also had one failed fourth-down pass. On fourth-and-1 from the Detroit 24 in the first quarter, Palmer threw incomplete to Denarius Moore into the end zone. Given the game's final margin, the Raiders probably regret that playcall.
Kyle Orton KC
Orton threw seven passes in the fourth quarter, each while protecting a one-score lead. He went 5-for-7 for 86 yards and four first downs. Of course, he may have been ahead by more than one score if he hadn't gone 1-of-6 for 8 yards and no first downs in the red zone.
Matt Ryan ATL
Atlanta defeated Jacksonville 41-14, and it easily could have been worse. Ryan threw six red zone passes. One was a 6-yard touchdown to Roddy White. Two were completed passes that failed to convert third downs, and three were incomplete.
Tony Romo DAL
Romo completed each of his first four passes. Then he threw one incompletion, then hit five in a row. He had some more muddled success, then hit seven in a row. This was all before halftime.
Michael Vick PHI
Vick on deep passes (more than 15 yards past the line of scrimmage): 5-of-7, 178 yards, one Hail Mary interception.
Tom Brady NE
New stat idea: In basketball, point guards get an assist for getting the ball into teammates' hands and putting them in position to make plays. So let's give quarterbacks an assist for any completion with at least 10 YAC. Brady had nine assists against Denver, and has 80 on the season. Nobody else has more than 67 (Matthew Stafford and Philip Rivers).
Alex Smith SF
Tarvaris Jackson SEA
Second-half numbers: 15-of-19 for 176 yards, nine first downs (including a touchdown), no sacks, no interceptions. At one point in the third quarter he completed four passes in a row for 33, 43, 21, and 10 yards.
Cam Newton CAR
Newton had six carries for 56 yards against Houston. Three of them were third-down conversions (with 10, 8, and 2 yards to go). One was a 13-yard gain on second-and-8. And one was an 11-yard gain on second-and-12. Third-quarter passing: 3-for-9, 6 yards, no successful plays. Fourth-quarter passing: 3-for-3, 57 yards, three first downs.
Aaron Rodgers GB
Not counting the Packers' bye week, Rodgers had ranked in the top 10 in quarterbacks every week of the year, and ten times in the top four. He finished 12th in Week 15, in perhaps the season's most surprising defeat. He gets a small boost over his raw numbers for playing the Chiefs' defense, which has been tougher than most against opposing quarterbacks. He gets a bigger boost from his rushing value -- his three carries resulted in a 5-yard gain on second-and-7, a 19-yard gain on second-and-4, and an 8-yard touchdown on third-and-3.
Matt Moore MIA
The entirety of Moore's value basically came on one three-play stretch over the first and second quarters: 30-yard completion to Charles Clay; 22-yard touchdown to Anthony Fasano; 31-yard completion to Brian Hartline. He actually had negative DYAR over the rest of the day.
Rex Grossman WAS
First quarter: 6-of-9 passing, 57 yards, but only three first downs, with two interceptions and a sack. His last pass of the quarter was a 7-yard gain on fourth-and-1, and it was more or less smooth sailing from there.
Jake Locker TEN
Hat tip to Tom Gower for pointing this out: At one point, Locker went 9-yard completion, 9-yard completion, 9-yard run, 9-yard completion. Then he was sacked for a loss of ...

(wait for it)

... 9 yards.
John Skelton ARI
Skelton's last eight passes of regulation: 3-of-8, 26 yards, no first downs. He still got credit for a fourth-quarter rally and an overtime win. Clutch, baby.
Tim Tebow DEN
Facing a fourth-and-17 down 18 points in the fourth quarter, Tebow was sacked for a 28-yard loss. That is, by far, the biggest mega-sack of the year, easily surpassing Cam Newton's 20-yard sack against Indianapolis.
Andy Dalton CIN
Dalton failed on his first three third-down plays. He then converted four in a row before failing on his final try.
Seneca Wallace CLE
What? He's still in the league? Well, yes. At the start of the fourth quarter, Wallace and the Browns were ahead 17-7. From that point on, Wallace went 4-of-10 for 38 yards, only one first down, plus two sacks and a fumble, and Cleveland lost in overtime. Earth to Mike Holmgren: Seneca Wallace is 31 years old and a 9-year pro. He's not going to get any better. Give up and move on.
Dan Orlovsky IND
You've got to actually see Orlovsky's play-by-play in its entirety to appreciate the pure nothingness of it all. It's fantasy football in the mind of Sartre. In the first half he went 8-of-14 for 48 yards with no sacks, no interceptions, and no first downs. His last play of the half would have gone for a first down, but Pierre Garcon fumbled the ball away. (Failed dreams of completion, indeed.) In the third quarter, he had three plays: a 7-yard gain on third-and-6; a 9-yard gain on first-and-10; and an 18-yard touchdown to Reggie Wayne. And then, perhaps having achieved fulfillment, he would pass no more. Keep in mind, the Colts won.
Kellen Clemens STL
Third downs: 5-of-7, 28 yards, two first downs, plus three sacks and a fumble.
Josh Freeman TB
First half: Two completions for 14 yards, two incompletions, two sacks, no first downs. And then, on his first play of the third quarter, he was sacked again. From that point he started picking up first downs and stuff, but by then Tampa Bay was down multiple touchdowns.
Ryan Fitzpatrick BUF
Fitzpatrick in his "back" zone (between his own 20 and his own 40): 11-of-17 for 84 yards, only two first downs, six failed completions, two sacks, and two interceptions.
Ben Roethlisberger PIT
Eli Manning NYG
For the season, Eli has 689 DYAR on deep passes, fifth behind Rodgers, Brees, Rivers, and Romo. Against Washington, though, he went 3-of-8 for 72 yards with three interceptions (and a 43-yard DPI) for -73 DYAR on deep balls.
Christian Ponder MIN
My God, there were some miserable quarterback performances this week. Ponder started 2-of-6 for 8 yards with no first downs and a sack. Then he threw a 10-yard touchdown to Toby Gerhart. His next 12 dropbacks produced three completions for 23 yards, five incompletions, three sacks, one intentional grounding call, and no first downs. That is one first down produced in 21 plays. And we haven't even gotten to Blaine Gabbert yet!
T.J. Yates HOU
Third down passing: 4-of-9 for 22 yards, two first downs, two interceptions.
Joe Flacco BAL
The Ravens were only down 17-7 at halftime. Flacco's first two passes of the third quarter were first downs to Torrey Smith and Ed Dickson. His next 15 dropbacks produced one first down (a 10-yard gain on second-and-5), three sacks, two interceptions, and one fumbled completion (which isn't Flacco's fault, but still).
Matt Hasselbeck TEN
First half: 8-of-16 for 78 yards, but only three first downs. The Titans were actually ahead 6-3 at halftime because, well, it's the Colts. Hasselbeck made a lot more big plays after that, but also had all three turnovers in the second half.
Blaine Gabbert JAC
Gabbert's first pass of the fourth quarter was an incompletion. To that point in the game, he had dropped back 13 times. One of those plays was a 21-yard gain on second-and-16. He had three other completions for a total of 1 yard. Meanwhile, he was sacked five times, fumbling twice. To repeat some mind-boggling info: In 13 dropbacks, Gabbert had one successful play, negative net yards, five sacks, and two fumbles. That's -152 DYAR. He then engaged in some major league stat-padding in the fourth quarter, when he was down by five touchdowns (and even then, he threw an interception). Gabbert now has -686 DYAR on the year. That would be among the bottom 10 seasons of all time, but David Carr's record of -1,113 with the expansion Texans in 2002 is safe.
Mark Sanchez NYJ
It probably wouldn't have given the Jets the win, but the final margin against the Eagles would have been a lot closer if Sanchez had played better in the red zone: 5-of-7 passing, 29 yards, two first downs, one touchdown, one sack, one interception. That touchdown came in the fourth quarter, and cut the lead to 26 points.
Caleb Hanie CHI
In 26 dropbacks against Seattle, Hanie had only six first downs. Meanwhile, he was sacked four times and threw three interceptions, including a pair of pick-sixes. His first-down performance was remarkably impotent: 1-of-6 passing with an interception, with his only completion going for -4 (that's minus-four) yards.

Five most valuable running backs
C.J. Spiller BUF
In the first seven months of his career, Spiller was a backup, and a consistently bad one. The former first-round draft pick posted a rushing DVOA of -10.5% in his rookie year, and precisely matched that figure in the first 14 weeks of 2011. He was a model of efficiency against Miami, though, with nine of his 12 carries gaining 5 yards or more. He finished with 12 carries for 91 yards and a touchdown. He was also thrown 11 passes, catching nine of them for 78 yards, including gains of 11, 14, and 19 yards.
Darren Sproles NO
Pierre Thomas NO
If you could combine Thomas and Sproles into one player — Pierren Sprolesmas, perhaps -- he would have been the most valuable running back of Week 15. As it was, Thomas and Sproles finished second and third in the weekly rankings. Thomas had just eight runs and only two pass targets, but six of those runs gained 5 or more yards, another was a 1-yard touchdown, and his two receptions gained 25 and 16 yards. Sproles wasn't nearly as steady on the ground -- six of his eight runs gained 3 yards or fewer, with the other two going for 6 and 14 yards -- but he caught five of the six passes thrown his way, and four of those receptions gained 9, 13, 18, and 39 yards. Their combined statline: 16 carries for 77 yards and a touchdown, with seven catches in eight targets for 120 yards.
Steven Jackson STL
Virtually all of Jackson's value came as a receiver. He caught nine-of-ten passes, including seven successful catches, five first downs, and four receptions of 10 or more yards.
Arian Foster HOU
Eleven of Foster's carries against Carolina were successful. He was stuffed for no gain or a loss just once, had five first downs (including a touchdown), and three runs of 15 yards or more. He also caught five of the six passes thrown his way, including gains of 22 and 25 yards.

Least valuable running back
Cedric Benson CIN
Benson's raw numbers (22 carries for 74 yards) may not earn him an invitation to Canton, but they hardly sound like the worst runner of the week, either. So why is Benson at the bottom of our rankings? Benson was stuffed for no gain or a loss nine times, but more importantly, he fumbled three times. The Bengals recovered all three of his fumbles, but our research has shown fumble recovery is more about luck than skill and so Benson is punished for putting the ball on the ground regardless of who eventually fell on it. Two of those fumbles came after Benson gained at least 10 yards, turning big-time positive plays into big-time negatives. He was also thrown two passes, catching one for 11 yards.

Five most valuable wide receivers and tight ends
Calvin Johnson DET
In addition to the numbers listed above, Johnson drew two defensive pass interference penalties for 14 and 17 yards. Seven of his receptions gained 17 or more yards. The other two gained only six yards apiece, but one was a touchdown and the other was a first down on third-and-4. The Lions threw him seven balls in the fourth quarter, and he caught four of them (plus the two DPIs) for 130 total yards.
Malcom Floyd SD
Five of Floyd's targets came on second down, one on third down. Each of his catches went for a first down, and each gained 12 or more yards, including a 28-yard touchdown.
Darrius Heyward-Bey OAK
Has the 2009 first-round draft pick finally shed the bust label? Heyward-Bey set career highs with eight catches and 155 yards against Detroit. Six of those catches gained first downs, and a seventh would have as well if Heyward-Bey had not fumbled at the end of a 29-yard gain.
Brent Celek PHI
Celek's final catch was a 6-yard gain on third-and-9. His first four catches went for 38 yards, 26 yards (and a touchdown), 13 yards, and then 73 yards.
Aaron Hernandez NE
Hernandez' big catches included gains of 16, 22, 25, and 46 yards. His short catches included two 8-yard gains on first-and-10 and a 1-yard touchdown.

Least valuable wide receiver or tight end
Damian Williams TEN
Dedicated readers may remember Williams' name from Week 11, when he was also the least valuable receiver in the league. Though he wasn't quite as bad against the Colts on Sunday as he was a month ago against Atlanta (one catch for 16 yards in 11 targets!), he was still virtually useless against one of the softest defenses in the league. Tennessee threw him six passes on Sunday. Two were caught, one for a 9-yard gain on first-and-10, the other for 6 yards on third-and-13. The other four fell incomplete.

Posted by: Vincent Verhei on 20 Dec 2011

224 comments, Last at 22 Dec 2011, 9:17pm by nat


by Will Allen :: Tue, 12/20/2011 - 11:16am

Weighted opponent adjustments may be something worth looking at. The Vikings defense has had problems all year, but right now they are playing with at least 4 defensive backs who are more vocationally suited to delivering packages for UPS or FedEx this holiday season, compared to playing a competent, or even incompetent, NFL qb, to say nothing of Drew Brees. I sincerely believe that Caleb Hanie might have a good day against them in two weeks.

by Athelas :: Tue, 12/20/2011 - 11:22am

I haven't watched much of the Vikings this year but could their secondary possibly be worse than that of the Patriots?

by Birdman84 (not verified) :: Tue, 12/20/2011 - 11:42am

Right now, yes. The Vikings' top four cornerbacks (Winfield, Griffin, Cook, Allen) have been injured or suspended. The safeties have also been injured and weren't good to start with. In fact, they're making Vikings fans nostalgic for Madieu Williams.

by Will Allen :: Tue, 12/20/2011 - 11:53am

If you had seen Tebow versus the Vikings secondary, compared to Tebow versus the Patriots seondary, you'd have your answer. Absent a sack in less than 2 seconds, they are completely helpless when facing the forward pass.

by akn :: Tue, 12/20/2011 - 6:28pm

Don't worry, Caleb Hanie won't be playing by the time Chicago rolls in to town. You might instead have to face the wrath of Josh McCown or Nathen Enderle.

by Mikey :: Tue, 12/20/2011 - 11:22am

Fitzpatrick seemed to rally a bit in not-quite-garbage-time. The game was in reach, but everyone in the stadium knew it wasn't going to happen. I'm guessing those plays really buoyed an awful performance.

by Vasilii :: Tue, 12/20/2011 - 11:23am

I'm not usually curious on specific things like these, but how did Reggie Bush not make it to the top-5 RBs? Stuffs? Short of 1sts? I'm curious.

by Mikey :: Tue, 12/20/2011 - 11:38am

Sounds like opponent adjustments. The Bills are not great run-stoppers, but even so, that seemed like a hell of a game.

by Eddo :: Tue, 12/20/2011 - 12:30pm

Didn't Bush also fumble?

by MJ (not verified) :: Tue, 12/20/2011 - 12:42pm

Yes, he lost a fumble. Probably fairly costly for him.

by MJ (not verified) :: Tue, 12/20/2011 - 11:29am

All anyone wants for Christmas is a Pierren Sprolesmas on their fantasy teams for Championship week.

by White Rose Duelist :: Tue, 12/20/2011 - 12:51pm

Merry Sprolesmas!

by Kevin from Philly :: Tue, 12/20/2011 - 3:53pm

Well, if they both have bad games, would the combination be Derrier Troles?

by IsraelP (not verified) :: Tue, 12/20/2011 - 11:34am

Roethlisberger > Flacco. There is some consolation there.

by MJ (not verified) :: Tue, 12/20/2011 - 11:49am

A Roethlisberger with playing with a high ankle sprain is better than Flacco playing fully healthy, no less...
I do feel sort of sorry for the Ravens. Flacco isn't terrible enough to just dump him, but he really can barely even play the part of a competent game manager at times, and he hardly ever rises much above that.

by GlennW :: Tue, 12/20/2011 - 12:47pm

Strangely, over his short career Flacco has played some of his better games against the Steelers (considering the quality of the Steelers' defense). I suppose this is just a coincidence resulting from his inconsistency, but as such I don't quite look at Flacco as some easy pushover. He can be dangerous or he can hand the game over; you just don't know.

by MJ (not verified) :: Tue, 12/20/2011 - 1:39pm

Either Mike Kurtz or Tom Gower referred to Flacco as a Jekyll/Hyde.

by tgt2 (not verified) :: Tue, 12/20/2011 - 12:55pm

This sounds like the description of a player ranking 14th in DVOA and 12th in DYAR, right? Granted, he's going to fall to maybe 16th this week, but that's not Trent Dilfer territory.

by DGL :: Tue, 12/20/2011 - 2:15pm

Not much.

by Julio (not verified) :: Tue, 12/20/2011 - 11:35am

How does a QB with fewer yards,
fewer TD's and more int's (Vick)
than Brady end up rated higher?
Please don't tell me the Jets
defense is so much better than
the Broncos. This is like Gronkowsi
last week being ranked 5th instead
of 1st or 2nd.

by navin :: Tue, 12/20/2011 - 11:43am

The Jets defense is so much better than the Broncos defense.


by Julio (not verified) :: Tue, 12/20/2011 - 1:33pm

And that's what happens when you use a
flawed rating system to calculate another
flawed rating.

Here are the numbers versus common opponents:

denver gave up 10 to KC Jets gave up 10
denver gave up 41 to NE Jets gave up 30 and 37
denver gave up 13 to SD Jets gave up 21
denver gave up 15 to Mi Jets gave up 6
denver gave up 23 to oak Jets gave up 34

and in the game they played each other, Denver
gave up fewer points than the Jets!
Obviously the Denver defense is about the same
or better than the Jets defense.

by Eddo :: Tue, 12/20/2011 - 1:36pm

Actually, when they played each other, the Jets' offense gave up seven of those points.

by tuluse :: Tue, 12/20/2011 - 1:41pm

The Patriots and Chargers also both scored on turnover returns, but obviously this rating is not flawed.

by RickD :: Tue, 12/20/2011 - 3:06pm

Obviously it's not flawed?

I don't think even Aaron would go that far.

My vague impression is that over the past month and a half, the Broncos defense has been comparable to the Jets. (Certainly their record has been better!)

by tuluse :: Tue, 12/20/2011 - 3:20pm

I meant Julio's points allowed rating.

by zlionsfan :: Tue, 12/20/2011 - 3:06pm

If points allowed is the metric you use to determine defensive performance, I doubt there's any reason for you to read these articles.

by RickD :: Tue, 12/20/2011 - 3:10pm

Points allowed is a fairly important statistic. I wouldn't be so dismissive of it.

You guys might do better than simply asserting that DVOA is better. The question is reasonable. Why does FO think Vick passed better than Brady did on Sunday? Is the answer entirely due to opponent adjustments or is there more to it than that?

For example, you could say that Brady is being penalized for being sacked. The hit by Dumervil was pretty convincing. Now you'd be hard pressed to blame Brady personally for the fact that Dumervil came in completely unblocked. But it's the kind of fact that would add something to the discussion.

by tuluse :: Tue, 12/20/2011 - 3:23pm

Aren't a lot of QBs at least somewhat responsible for calling out protections? Aren't QBs supposed to read the defense and if a defender is unblocked hit a hot route?

by MPavao (not verified) :: Tue, 12/20/2011 - 11:44am

I'm pretty sure that's exactly why Vick is higher. The Jets secondary is very, very good. The Broncos secondary is not.

by Birdman84 (not verified) :: Tue, 12/20/2011 - 11:50am

Well, a quick look at the numbers shows that Vick had a higher completion percentage (75% versus 70%) and much higher yards per attempt (13.7 versus 9.7). Vick only had 9 more passing DYAR than Brady. I see no reason to be surprised by this.

by UConnymous (not verified) :: Tue, 12/20/2011 - 11:54am

Vick had fewer incompletions and was 32 rushing DYAR better. Or maybe it's just this site's obvious bias against the Pats! ;-)

by Led :: Tue, 12/20/2011 - 1:01pm

Vick's INT was a hail mary on the last play of the 1st half. It was the equivalent of an incompletion where the baseline must be very low. I doubt DVOA attributes much negative value to it.

by chemical burn :: Tue, 12/20/2011 - 4:14pm

I feel like this is a case where it was a such a stupid, pointless, poorly executed hail mary that he SHOULD have been penalized. For a second, I thought "god, that was such a terrible under-thrown ball, the Jets are going to return it for a TD!"

by Illmatic74 :: Tue, 12/20/2011 - 11:54am

I am just wondering what the similarity scores would be for Matthew Stafford?

by zlionsfan :: Tue, 12/20/2011 - 3:08pm

I'd be interested to know for 2011. For 2010 (or 2009-10), there is no best-case scenario among the players listed.

by andrew :: Tue, 12/20/2011 - 11:57am

Jake goes for 9, 9, 9, 9, -9?

I would point out who Jake Locker must be a fan of but that would violate rule#1.

by White Rose Duelist :: Tue, 12/20/2011 - 12:56pm

Nine, nine, nine
That crazy number nine
It's perfectly consistent
It works oute very time

by RickD :: Tue, 12/20/2011 - 3:11pm

Jack Locker is a fan of Fight Club?

by Kevin from Philly :: Tue, 12/20/2011 - 3:56pm

No, the Beatles white album.

by dryheat :: Tue, 12/20/2011 - 3:58pm

Nope. Magic Hat Brewery.

by Anon (not verified) :: Tue, 12/20/2011 - 12:13pm

It's a shame that for a second time Drew Brees will finish the season with a record-breaking performance, a spot in the playoffs, and no MVP award. In 2009 he broke the completion percentage record (70.6%, held since 1982 by Ken Anderson), had 30+ TD and 4500+ yds and yet he was snubbed in favour of Manning, who trailed him in both cathegories, but was crowned for his suppossed comeback performances earlier in the year.

This season he's on pace to break his own record for completion percentage, obliterate Marino's passing yards record, and yet it's almost a given the press is going to crown Rodgers as MVP.

I'm not saying Rodgers doesn't deserve it like Manning, I'm just pointing out how unlucky Brees has been, and how unfair it is to compare QBs starting with awards and All-Pro votes. If I had to pick the best QB in the last 5 years, it's most definetly Brees, by a landslide. And yet, all he has to show for it are Pro Bowl berths. I wonder how he's going to be remembered in the future, given his lack of big trophies compared to his contemporaries (Rodgers, Brady, Manning, Rivers, etc.)

by Anon (not verified) :: Tue, 12/20/2011 - 12:15pm

Correction, in 2009 he had 4300+ yds passing. Manning had more passing yards than him, though also more INTs and a bigger INT%, smaller YPA and TD%.

by AnonymousA (not verified) :: Tue, 12/20/2011 - 12:22pm

I think you're looking for ESPN. This is not the sort of site where a QB with 100 extra attempts gets a lot of credit for having more yards.

by Anon (not verified) :: Tue, 12/20/2011 - 10:46pm

I used yards as an example of the kind of stats used by the people that vote for the MVP, following their logic. Besides, that year Manning passed more than Brees (591 passes against 538). If I were to argue with the FO commentariat, I'd point at Brees' DVOA was better than Manning's (true, Manning was the No.1 QB that year in DYAR and YAR, but he also led the league on passes attempted).

by tuluse :: Tue, 12/20/2011 - 12:22pm

"If I had to pick the best QB in the last 5 years, it's most definetly Brees, by a landslide"

If you say so. I would say Peyton in landslide, followed by Brady, and then Brees.

by BSR :: Tue, 12/20/2011 - 12:39pm

The last five years I would say Brady although it would be close. The five years prior to that definitely Manning.

by JIPanick :: Tue, 12/20/2011 - 1:28pm

I would say Brady and not close. Manning is second, Brees a somewhat distant third, Rivers fourth.

Just my opinion, of course.

by MJ (not verified) :: Tue, 12/20/2011 - 1:42pm

What thread would be complete without a Manning/Brady debate?!

by zlionsfan :: Tue, 12/20/2011 - 3:09pm

all of them. There's a special thread for it for a reason.

by dryheat :: Tue, 12/20/2011 - 1:44pm

My opinion mirrors yours, with Roethlisberger clearly 5th and then......Eli? Romo? Three years of Kurt Warner? The list seems to fall quickly into the "not elite, but good enough to win with" category. From 2007 to 2011, it's hard to form a coherent argument for anybody except Brady...unless you want to ding him on 2008.

by tuluse :: Tue, 12/20/2011 - 1:52pm

"it's hard to form a coherent argument for anybody except Brady"

Must not reply, must not reply... shit, can't help myself.

I think Peyton has done more with less than any QB in recent memory. I'm sure Will Allen would say that Tarkenton is the all time king in this category, but Peyton has to come in 2nd. Dragging that 2010 team into the playoff should basically be seen as a super human feat.

Was that coherent?

by dryheat :: Tue, 12/20/2011 - 2:06pm

Well, Brady's 2007, 2010, and 2011 seasons are far better than anything that Manning's done since then....so no?

I mean, not to be argumentative, but I think Brees-backers might have the better argument. I think the gap from Manning to Brees is fairly slight.

I'm not sure what this "Doing more with less" myth comes from. Manning has had Reggie Wayne and Dallas Clark for the entire stretch. The Patriots have never had a RB as good as Addai in that time. The Colts defense is built to perform with a lead, and the Colts OL is suited for Manning. Manning goes down and Coltersky replaces him and the whole team looks much worse, obviously. The defense never gets to play with the lead. Coltersky doesn't know how to handle a pass rush or hit the open man. The running game has to deal with an 8th man in the box. It doesn't mean the surrounding talent sucks...it means they were constructed to complement QB1 and can't perform as well with the quarterback equivalent of Ishtar at the helm.

A guy like Kyle Orton has this team at ~5-9. Manning has them at ~8-6.

by tuluse :: Tue, 12/20/2011 - 2:27pm

I guess we have different definitions of coherence then.

by GlennW :: Tue, 12/20/2011 - 2:44pm

Unfortunately using similar logic, the Matt Cassel 11-5 season where the Patriots still put up 25+ points/game (yeah, I know, weak schedule) doesn't work in Brady's favor. It doesn't badly hurt him, but it certainly doesn't help his case.

by dryheat :: Tue, 12/20/2011 - 3:10pm

Why not? The team went from 16-0 to 11-5. I mean, the sample size couldn't be smaller, but the Patriots won 5 fewer games with Matt Cassell with the same supporting cast (in addition to a weaker schedule).

Brady worth 5 wins (and however many more PPG) more than Cassel passes the sniff test.

by GlennW :: Tue, 12/20/2011 - 3:33pm

And then in 2009 Brady returned and the Patriots finished 10-6 while scoring all of 17 more points than they did with Cassel. The delta argument works on both sides of the lost season.

But regardless, even from the historic 16-0 to 11-5 the impact wasn't close to what the Colts are experiencing. And yes, I fully understand that Cassel is better than Painter/Orlovsky, but we're still talking about a massive performance delta, even as Cassel was posting a passer rating of 89.4 in spite of having NO prior NFL or collegiate experience. Pretty obviously that's indicative of a nice little offensive system in its own right, which is taking nothing away from Brady (who makes the system better-- much better even). It just highlights the problem with looking at straight DYAR and attributing all of it to the QB as if in a vacuum, and then declaring him "hands down" the best QB of the era or whatnot.

by dryheat :: Tue, 12/20/2011 - 3:54pm

Yes, yes...understood. Cassel had no starting experience, but it was his fourth year as primary backup to Brady, and he must have been showing something in practices and preseasons, because Belichick was comfortable not having a #3 QB on the roster for much of that time. Still, I agree with everything you say here, except I'd say the Patriots scaled back the offense to accomodate him and played to his strengths, rather than they use a system that you can plug average QBs in and have them perform like great QBs.

The only anointing I'm doing is of Brady being, by a significant margin, the best QB in the NFL from 2007 to 2011, which was the time frame the initial poster used in support of Brees. I think that is borne out by pretty much any reasonable measure that one would like to use, unless one of the criteria is "games played" or "playoff wins". If the question was who was the best QB in the NFL from 2003-2008, 2001-2004, or 2009-2011, the answer might be different.

by GlennW :: Tue, 12/20/2011 - 4:02pm

I *think* you're saying that Brady was the best QB over that period based solely on his statistics, absent any further analysis of context within a team game. I wouldn't disagree with that. But if you're saying that Brady was observably and demonstrably the best QB ("by a significant margin") independent of context or circumstances, yeah, I'd take issue with that. I'd have a hard time separating Brady from Manning or even Brees in that regard. (Seriously, I've always been an agnostic on the Brady/Manning debate-- it's a tough call.)

by BSR :: Tue, 12/20/2011 - 4:03pm

Its not really indicative of much as you have to take each year on its own. There are far too many factors that change from year to year to really make rhyme or reason from these types of delta arguments. Players come, players go, players improve, players regress, players get injured, players get healthy, injures, schedules, weather, etc.. Its an exercise in futility trying to put all the different factors together.

The real ironic thing about this argument, is that the irrational thread is littered with the supporting cast argument but in favor to Brady. I think that was the number one argument of Brady supporters, at least that and not playing in a dome. But at the end of the day, you can really only measure a player based on what he accomplished, not what he might have done if he had better players around him. Both have accomplished a hell of a lot. Trying to look at it objectively, its hard to say that Manning never had good players around him. Same is true for Brady, although 2006 was incredibly low on talent. This is why I say the comparison is pretty close. I give it to Brady because overall his numbers were better during the time.

by milo :: Tue, 12/20/2011 - 4:40pm

It's the dome, it's the dome.

Brees 2011: 370 ypg outside. 326 ypg inside.

Oh, wait..

by GlennW :: Tue, 12/20/2011 - 6:22pm

> Its not really indicative of much as you have to take each year on its own.

True, but there's at least enough evidence there that no one should ever say "you know, those were pretty average Patriots teams if it weren't for Brady". I'm less sure of the Colts' situation minus Manning, and only partly because of their abject collapse this season. Admittedly some of this is subjective-- Peyton Manning more consistently made throws down the field that I don't see from anybody else, not from Brady, not from Brees. On the hand Brady (or Brees) may be more adept at making percentage calculations from his reads, and delivering the ball to the right receiver with a minimum of risk. I don't which is more valuable on the final balance sheet-- I'm not that smart.

by BSR :: Tue, 12/20/2011 - 6:56pm

I don't know if I've seen and compared enough film from each to say definitively who throws what better. As you said it is subjective, and I think much of those are anecdotal rather than a real comparative study. Plus I think the differences at that level are small.

And I think there are a few teams that were fairly average without Brady. The 2005 and 2006 teams were mediocre, especially the 2006 team. If you put Painter et al on this current Patriots team and had Caldwell coach they might lose every game by 50 points. This Patriots team probably has the least depth of any that I can think of in the Belichick era. They have less than average quality players starting in several positions, particularly on D. They are less then pretty average without Brady.

by Nathan :: Tue, 12/20/2011 - 4:04pm

And then in 2009 Brady returned and the Patriots finished 10-6 while scoring all of 17 more points than they did with Cassel.

With a freshly rehabbed knee. I don't know if you're a Pats fan, but in 2009 it was obvious to anyone who watches all his snaps that Brady had some serious trepidation about stepping into throws. This observation is substantiated by the NFL Films "Football Life" doc on Belichick BTW, Belichick mentions it to Brady a couple times.

by RickD :: Tue, 12/20/2011 - 3:14pm

The problem with picking "the last 5 years" is that both Brady and Manning missed entire seasons over that period.

I think it's ridiculous to pick any of the three in "a landslide."

by Danish Denver-Fan :: Tue, 12/20/2011 - 12:29pm

The last five years? Counting 2011...

Brady total DYAR (four seasons): 8937
Brees total DYAR (five seasons): 8301

Not. Even. Close.

by Mr Shush :: Tue, 12/20/2011 - 12:30pm

Production's not everything either. I'd rank QB performance over the last five years (when healthy) Manning-Brees-Brady-Rivers (let's leave Rodgers on one side because he's obviously playing at a very different level now to three years ago). This is because I think over that time frame Manning has had a significantly weaker offensive supporting cast than the other three (earlier and his and Brady's careers is of course a different story).

But I absolutely regard Brees as a truly great quarterback and first ballot Hall of Famer, and agree that it's a bit fluky he's never been league MVP.

by Otis Taylor89 :: Tue, 12/20/2011 - 1:25pm

I love me some Drew, but he's has had the total package working for him over that time span: a great line, pretty darn good receivers and TE's, a very good coach who's a great OC and plays in a dome.

The other power QB's have had some of that, but not all.

by BSR :: Tue, 12/20/2011 - 12:42pm

I don't think the MVP should be a forgone conclusion this year either. I could definitely see a case made for Brees and much should depend on these last couple of weeks especially since home field advantage isn't locked up between the two teams.

by Eddo :: Tue, 12/20/2011 - 12:51pm

The Packers haven't locked up home field advantage, but they have locked it up over the Saints (thanks to their week one head-to-head victory). The 49ers are the only team that can pass the Packers as of now.

by BSR :: Tue, 12/20/2011 - 12:55pm

True. The point is that at least for another week, all teams are continuing to play to win.

by The Powers That Be :: Tue, 12/20/2011 - 12:54pm

Yes it is. GB has the tiebreaker and is guaranteed a higher seed than NO.

by tgt2 (not verified) :: Tue, 12/20/2011 - 1:00pm

Does GB or SF have the 3-way tiebreaker over NO?

by Supadome (not verified) :: Tue, 12/20/2011 - 2:17pm

Hat tip to ESPN Playoff Machine:

"1st Seed - San Francisco
Wins tie break over Green Bay based on best win percentage in common games. Wins tie break over New Orleans based on best win percentage in conference games."

by Supadome (not verified) :: Tue, 12/20/2011 - 12:58pm

Well, in 2009 he could console himself with a SBMVP award, so maybe that'll happen again

by Joseph :: Tue, 12/20/2011 - 3:24pm

You beat me to it, Supadome.

Not only that, but he will forever be revered in NOLA. If he decided to run for any political office in Louisiana, he might get elected unopposed--as it would probably be futile for anyone to run against him.

by Supadome (not verified) :: Wed, 12/21/2011 - 1:22pm

I dunno, I hear there are C*wb*ys fans way up north in Shreveport and Monroe. Whole 'nother country up there.

Also: http://www.onionsportsnetwork.com/articles/drew-brees-casually-wonders-a...

by LionInAZ (not verified) :: Tue, 12/20/2011 - 8:00pm

I tend to agree. Drew Brees deserves as much consideration for MVP as Rodgers does.
The season's not over yet, so I wouldn't count him out yet.

by dryheat :: Tue, 12/20/2011 - 10:52pm

Well, the biggest thing Rodgers had going for him was being the quarterback of an undefeated team. I think the vote is going to be very very close.

by Arkaein :: Wed, 12/21/2011 - 10:11am

He also has more passing TDs, fewer INTs, and gains a full extra yard per passing attempt. Rodgers holds the same advantage over Brady in these same categories, with a small edge in YPA, but a higher completion percentage. And of course, Rodgers offers much more with his running ability than either of these other QBs.

Brees has essentially played 3 extra games with all of the passing attempts he has. Rodgers could play all of the last two games and would not have as many passing attempts as Brees has now. Brees has the higher completion percentage, but his lower YPA demonstrates that he is not throwing nearly as deep as Rodgers this season.

Barring a collapse in the last few games, there is no good reason why Brees would deserve the MVP over Rodgers. Though considering that Brees made the Pro Bowl over Rodgers last year despite 20 INTs, anything is possible, I suppose.

by Supadome (not verified) :: Wed, 12/21/2011 - 3:13pm

Not seeing what makes Rodgers such a clear choice. Rodgers leads by 3 TD and 5 INT (more than a few of which were dropped passes into opponent's hands). Brees is leading by 40 yards per game. So Brees is more productive, Rodgers is more efficient, and Brady is between the two.

This is going to be a 3 man race, with the difference probably being Brees breaking multiple records this year.

by Arkaein :: Wed, 12/21/2011 - 3:39pm

Ah yes, defenders drop Rodgers passes to keep his INT total low, while I'm sure every Brees INT was a phenomenal play.

Please, every QB has a few potential INTs dropped. Can Brees say that fully half of his actual INTs first touched his own players' hands before a defender was able to get a hold of the ball? Rodgers can, this season. Of 6 INTs, 3 either deflected off of his receivers hands or were pulled away from his receiver.

All in all, Rodgers has been every bit as good at avoiding INTs this season as his numbers say he has.

by Supadome :: Wed, 12/21/2011 - 5:34pm

I'll give you the INT argument. I know there have been a few, but I can't recall exactly, since it's been 6 weeks since the last one.

But my point is that we're looking hard into the microscope to separate these two QBs, who are both having phenomenal seasons.

by dryheat :: Wed, 12/21/2011 - 4:08pm

I think you're seriously over-estimating the thought that voters put into their vote. Do you think a guy like Peter King considers things like YPA or scrambling ability or passing attempts when casting their ballot?

It's a two-man race right now, with Brees trending up and Rodgers trending down. Considering Brees is probably going to shatter Marino's record, I'd put my money on him. If this commentariat each had votes, maybe I'd feel differently.

by GlennW :: Wed, 12/21/2011 - 5:23pm

Yeah, from MMQB Peter King's latest on the MVP race:

"I'd still give it to Rodgers this morning. The MVP should be his to lose at this point. But let's see how it plays out the next two weeks. If Brees sets records for accuracy and passing yards this year, and the Saints finish the year on an eight-game winning streak, it should be a contest."

Methinks PK is just itching to throw over to Brees once these records are broken. Undoubtedly he wouldn't do this if the Packers were to have finished 16-0.

by Arkaein :: Wed, 12/21/2011 - 5:56pm

You make some good points about the minds of the voters, but I would be seriously upset if Rodgers loses the MVP to Brees for the reasons you mention.

Looking at how two QBs are trending at the end of a season, when the QB trending down already has a first round bye locked up and is all but guaranteed to have the #1 seed for his team, vs. a QB whose team is fighting for a first round bye would be shameful. Peyton wasn't hurt in MVP voting when his team took the last few games off, so I'd hope the voters keep that in mind.

With regards to scrambling ability I think you're wrong, this gets plenty of coverage for Rodgers, and every coach and announcer knows that he's an above average runner for a QB. I doubt the voters ignore it.

As far as the yardage record, I don't know. Voters might weight it heavily, but the fact that there are four QBs with an outside chance at 5000 yards this season (Brees, Brady, Rogers, Eli Manning are all over 4300 yards at this point) makes it a bit less impressive. In any case, I'm not sure the voters put much weight on total yards. Brees' near record total in 2008 wasn't enough to push him past Peyton Manning, and in that year the only stat Brees significantly trailed Manning in was INTs.

by huston720 :: Tue, 12/20/2011 - 12:18pm

Picking on Holmgren for having Seneca Wallace as a backup QB is a little strange, even for a Seattle fan with a grudge. Wallace isn't that old, he is very familiar with the offense, he is cheap, and he basically performs at replacement level. Isn't that pretty much what you want in a backup QB, especially when your starter is a young QB in only his second year? It's not like Holmgren decided to start him because he thinks he is the future or anything. Plus the Browns do have a developmental 3rd string QB.

At least the Browns were smart enough to have a decent backup that isn't going to be among the leagure worst at theposition if forced to play like some teams. How is that working out for the Bears, Chiefs, Colts, Seahawks, Rams, etc.

by Anonymous1 (not verified) :: Tue, 12/20/2011 - 12:41pm

Agreed. I'd even contend that Cleveland would have beaten Pitt if Wallace stayed in and finished the game last week.

There is no reason to chide Holmgren about having Seneca as his backup.

by dryheat :: Tue, 12/20/2011 - 1:15pm

Thirded...I was going to call out that comment. He's a solid backup, and you can bet that Chicago, Houston, Indianapolis, and maybe another team or two wishes they employed Mr. Wallace as a backup.

Give up on him because he's 31 and unlikely to improve? That's just non-sensical logic for a back-up quarterback....especially for a team that has a ~23 year old starter who was talked about as a franchise quarterback two months ago.

by RickD :: Tue, 12/20/2011 - 3:16pm

Agree completely with this comment. Wallace was the QB who led the Browns into position to get a go-ahead TD. Then the Browns put in a QB who was suffering from a concussion, who promptly lost the game.

by jfsh :: Tue, 12/20/2011 - 12:22pm

Personally, I prefer "Darriere"

by Nathan :: Tue, 12/20/2011 - 12:25pm

Earth to Mike Holmgren: Seneca Wallace is 31 years old and a 9-year pro. He's not going to get any better. Give up and move on.

Seneca Wallace is a backup forced into service because the starting QB had a concussion. It's not like Holmgren is trying to develop him to be the QB of the future. Plus, there are way worse backups in the league. I really don't get this comment.

by Will Allen :: Tue, 12/20/2011 - 12:53pm

I think a great research project for next year might be an investigation of how the rules changes and changes to interpretation of rules, has had a huge impact on on putting greater emphasis on qb performance, which has had made starting caliber qbs ever more valuable. This, in turn, has had, due to the salary cap, perhaps had a real impact on how rosters are constructed, in terms of offensive line, total defense, and back up qbs, etc.. So much of the available money is devoted to the starting qb, necessarily so, given the importance of the forward pass, compared to previous eras, that the margin of error on the rest of the roster tends to become razor-thin.

I might be wrong, but it just seems to me that when the game increasingly revolves around the performance of the starting qb, teams are forced to take greater chances eleswhere. I don't think we are likley to ever see a team, for instance, with two high quality veteran qbs, like Unitas/Morrall, Griese/Morrall, Jurgensen/Kilmer. A Young can succeed a Montana, but Montana has to go, like Favre had to go. Or a team with a great qb has to take chances on the offensive line, like the Colts have done recently. The Packers this year are one of the great passing teams ever, but they certainly are not a great pass blocking team, even before recent injuries.

by tuluse :: Tue, 12/20/2011 - 12:54pm

I've been having similar thoughts. I think there has been an unintended consequence to the NFL's policy of changing rules to increase passing. Which is that they've magnified the difference between good and bad QBs rather than elevating all QBs.

The Bears in the 80s were able to keep winning with Steve Fuller in at QB. The Rams, in the same division with the 49ers, were able to make the playoffs 4 times in a row with 4 different QBs. Including someone named Deiter Brock. I don't think things like that can happen anymore.

by Will Allen :: Tue, 12/20/2011 - 1:02pm

Again, I try to avoid nostalgia, but I think I liked the game better when superior offensive line play, running and blocking, was a more viable path to a championship. Home team aside, my favorite NFL team since I was a kid was the Gibbs I era for the Redskins; a team whose first principal, for more than a decade, was that they would receive superlative offensive line play as a unit, and the quarterback would be protected, and the running game would be extremely good.

by Will Allen :: Tue, 12/20/2011 - 1:14pm

Yes, I am rooting hard for the 49ers in the playoffs.

by Mr Shush :: Tue, 12/20/2011 - 1:24pm

There's a team in the playoffs with a superb offensive line and running backs and a bad quarterback, but it's not the 49ers . . .

by Will Allen :: Tue, 12/20/2011 - 1:32pm

Them too, but I've always been more of an NFC guy, and it gives me indigestion to think about two of the offensive linemen on that team.

(edit) As for the other possibility, I haven't seen them much this year.

by Mr Shush :: Tue, 12/20/2011 - 10:17pm

Yeah, I was talking about the other possibility, but the Ravens too, I suppose.

by bravehoptoad :: Tue, 12/20/2011 - 1:52pm

I'm all excited about the 49ers breaking the single-season streak for not allowing a rushing touchdown. Marino's single-season yardage record has only stood for 27 years. Hah! That's nothing! The Chicago Bears (nee Decatur Staleys) have held this record for 91 years. How often do you see 91-year-old records being broken in the NFL?

Shoot, it might even be a new record for longest-held record broken.

by tuluse :: Tue, 12/20/2011 - 1:55pm

Considering that was the 2nd year of the league, the only way it isn't the longest held record is if there was a record set in the first year.

by DGL :: Tue, 12/20/2011 - 2:16pm

I'm sure there were a lot of records set in the first year. All of them, in fact. But I don't know if any of them remain unbroken :-).

by Aaron Brooks Go... :: Tue, 12/20/2011 - 2:08pm

They can break the Staley's consecutive games within a season streak. The single-season streak is different. The 1921 Akron Pros also allowed no rushing TDs in 1921 (and two fewer passing TDs). They only played 11 games, though.

by Karl Cuba :: Tue, 12/20/2011 - 2:41pm

I'm all excited about Aldon Smith, he was outstanding yesterday. Chaos in the backfield.

by Will Allen :: Tue, 12/20/2011 - 3:00pm

Anybody who saw last night's game, and claims the 49ers would not have a decent chance to beat the Packers, as the Packer's o-line is currently constructed, is partaking too much cheese.

by Aaron Brooks Go... :: Tue, 12/20/2011 - 3:48pm

Not every QB is as immobile as Roethlisberger With A Sprained Ankle.

Indeed, there are statues in Canton that are more mobile than that.

by Will Allen :: Tue, 12/20/2011 - 4:03pm

Never said they were. The 49ers still have a decent shance to beat the Packers, if they meet. Mobility didn't help Rodgers that much against the Chiefs, as an analogy.

by MCS :: Wed, 12/21/2011 - 11:04am

The Chiefs just proved that a team built like the 49ers can beat the Packers.

by MCS :: Wed, 12/21/2011 - 11:05am

Double Post

by Aaron Brooks Go... :: Tue, 12/20/2011 - 1:22pm

Rex Grossman, Superbowl QB.

by tuluse :: Tue, 12/20/2011 - 1:27pm

Yeah, the 05/06 Bears came close to what the Rams in the mid 80s did. However, 1) they couldn't keep it up, 2) Grossman is better than Deiter Brock, Jeff Kemp and possibly Vince Ferragamo.

by Will Allen :: Tue, 12/20/2011 - 1:27pm

Yeah, it's obviously possible to string together a couple playoff wins, but trying to win, year in, year out, without devoting a large chunk of your roster's salaries to the starting qb, compared to previous eras, seems to be increasingly problematic.

by bigtencrazy (not verified) :: Tue, 12/20/2011 - 1:32pm

As a mild defense of the Packers offensive line, the issue all season has centered on Newhouse.

Belaga has played really well.

The center Scott Wells has been very good

Sitton has not played to his 2010 level but he has been playing hurt.

TJ Lang has been solid.

If the Packers line were at 90% health the pass-blocking would be regarded among the best in the league

by Will Allen :: Tue, 12/20/2011 - 3:05pm

Yeah, I don't mean to dog them unfairly, but you are correct in implying that unit performance tends to track the weakest link in the chain, and obviously, with what happened Sunday, some links just got weaker.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 8:06pm

It's hard to say. A lot of Packer fans say that Colledge was the weak link on the O-line last year, but Colledge has played better with the Cards this year than he did for the Packers. Bulaga has obviously improved. It's not so clear that having Clifton and Lang together would have been that much of an improvement over last year, though.

by dmstorm22 :: Tue, 12/20/2011 - 3:13pm

Depending on your view of Sanchez and Flacco, the Jets and Ravens have won consistently (Baltimore for years with a number of QBs, with Dilfer, Grbac, Wright, McNair and Flacco all starting playoff games since 2000).

I agree with you though. There is too much importance on passing, because with all the rules passing is just easily the most effective way to play offense. Sadly, I don't think this will ever change, because there is no indication that this is a worse version of football. But it could become like basketball where the teams with the stars (the best QBs) are the only teams that can legitimately win a title.

by Will Allen :: Tue, 12/20/2011 - 3:59pm

Yeah, I'm not trying to make a subjective good vs. bad argument either, and obviously the NFL has spent a lot of time and money to determine what attracts eyeballs best. More than anything, I just miss a greater emphasis on really good offensive line play. It may be that the size and speed of defensive players precludes that type of play being so prominent again.

by dmstorm22 :: Tue, 12/20/2011 - 5:19pm

It also comes down to how much importance are you going to put on the postseason. Many teams can make the playoffs with mediocre quarterback play and great o-line play. Hell, the Texans are going to do it, and yes, they were great with Schaub, but even after he's gone, the running game's been doing just fine.

Great defense, great running, great anything really, can get you to the playoffs. I think we have reached a state where only great QB play can really win a title, but then again, Roethlisberger wasn't great in 2008.

If you are like me, and think that the postseason is way overvalued, and that there should be a lot more time spent enjoying the regular season (MLB does this really well), then you can see that these things can still work. The 2008 Panthers got a bye because they played solid defense and had one of the best running seasons I've ever seen, especially with that group of lineman (I've only watched football since 1999, and other than the 2005 Seahawks and 2002-2004 Cheifs, those Panthers were the best o-line unit I've seen). I think it is sad that it will be really hard for these teams to actually win a Super Bowl, but they can still bring a lot of happy Sundays from Sept-Dec. January is really just about having luck and a QB able to capitalize on that luck.

by Mr Shush :: Tue, 12/20/2011 - 10:23pm

I thought the 2008 Broncos were on that level. Sadly, we never got to see if they'd be able to maintain it.

Well, as a Texans fan I'm not that sad. Rick Dennison would still be in Colorado if we had.

by dmstorm22 :: Tue, 12/20/2011 - 10:26pm

I think they would have been waxed by Indy in round 1. I guess you never saw if they could maintain it in 2009, but at least for that year, I think they would've been one and done.

by Mr Shush :: Wed, 12/21/2011 - 11:06am

I was really just talking about the offensive line. I agree that the team wasn't that special (though the offense was - they were a bit like the 2010 Texans). But Clady et al., coached by Dennison in that system . . . they were awesome.

by nat :: Tue, 12/20/2011 - 12:34pm

Brady's rushing DYAR is interesting. Checking the play-by-play, he had a first down, a TD, and two fourth quarter stuffs from one yard out. I don't think the aborted snap counts as a run in DYAR.

So a success, a maximum-possible success, and two plays one yard short of maximum possible success = -12 DYAR. Hmmmmm.....

by Tom Gower :: Tue, 12/20/2011 - 12:56pm

Right now he's getting punished for the aborted snap, which is listed in the Gamebook as a rushing attempt and a fumble for him.

by nat :: Tue, 12/20/2011 - 1:26pm

Ah, that explains some of it. I thought aborted snaps were supposed to be treated differently. My bad.

by Tom Gower :: Tue, 12/20/2011 - 2:49pm

Ah, see Vince's comments below, he knows this stuff better than I do.

by White Rose Duelist :: Tue, 12/20/2011 - 1:01pm

Well, when "one yard short of maximum success" is zero yards, it can't really be considered a positive play. Remember, Brady is being compared to all goal to go from the 1 plays in silimar situations, and a lot of those are punched in.

One of the almanancs (IIRC) had a piece comparing players' gola-to-go TDs to the number an average player would score from the same distance, giving players a sort of plus/minus for these runs. Brady would definitely be minus for 1 TD on three rushes from the 1.

by Otis Taylor89 :: Tue, 12/20/2011 - 1:28pm

Brady's rushing attempts at the goal line should count as kneel downs as he has said on a radio show that they weren't trying to score, but take time off the clock.

by dryheat :: Tue, 12/20/2011 - 1:38pm

Yet I was pretty sure he scored on the second attempt. Didn't make sense to challenge though.

by Anonymouse (not verified) :: Tue, 12/20/2011 - 12:37pm

I'm not really a fan, but I think Tebow was far better than stats show this week. Because of the game situation, the 28-yard sack was no worse an outcome than a 17-yard gain would have been, so risking a big sack was worth it if there was any chance of escaping for a big gain/completion; it's the equivalent of an interception on an end-of-half/game hail Mary pass.
Similarly, an earlier play only goes as an incomplete (and possibly a fumble), when he actually escaped being tackled for a safety (average QB would have gone down), then had the presence of mind to not fall on a loose ball, but pick it up and make a credible attempt at a completion (semi-intentional grounding is probably the best description). At a minimum he saved 2 points and a turnover right there.

I know it's hard to adjust for 'should have been sacked but escaped', but surely a good stat can control for end-of-game desperate attempts that don't succeed (but don't really leave a team worse off, winning-wise)? If nothing else by manually dropping desperation plays? (Otherwise, we're particularly punishing QBs with bad teammates, who are more often in desperate situations at the ends of games).

by Eddo :: Tue, 12/20/2011 - 12:49pm

Two things:

1. You're trying to spin a very negative play (he fumbled in his own end zone!) into a positive play. It was great that he turned that bad situation into a slightly better one (a incompletion is still a negative play), but he wouldn't have had to do so had he not fumbled the ball in the first place.

2. I agree that the long sack should be treated as any other failed fourth-down play in that situation. I'm not sure it isn't, though; that's a good question for Aaron.

by Jerry :: Wed, 12/21/2011 - 7:51am

I'm reminded of Roethlisberger taking sacks late in the game against New England rather than throwing clock-stopping incompletions. It hurt Pittsburgh's DVOA, but was the right thing to do. Tebow's fourth down sack is the same thing - one play where DVOA isn't designed to understand all the context, and which will be a tiny part of the final totals.

by lester bangs (not verified) :: Tue, 12/20/2011 - 12:57pm

Stop trying to be funny, Vince. Know who you are.

by Anon (not verified) :: Tue, 12/20/2011 - 4:17pm

Come on! Orlovski is the existential qb. That's priceless.

by lester bangs (not verified) :: Tue, 12/20/2011 - 1:00pm

Re: Brees - The Vikings defense had allowed 26 TD passes before Sunday. It's now 31, obviously.

by jimm (not verified) :: Tue, 12/20/2011 - 1:15pm

given the state of the Vikings defence I would say Brees had a slightly above average game. Over the past nine games the vikings have not intercepted a pass. Have given up about 73% completion rate.

Brees 80% - not ints...

see just slightly above average

by Will Allen :: Tue, 12/20/2011 - 1:24pm

It's really a frightening spectacle to watch the Vikings defensive backs. I don't mean to hold the head coach responsible for a roster he had very little control over, but it is concerning that a head coach with a background in the defensive secondary would be coaching a group that looks so pathetic.

by jimm (not verified) :: Tue, 12/20/2011 - 2:34pm

It really is mind boggling how open receivers are - if a db is within 3 yards of a receiver it's a surprise.

The last time the Vikings had strong secondary was the 1987-88 team with Browner, Lee, and Holt. Since that time the Vikings haven't been good against the pass. Even the recent teams with the amazing run defence weren't particularly good against the pass.

I would really like to see them change from the Tampa 2 - for no other reason than I'm tired of watching the same thing over and over every year done so poorly.

Towards that end...I get the sense Everson Griffen might be a real star as the rush linebacker in the 3-4 defence...not sure how Allen would do, but my sense is he's a good enough athlete to do that as well.

by dmstorm22 :: Tue, 12/20/2011 - 3:16pm

It still stuns me how well the Vikings defense played Brees in the 2009 Title Game. As someone who watches the Vikings all the time, what has been the drop-off the last two years? Interior pass rush is the only noticeable difference, as it is no longer great. I guess the secondary is worse too, but it is amazing how much this defense has fallen.

by tuluse :: Tue, 12/20/2011 - 3:22pm

I don't think a single starter in the secondary from that game has played for the Vikings in weeks.

by LionInAz (not verified) :: Tue, 12/20/2011 - 5:59pm

Jamarca Sanford has been there at safety, but otherwise, you're right -- Abdullah and Winfield are hurt, Cedric Griffin hasn't played up to his pre-injury level, and Chris Cook got himself arrested. It's been 2nd stringers for weeks.

by milo :: Tue, 12/20/2011 - 4:54pm

Of all the gin joints, in all the towns, in all the world, with all the banged up secondaries that ever existed....Brees is the first quarterback in NFL history to complete 80+ percent of his passes and throw for at least 400 yards and five touchdowns with no interceptions in a game.

just. slightly. above. average. .....sure

by Aaron Brooks Go... :: Tue, 12/20/2011 - 7:35pm

The average QB completes 70% against the Vikings.

In NFL history, only 7 starters have accomplished this, and only 4 who played the majority of the season.

Granted, two of those are Drew Brees. But basically, the 2011 Vikings make the average NFL start look like 2009 or 2011 Drew Brees.

by milo :: Tue, 12/20/2011 - 8:40pm

Carson Palmer and Josh Freeman completed over 70% against Minny, with a touchdown apiece.
Matt Flynn completed over 70% against Minny, with three touchdowns. All of them for less than 8 yards/attempt.
Aaron Rodgers completed over 70% twice and scored 3 and 4 touchdowns with over 8 yards/attempt. One game was at 80%.
So 4 other QBs completed 70%. One other completed 80%.

8 other QBs did not complete 70% of their passes. 11 other QBs did not complete 80%.

So the 2011 Vikings do not make the average NFL start look like Brees. Aaron Rodgers does, but I am not about to say that he plays like a slightly above average QB. I will leave a statement like that to someone with just slightly above average intelligence, more or less.

by Keir (not verified) :: Tue, 12/20/2011 - 1:23pm

The Cedric Benson comment spurred a question in my mind. Do players get the benefit of the play before the fumble? Clearly running 80 yds forward and fumbling is better for your team then 10 yds back and fumbling. If so, is there some 0 value play where the yardage gain pre-fumble negates the negative value of a fumble?

by Eddo :: Tue, 12/20/2011 - 1:39pm

There would indeed be.

The outcome value of a play is basically derived from The Hidden Game of Football, which assigns an expected points value to each yardline on the field.

So if you started from a situation where you were expected to score +2 points, and rushed for a big enough gain to where the fumble resulted in your opponent having the ball at a point where they were expected to score -2 points, it would be a net neutral play.

by bigtencrazy (not verified) :: Tue, 12/20/2011 - 1:28pm

I think Jermichael Finley just dropped another pass. Kind of thought he might be in the running for worst receiver even with the other things he DID accomplish.

The guy has real talent and at times makes some incredible catches. But he has been dropping catchable balls now for what seems like the last six games or so while also insinuating in the press that he isn't getting the ball enough.

Be nice if he backed up the talk with a big game against the Bears.

by MJ (not verified) :: Tue, 12/20/2011 - 1:47pm

No, no, I'm fine with him dropping passes for another week! (yes, my opponent in the championships has him)

by Arkaein :: Tue, 12/20/2011 - 5:57pm

Well, the Bears are the one team he's actually lived up to his hype for this season.

But his hands this year are so strange. After his late season 2009 through 2010 before he was injured I would have said he had the best hands on the Pack, bar none. This year he has the worst. It's like James Jones stole his gloves.

by Vincent Verhei :: Tue, 12/20/2011 - 1:53pm

How does a QB with fewer yards,
fewer TD's and more int's (Vick)
than Brady end up rated higher?
Please don't tell me the Jets
defense is so much better than
the Broncos. This is like Gronkowsi
last week being ranked 5th instead
of 1st or 2nd.

Others have pretty much answered this, but A) rushing value had a lot to do with it, and B) yes, the Jets passing defense is that much better than the Broncos passing defense. Without opponent adjustments, Brady would have ranked higher than Vick.

On Brady's rushing day: He had four carries, all needing exactly one yard for a first down or touchdown. The baseline for those carries is very high, and thus failure is severly punished while success is lightly rewarded. His two-yard gain on second-and-1 was worth essentially zero DYAR. (It was actually a negative YAR play before we add in adjustments for playing the Broncos, whose run defense has been OK.) Each of his two stuffs at the goal-line was worth more negative DYAR than the positive DYAR he gained for the one time he crossed the goal-line.

On Seneca Wallace: You know what? All of you are right. I was wrong. Wallace has been a much better player in Cleveland than I realized. In my defense, his 18.2% DVOA in 101 attempts last season was much better than anything he ever did in Seattle, when I actually had the misfortune of watching him play.

I agree 100% with the point Will Allen and others have raised about how the soft passing rules harshly punish those teams without good quarterbacks. There's really only one way to build a team now, and if you don't have one of the top 5 or 10 passers in the league, you're going to have trouble winning year to year. It's kind of like the NBA, where you can't win without a superstar player, but only one or two superstar players enter the league each year.

I'm not really a fan, but I think Tebow was far better than stats show this week. Because of the game situation, the 28-yard sack was no worse an outcome than a 17-yard gain would have been, so risking a big sack was worth it if there was any chance of escaping for a big gain/completion; it's the equivalent of an interception on an end-of-half/game hail Mary pass.

You're reading too much into my comment. My point wasn't to rake Tebow over the coals or taking a megasack. It was just an interesting trivia note. By DYAR, that sack was essentially the same as his other three sacks on the day. Each was worth -14 to -17 DYAR. His 10-yard sack on third-and-8 in the third quarter was a worse play.

Stop trying to be funny, Vince. Know who you are.

Aww. Anyone agree/disagree with this? I liked it.

Re: Brees - The Vikings defense had allowed 26 TD passes before Sunday. It's now 31, obviously.

My God, you're right.

I think Jermichael Finley just dropped another pass. Kind of thought he might be in the running for worst receiver even with the other things he DID accomplish.

-1 DYAR. Only caught three of ten passes, but those three receptions went for 83 yards.

by Aaron Brooks Go... :: Tue, 12/20/2011 - 2:13pm

The NBA is dire. The number of NBA Champs without at least one confirmed or probable HoF member: 1.

The number of NBA Finalists without at least one confirmed or probable HoF member: 5.

by nat :: Tue, 12/20/2011 - 2:25pm


I understand how you arrive at two "stuffs" (arguably the best results possible, considering the score and time remaining) would each? cancel an earlier TD in a one-point game. I understand that a two yard gain to get a first down might be below average. But this isn't a sign that Brady played badly on those plays. It's a sign that DVOA has problems judging short yardage situations, and deep problems when clock management becomes important.

Just what was the baseline success value on a first-and-goal from the one, up eleven with less than five minutes left to play? Where those plays treated as similar in value to, let's guess, a six yard sack on second-and-ten?

by jimbohead :: Tue, 12/20/2011 - 2:29pm

re: seneca wallace - its ok Vince. We all understand the horrors of being a Seahawks fan during the late Holmgren era.

by nat :: Tue, 12/20/2011 - 2:33pm

Regarding Vick and Brady, you might take a look at the total baselines for each team in the games. Is it possible that Brady and Vick were judged differently, not because of the success points over average they each earned, but because those success points were divided by different denominators?

As you point out, first-and-goal from the one sets a very high bar, and that high bar devalues ALL plays in the game, not just one.

by Vincent Verhei :: Tue, 12/20/2011 - 2:41pm

Is it possible that Brady and Vick were judged differently, not because of the success points over average they each earned, but because those success points were divided by different denominators?

I don't even need to check that. Not only is that possible, I'm sure it happens all the time. What's your point?

by nat :: Tue, 12/20/2011 - 3:14pm

My point is that they were graded on different scales. Each success point is supposed indicate equal amounts of success - at least after subtracting out the "average" success for the situation. That's how the success value was designed to work.

Consider: If Vick and Brady each make the same play in the same situation in the first half of his game, why does YAR wait to the end of the games to figure out which play had the most value? They were exactly the same. Obviously. Any other answer is wrong.

But that's exactly what YAR does when it divides by a game-specific total baseline. It waits (unnecessarily) until the end of the game to determine (wrongly) that the two plays had different value. (edit: I am assuming here that YAR is figured the same way as VOA. Correct me if I am wrong.)

Subtracting out a situation-specific baseline to get "success over average" is good. Applying a game-specific scaling factor is bad.

by Eddo :: Tue, 12/20/2011 - 3:51pm

YAR doesn't divide by anything, it's a counting stat. And I'm fairly sure the statement "If Vick and Brady each make the same play in the same situation in the first half of his game, why does YAR wait to the end of the games to figure out which play had the most value?" is wrong. If both Brady and Vick throw an eight-yard pass on first-and-ten with 5:00 left in the first quarter, they'll both get the same YAR for it.

The "D" part, I believe, is applied afterwards, though.

by nat :: Tue, 12/20/2011 - 4:42pm

YAR converts (success points minus replacement level) to YAR-yards. If that factor is game-specific or play-specific, my concern remains. The grading scales for two QBs would be different - albeit in two different ways. If the factor is constant, then life is good.

For VOA, the scaling factor isn't constant, which leads to problems. I assumed YAR had the same issue. I'd love to be wrong.

by Joseph :: Tue, 12/20/2011 - 9:57pm


I believe Eddo is correct about YAR--it's a counting stat, and so the factor/grading scale/etc. is constant. If I understand it correctly, it is somewhat based on Success Rate. The "D" in DYAR is a different factor for each team, and I believe is divided into running plays vs. passing plays. In other words, when teams played MIN the last few years, success on runs was rewarded more than passing success.

by nat :: Tue, 12/20/2011 - 11:08pm

Well, I hope you're both right.

I understand about the D factor being different for different teams. That's as it should be, which is why I focused on YAR instead of DYAR.

by Vincent Verhei :: Wed, 12/21/2011 - 1:09am

Both VOA and YAR are based on the same principles:

* Calculate the Point Value for each play.
* Calculate the average-level and replacement-level performances for all plays in similar situations.
* Calculate the difference between Point Value and average Point value or replacement point value
* For VOA, the resulting number is then expressed as a percentage of the average point value. For YAR, the resulting number is then modified to resemble value in terms of yards of field position. This modification is the same for all plays and all teams. It is set for each play at the point it happens, and does not change at the end of the game or at the end of the year.
* After all that, both VOA and YAR are modified for opponent.

by nat :: Wed, 12/21/2011 - 9:59am

That's a relief. YAR, at least, isn't hit by the team and game specific scaling problem.

Thanks for a clear explanation. Excellent job.

by tuluse :: Wed, 12/21/2011 - 1:52pm

Why would you think it was?

The "D" that prepends YAR or VOA is the opponent adjustment part that does game specific adjustments.

by nat :: Wed, 12/21/2011 - 2:13pm

Because VOA itself is subject to a game-specific scaling factor when it is converted from "total success points above average" to a percentage. This happens before the "D" adjustment is applied.

If your team gets into an unusually large number of short yardage situations, its total baseline for that game will be unusually high. That will devalue the excess success points you team earns from each and every play (not just the short yardage ones), forcing your team's VOA closer to zero for reasons having nothing to do with quality of play.

Imagine two teams, each with 60 offensive plays, each with a 10 yard gain on second-and-nine from midfield in the first half. Those two plays could contribute a different amount to those teams' VOAs, not because the plays are different, but because the teams got into different situations in the second half.

Why should a second half play change the value of a first half play?

by tuluse :: Wed, 12/21/2011 - 2:21pm

"Why should a second half play change the value of a first half play?"

I don't think it does. Unless you have a specific example of this.

by nat :: Wed, 12/21/2011 - 2:35pm

If you go back to the "Last Word on the Jets Week Five" article, you'll see that

VOA = sum(success-baseline) / sum(baseline)

By beginning-level algebra, that's the same as

VOA = sum( (success-baseline)/ sum(baseline) )

The inside term is just this: the success value above baseline from one play, divided by a factor related to the overall mix of plays for the entire game.

QED: each play's contribution to VOA is scaled by a factor determined by other plays. That includes first half plays whose value is largely determined by what situations (therefore baselines) come up in the second half.

I hope that's clear enough. I could walk through an example from that game if you want.

by tuluse :: Wed, 12/21/2011 - 3:42pm

That's how you calculate VOA for an entire game, not for one play.

So yes, doing worse later in the game will lower your VOA for the game, but if you took a 1 play section out and looked at it, it wouldn't.

by nat :: Wed, 12/21/2011 - 4:49pm

The effect I'm talking about has nothing to do with how well you do in the second half. It has to do with the baselines you face throughout the game.

You are right that you can take out a section of a game and create yet another scaling factor for those plays. But that's not the scaling factor that is used in calculating the game's VOA. At best, that's a red herring. But really it's just another example of how a play's value is scaled inconsistently.

These are the steps of calculating VOA (leaving out redzone for now):

(1) A success over baseline score is calculated for each play.
(2) A total baseline is calculated for the whole game.
(3a) Each play's score is divided by the total baseline to give its contribution to VOA.
(4a) Those impacts are totaled to give VOA

Technically, the VOA calculation reverses steps (3) and (4) like so:
(3b) The plays' scores are totaled.
(4b) The total is divided by the total baseline.

But division is distributive over addition. So (3a) and (4a) give the same result as (3b) and (4b). This is not a coincidence. It's mathematically the same calculation and means the exact same thing.

It's not that doing badly late in the game will lower your whole game VOA. It's that getting into a lot of high-baseline situations in the second half will lower the value (step 3a) of your first half plays. It doesn't matter how well you do on those second-half plays at all.

Why should getting into a lot of short yardage situations in the second half effect the value (3a) of a first-half play? It shouldn't. And yet, it does.

by Vincent Verhei :: Wed, 12/21/2011 - 4:47pm

You're off on this. The baseline for each play is just that, the baseline for each play. We do not use the same baseline for entire games. Otherwise, we couldn't do things like calculate DVOA on first downs across multiple games. The "sum of baselines" is determined by all plays in whatever sample you're talking about. Therefore, the "sum of baselines" for one play is in fact that one play.

If you're point is that a few unexpected results on high-leverage plays (high success on a play with a very low expected baseline, or vice versa) can skew DVOA, then you're correct. But don't you think teams SHOULD get extra credit for converting a third-and-20, or turning the ball over on first-and-goal at the 1?

by nat :: Wed, 12/21/2011 - 5:02pm


Please look at the data presented in this article.

You will see that the games' VOA is indeed calculated from

sum(value over baseline)
divided by

You may never present a single play's VOA impact as a separate number. But your calculations do indeed scale a play's contribution to VOA in the way I have described.

by Eddo :: Wed, 12/21/2011 - 5:53pm

Yes, but your initial statement was:

"Consider: If Vick and Brady each make the same play in the same situation in the first half of his game, why does YAR wait to the end of the games to figure out which play had the most value? They were exactly the same. Obviously. Any other answer is wrong."

YAR does NOT wait until the end of the game to figure out which play had the most value. In fact, YAR will judge each individual as exactly the same.

What you're citing with the sums is the full-game VOA.

by Vincent Verhei :: Wed, 12/21/2011 - 6:15pm

But your calculations do indeed scale a play's contribution to VOA in the way I have described.

OK, when you word it that way, yes, you're 100% correct. When we analyze a group of plays -- whether it's in one game, one quarter, all third down plays, all red zone plays, whatever -- performance on those plays with the highest or lowest expected baselines will have the biggest impact on VOA. No question about that.

I guess what you're proposing is that rather than sum the values and baselines from all plays, we should calculate VOA for each play individually and then use the average of all DVOAs. Is that a fair summary of your argument?

by nat :: Thu, 12/22/2011 - 10:20am

Actually not quite. I don't want to move from a game-specific scale (total baseline) to a play-specific scale (baseline for each separate play). I want to move to a uniform scale, so comparing VOAs and DVOAs is valid.

Here's my thinking: A success point is supposed to be worth - in a predictive skill sense - the same in all plays. Because of the way success value is defined, the "zero point" for success may have a bias one direction or another for different situations. That should be okay, because we always compare to an average or replacement level for the situation.

The key point is that any additional success point represents the same skill over the opponent as any other success point. Success points are all the same size.

We'd like a VOA percentage point to have the same consistent size property. An extra 1% VOA for one team should represent the same increase in per-play dominance over their opponent as an extra 1% VOA represents for any other team in any other game. Otherwise, VOA won't work to compare raw performance in games and DVOA won't work to compare team quality.

The only way to accomplish that goal is to convert success points into VOA% using a factor that is proportional to the number of plays (in the game or season being judged) and no other factor.

The best divisor I can think of - it's the obvious one - is the number of plays in the game times the average baseline for all plays. In other words, the amount of success an average team playing against an average team with an average mix of situations would have had in the same number of plays.

That way you really could use VOA to compare raw performance, and DVOA to compare team quality. Every team, every game, every play would be graded on exactly the same scale.

Most of the time this hasn't been an issue, because the mix of situations in each game tends to be near some kind of "average" mix anyway. But every once and a while, DVOA coughs up a fur-ball. This inconsistent VOA scale could be a part of the problem. It may even be most of the problem.

by nat :: Thu, 12/22/2011 - 10:44am

If the previous post was too long to read, here's the short form...

Success points are all the same "size".
VOA percentage points should all be the same "size". (They aren't today.)
Use the average success for all plays in the database as the scaling factor.
VOA for a game or season will then be the average VOA for all the plays in the game or season, as it should be.

VOA would be improved for comparing teams' raw performance. DVOA would be improved for comparing teams' quality.

by Vincent Verhei :: Thu, 12/22/2011 - 1:09pm

So you do think that a third-and-2 conversion should have the same value as a third-and-20 conversion. Because they'd be worth the same number of Success Points, and what you're proposing sounds like a uniform baseline.

In fact, why bother with a baseline at all, then? Why not just use Success Points Per Play?

by nat :: Thu, 12/22/2011 - 2:16pm

Success Points Per Play wouldn't work because while Success Points are intended to be of equal value at the margin, there's nothing in the methodology that makes them have the same "zero point". A team's SP/P score would depend heavily on the situations they got into rather than their performance in those situations.

You could use Success Points over Baseline per Play (SPoB/P) to compare raw performance. But dividing by an average baseline instead of just the play count lets you quote a percentage, and thus to avoid have to indoctrinate everyone into the Success Point concept in detail.

by Vincent Verhei :: Thu, 12/22/2011 - 3:49pm

Success Points Per Play wouldn't work because while Success Points are intended to be of equal value at the margin, there's nothing in the methodology that makes them have the same "zero point".

But dividing by an average baseline instead of just the play count lets you quote a percentage, and thus to avoid have to indoctrinate everyone into the Success Point concept in detail.

I still don't understand what the difference is between these two, except that dividing by a uniform baseline average will result in a percentage. You'd still have the same teams ranked in the same order, by the same margins.

by nat :: Thu, 12/22/2011 - 4:41pm

Your quote skipped the important step: the difference between "Success Points" and "Success Points over Baseline". Note: "Over" here means "in excess of" and not "divided by".

Let's call them SPs and SPoBs.

SP/Play has the same "zero point" problem that SPs have.

SPoB/Play doesn't have a "zero point" problem, since it's measuring relative to a baseline. But it does end up being measured in Success Point units. That's fine for experts, but useless for most people. Most people don't understand or care what a success point is.

SPoB/(plays*average baseline in database) is as good as SPoB/Play, and puts all plays on an equal scale. It's effectively the same comparison, but converts it into units that make intuitive sense and are uniform: percentage of a typical play's success.

Using this approach, a +12% VOA would mean per play value "Above average by 12% of a typical play".

by Vincent Verhei :: Thu, 12/22/2011 - 4:54pm

OK. Followed most of that. One question: The "Baseline" in "SPoBs" -- would you have that be situation-specific, or use the league-average, all-situation baseline average?

by nat :: Thu, 12/22/2011 - 5:14pm

Great question!

Each play's success should be compared to the average or replacement-level success for that situation. That's what makes it adjust properly for situations.

But the league-average all-situation baseline should be applied as the scaling factor. That's what would make it valid for comparisons, and avoid the occasional truly weird VOA result.

All of this is "in my opinion" and "subject to off-season testing" etc. And all of this is targeted at making DVOA better, rather than complaining about any team's ranking.

by dryheat :: Thu, 12/22/2011 - 5:23pm

Oh my God...you nerds are still at this?

by nat :: Thu, 12/22/2011 - 9:17pm

Just about wrapped up for now. but thanks for asking.

by nat :: Thu, 12/22/2011 - 2:32pm

Your specific example just points out that the Success Point formula isn't perfect. (We knew that. Certainly for clock management situations, and perhaps for short or long yardage situations as your example suggests.) Compared to getting no gain in each situation, both of these plays get a new set of downs, but one of them improves field position more. They should be valued (a bit) differently.

But whatever the flaws in the Success Point formula, dividing by a per-play baseline isn't going to help. Nor is dividing by a per-game baseline.

by Karl Cuba :: Tue, 12/20/2011 - 2:46pm

"Stop trying to be funny, Vince. Know who you are.
Aww. Anyone agree/disagree with this? I liked it."

I think your writing has really improved this year and you are at least a bit funny.

by tuluse :: Tue, 12/20/2011 - 2:59pm

I enjoy Vince's comments and I think he's a huge upgrade on Barnwell.

by RickD :: Tue, 12/20/2011 - 3:21pm


by Faint Praise (not verified) :: Tue, 12/20/2011 - 3:34pm

This is me damning you.

by Eddo :: Tue, 12/20/2011 - 3:51pm


by bigtencrazy (not verified) :: Tue, 12/20/2011 - 3:56pm

Strongly agree

Vince educates and shares.

Others.........did not.

HUGE change for the better.

by BroncosGuyAgain :: Wed, 12/21/2011 - 11:58am

Funny is in the eye of the beholder, so I wouldn't worry about one comment. I think you are doing a great job with this column and have set the tone properly.

by MJK :: Tue, 12/20/2011 - 4:12pm

His two-yard gain on second-and-1 was worth essentially zero DYAR. (It was actually a negative YAR play before we add in adjustments for playing the Broncos, whose run defense has been OK.)

Hmmm, I find this surprising. So, leaving out opponent adjustments, a 2 yard gain on 2nd-and-1 has a negative YAR? How high is the baseline for a 2nd-and-1 play? A negative YAR means that you expect a REPLACEMENT level quarterback to gain more than 2 yards on a sneak on 2nd and 1? Or just that you expect a replacement QB to gain more than 2 yards in some way, shape, or form? Given that most QBs either sneak the ball (generally gaining only a yard, if that) or attempt a pass, which probably falls incomplete a fair portion of the time, I'm surprised that the baseline for an average QB, let alone a REPLACEMENT QB, is more than 2 yards. Do we really expect Curtis Painter to consistently gain more than 2 yards on 2nd and 1?

I would buy that a replacement RB might have a higher expectation on 2nd and 1. I would also buy that a QB *might* have a higher expectation if you take both pass and rush attempts into account. But not that a QB rushing attempt would have that high of an expectation. Does YAR (and DYAR) take the ballcarrier's position into account when setting baseline, or just the scenario? Does it take the playcall into account? Are you claiming that Tom Brady didn't gain more yards then you would expect a replacement player of unspecified position to gain using any kind of play, or that Brady didn't gain more yards than you would expect a replacement QB to gain on a rush? Or something in the middle.

It seems counterintuitive to compare a rush by a QB to the expectations of a pass by a QB...it means that QB's that play for coaches that like to call safe, get-the-first-down-with-a-sneak-instead-of-trying-the-homerun-pass type of call will always end up with a worse DYAR. It also seems counterintuitive to compare a QB sneak to a designed run by a RB...obviously, your expectation for the RB will be higher.

by tuluse :: Tue, 12/20/2011 - 4:18pm

I think they expect a replacement level rusher to gain 2 yards on that play.

by nat :: Tue, 12/20/2011 - 11:18pm

I think the sneak is being compared to all designed runs for the QB, including bootlegs, draws, options plays and the like. I'm assuming Michael Vick is in the average somewhere. He's probably not just running sneaks. So perhaps gaining two yards is below replacement level as they calculate it. Especially when you consider that yard line and yards to go are always reported without any fractional part in the play-by-play.

Surprising. But not bonkers.

by dryheat :: Tue, 12/20/2011 - 5:10pm

It's just a weakness in calculating rushing DVOA for quarterbacks. Sneaks are essentially dives to get one yard, give or take a foot. The QB is usually going to the ground, forfeiting further yards. In these cases, if the QB gets the yard, he's gotten 100% of the yardage available on that play.

We all know this, so what to do about it? We could live with it, which isn't a major problem. Or somehow grade sneaks on a pass/fail basis....maybe treat them similar to the way DVOA treats runners getting a touchdown from a 1-yard plunge?

by Vincent Verhei :: Wed, 12/21/2011 - 1:16am

When quarterbacks run, they are compared to other quarterbacks who run. Yes, this means a Tom Brady sneak is compared to a Michael Vick scramble or a Cam Newton power dive play.

by nat :: Wed, 12/21/2011 - 10:07am

Shit. I forgot "undesigned" runs (scrambles). That makes QBs look like really good runners, since a scramble that loses yards isn't counted as a run, but as a pass play. Understandable. Problematic. But a hard nut to crack.

This would also explain the high replacement value for a goal-to-go-from-the-one QB running play. All scrambles from the one yard line score, by definition. If they don't score, they're not scrambles, and thus not counted as running plays.

by MJK :: Wed, 12/21/2011 - 1:57pm

Ah, yes. I see that this is a fundamental problem created by going off the official play-by-play and not charting data. Most negative QB runs are scored as "sacks" on the play-by-play, whereas any positive QB run (designed or undesigned) is scored as a rush. Hence the baseline for any QB rush is necessarily going to be very high...probably higher than for a RB.

So a coach by calling a sneak is automatically dooming his QB to a negative YAR play, even if it is the tactically correct thing to do.

This is going to grow to be an even bigger challenge if Cam Newton and Tim Tebow's successes cause the college option play to be seriously revived in the NFL. On an option, if the QB laterals it to the RB running behind/beside him, it's a rushing attempt by that back and won't affect the QB. If he passes it forward to the TE, it's a passing attempt. If the QB keeps the ball and is tackled for a loss, it will be recorded as a sack, I think, and hence not contribute to the rushing YAR baseline? If he keeps it and gains yards, he gets positive yards for rushing, pulling the YAR baseline up.

by tuluse :: Wed, 12/21/2011 - 2:00pm

It might cause weird splits for passing/rushing YAR, but the over all YAR should still do as good of a job as it was.

Well except QBs won't get any credit for pitching it on an option.

by Raiderjoe :: Tue, 12/20/2011 - 2:21pm

Andre roberts ztrArting to burst out

by Vincent Verhei :: Tue, 12/20/2011 - 2:29pm

It's a sign that DVOA has problems judging short yardage situations, and deep problems when clock management becomes important.

I disagree with your first point, but agree 100% on the second. Aside from wiping kneeldowns and clock-kill spikes off the record, there is no clock management variable in DVOA. We've never said there was.

Remember last week, when Marion Barber failed to stay in bounds on a 5-yard run against Denver, stopping the clock and giving the Tebows enough time to rally and win? That was a (barely) positive play in DYAR. Check the Audibles for that day, though, we ripped Barber for his dumb play. DVOA has flaws. We acknowledge that.

by nat :: Tue, 12/20/2011 - 2:47pm

I never said you never said....

It's okay that DVOA has flaws. It is, and should be, a work in progress. Although I harp on its flaws because I want to see DVOA improved, I appreciate what DVOA does: The focus on play-by-play results, the comparison to average, the non-linear value of yards gained, the value assigned to turnovers, the emphasis on repeatable skills... the list goes on.

I could be wrong about short yardage. Or there might be a problem in the huge step in value between 0 and 1 yard on first-and-goal from the one, and a problem with the huge baseline of such a play diluting the value of all other plays.

by Jimmy :: Tue, 12/20/2011 - 4:19pm

Couldn't you add an adjustment for a play when someone blows a chance to ice the game?

by Maltodextrin (not verified) :: Tue, 12/20/2011 - 3:12pm

On fourth-and-1 from the Detroit 24 in the first quarter, Palmer threw incomplete to Denarius Moore into the end zone. Given the game's final margin, the Raiders probably regret that playcall.

But Moore was wide open. Palmer overthrew him. That's not a reason to regret the playcall. It was a perfect playcall, bad execution.

by zlionsfan :: Tue, 12/20/2011 - 3:29pm

Agreed. It was a great call: fourth-and-one is an aggressive down, and with Detroit's success against short-yardage plays, something other than a run was a great idea. (I say that as a fan of strategy, not as a Detroit fan. My reaction to the play live was "oh, ****", which means it was a good call.)

by Aaron Brooks Go... :: Tue, 12/20/2011 - 3:56pm

You're still arguing a performance-based evaluation, and not a method one.

Moore was wide open as much because Detroit performed poorly on the play, but Oakland had no way going into the play to know that. You're couching your argument as the playcall being good because Moore performed well. Vince isn't arguing that the playcall was bad because Palmer performed poorly, but because in the end-game situation, the difference between 0 pts and 3 pts was worth more than the difference between 3 points and 7 points, and despite volume-averaging of probability, scores still have discreet values.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 6:26pm

I think there's a psychological factor as well. The Raiders were struggling to score points the previous three weeks. Their defense is suspect. They're facing a team that can go downfield and score in a hurry. You want to put them in a hole so that they're forced to eliminate some calls from the playbook. If you're up against a ponderous team like the Broncos or Titans, you'd probably go for the FG. But if you're up against the Saints or the Pats, you want every TD you can get, and you want to get them early because it's harder for you to play from behind than them.

by zlionsfan :: Tue, 12/20/2011 - 3:21pm

I think this was an accurate assessment of Stafford's season. (We might as well call it his "career", considering that it's been so different than his first two seasons. At worst, he's had little, nagging injuries that a lot of QBs get, but nothing like 2009 or 2010.) He's shown the ability to put points on the board in a hurry against defenses that are average or worse; against better defenses, his success is spotty at best.

I don't think it's a coincidence that the Lions are just 2-4 against current playoff teams, and both wins came against teams of questionable quality that haven't locked down playoff spots yet. Detroit has done a good job cleaning up against weaker teams (just one loss, and that was to good Chicago), but they've yet to demonstrate they can play at the same level against better teams. Too many mental mistakes (more than a few by Stafford himself) took away any chance they had of beating those better teams.

9-5 is certainly a pleasant change from everything that happened since Millen destroyed the team, but I've no illusions about what it means. After all, I remember a 12-4 team that arguably wasn't much better than this one ... hopefully the DVOA era will eventually extend back that far. The Lions that season still had trouble against quality opposition, getting blown out twice by the Redskins (although Barry missed the first meeting).

It didn't matter too much, though. When your expectations are perennially low, even a taste of success is enough for the moment. Following up 0-16 and 2-14 with 6-10 and then maybe 10-6 would be just fine ... a wild-card win would be even better, but even if it's a playoff loss, there is, for a change, hope for the future.

by Aaron Brooks Go... :: Tue, 12/20/2011 - 4:18pm

Stafford had a putrid game @CHI (Cutler version) and poor ones against GB and @Atl.
He was decent home against CHI (Cutler version). The 2nd Chicago and GB games were played with a broken finger.

He played well against SF, Dallas, NO, and lit up a decent Denver defense. He's been more betrayed by a complete absence of a rushing attack in his losses. Probably because the Lions are down to their #4 RB and a street FA.

That 1991 team went 4-4 against teams with 10 or more wins and 9-1 against teams with 9 or fewer wins. There's no shame in the 1991 49ers or 1991 Redskins having your number.

by Independent George :: Tue, 12/20/2011 - 6:06pm

I haven't seen many Lions games this season, but it still looks to me like Calvin Johnson has a bigger impact on the offense than Stafford. I know it's by design, but it looks like there are at least 4-5 plays per game where he flings it up in the air and waits for Megatron to bring it in. It's reminiscent of Randy Moss in Minnesota.

by tuluse :: Tue, 12/20/2011 - 6:14pm

It's not just CJ, though he is the best, all the Lions receivers are big and tall and Stafford throws high all the time. It's hard to tell if this is an Eli-esque accuracy issue or if intends to do this every time.

It does mean that a really disciplined secondary has a solid shot at breaking those up, but it also leads a lot of CJ jumping over 4 guys for a TD.

I think it also simplifies things for Stafford.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 7:01pm

CJ and Stafford practice those plays a lot. I think they've been doing it with Burleson too, but Stafford, CJ, and Pettigrew have had the most time playing together, so it makes some sense that they've worked together better.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 6:57pm

It's also easy to underestimate the impact that Jahvid Best had on the passing game.
He was enough of a threat to break a long play on outlets that defenses could not afford to over-commit to covering CJ. It's been up to the TEs to make up for that, with mixed results.

by Aaron Brooks Go... :: Tue, 12/20/2011 - 7:39pm

Johnson has a large effect, but the plays Johnson makes when Stafford is in are very different than the ones when Hill or Stanton are in.

Stafford throws those high bombs because CJ can catch them, not because he just throws high bombs. Hill and Stanton just can't make those throws.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 8:19pm

To be fair, Shaun Hill made those throws a few times for TDs from within the 30 yard line last year (at least once against the Packers IIRC).

Stanton, on the other hand, has always been too erratic a passer to trust on that kind of play.

by greybeard :: Tue, 12/20/2011 - 8:12pm

Regardless of whatever his numbers were, he did not play well agains 49ers.

I watched a few Detroit games this year and Stafford has play is above average QB play. But, IMO, if you were to replace Megatron with a league average #1 WR, Stafford would not have looked good. Or if you were to replace Stafford with Shaun Hill the drop off would not be that big.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 8:22pm

Talk all you want, he played a hell of a lot better than Alex Smith.

by greybeard :: Tue, 12/20/2011 - 9:59pm

I watched all of the games Alex Smith played and 4 games Stafford played this year Chi, SF, GB, Oakland.Alex Smith has played better overall IMO. Especially given the huge difference between their #1 WRs and their OLine quality I would say Alex Smith played head and shoulders above from Stafford.

by LionInAZ (not verified) :: Tue, 12/20/2011 - 6:45pm

I agree, this was an informative read and demonstrated some of the obvious shortcomings we've seen through the season (such as 3rd down performance).

It's hard to say how much of an impact the broken index finger has had on Stafford's play in the middle of the season. No one will ever admit that it affected his play.

The other factor is that he's only 23 years old, and he has played considerably fewer games than any of the other QBs listed above. I think it's fair to say that he hasn't peaked yet. Even small improvements in the running game and O-line performance would help a lot.

It's just very gratifying right now to have a QB who could even be considered in the top ten and could very well turn out to be the best QB in the history of the Lions team.

by ASmitty :: Tue, 12/20/2011 - 3:42pm

"Stafford won the game with two late touchdown drives, but his "clutch" performance (fourth quarter, within one score) wasn't as good as you'd expect: 7-of-12 for 96 yards, six first downs, one touchdown, plus a 17-yard DPI."

Uh. Part of what made his fourth quarter "clutch" was because, as you noted, he nedded two drives to win it: one to get within a score, and the other to take the lead.

Now, granted, they were down two scores because he was stripped for a touchdown, but still, the sentence as written makes little sense. It's like saying "Stafford won the game with two late touchdown drives, but if you take one of them out, his numbers aren't as good."

by LionInAZ (not verified) :: Tue, 12/20/2011 - 8:25pm

I think they're saying that a 'clutch' performance would have looked more like 10-12 passing without a big DPI call to help things along.

I don't care because the Lions won anyway.

by Biebs :: Tue, 12/20/2011 - 3:42pm

I'm curious as to how different Sanchez's numbers would have been if Holmes caught the pass in the red zone, instead of tipping the ball and it getting picked off.

I understand that tipped passes are a part of life for the QB. But I suspect that int in the red zone pushed sanchez numbers WAY down

by BroncFan07 :: Tue, 12/20/2011 - 7:21pm

"Orton threw seven passes in the fourth quarter, each while protecting a one-score lead. He went 5-for-7 for 86 yards and four first downs. Of course, he may have been ahead by more than one score if he hadn't gone 1-of-6 for 8 yards and no first downs in the red zone."

Yep, that's a Completely Competent Kyle Orton* kind of day if I ever saw one. 299 yds passing, 1 total TD, 4 FGs. No one is better between the 20s.

*Copyright V. Verhei

by Bryan and Vinny Show (not verified) :: Wed, 12/21/2011 - 1:35am

Yep. I liked Orton well enough here in Denver, but, sweet cripes, whenever they would get the ball inside the red zone, he was always dreadful. I don't know what the problem is.
At least with Tebow, he'll eventually just tuck the ball away and run toward the end zone. He may not get in, but there's usually at least some sort of progress.

by TreeRol (not verified) :: Wed, 12/21/2011 - 4:37pm

I'm not sure I get the point of assists. A QB who throws a 15-yard pass that is caught with 0 YAC is credited the same as a QB who throws a 2-yard pass that is caught and followed by 13 YAC. He gets the completion, he gets the yards, and he gets the first down or touchdown. But in the latter case he gets an "assist."

Want to lead the league in assists? Throw a lot of screens. Sure, the OL does all the work, but it's another stat that goes to the QB.

Seems like it could be a useful way to characterize an offense, but why is this a stat for the QB?

by Vincent Verhei :: Wed, 12/21/2011 - 4:50pm

It has some descriptive value, about the style of the QB and his offense. I would never say that more assists = better. I would say that more assists = different.