In this week's Varsity Numbers, Bill Connelly takes a page out of baseball's playbook and attempts to isolate power from efficiency.
26 Aug 2003
by Aaron Schatz
This is the sixth of a series of eight articles taking a closer look at every team in the NFL, division by division, using our new statistics such as line yards and DVOA (explained here). Don't be scared away by all the numbers -- the goal of these IN FOCUS articles is to go through all the numbers and translate the most interesting trends into actual English paragraphs for those allergic to endless tables of stats. You enjoy the stats separated by team, or just enjoy the insights of each team's commentary.
The format of these articles is explained at the beginning of the one on the AFC East. Remember, offensive numbers are better the more POSITIVE, defensive numbers are better the more NEGATIVE. Overall total is offense minus defense, so the more positive the better. Schedule strength is harder the more negative, with the hardest schedule ranked #1 and the easiest ranked #32. New players on a team are colored blue and players who have left the team are colored red.
Yes, DVOA believes that Jacksonville was a better team than Indianapolis even though the Colts won four more games. We're still trying to figure out why -- it is somewhat addressed in the Colts section below, and is a definite topic for further study.
OUT: AFC North, NFC East
IN: AFC East, NFC South
HURTS: Passing offense and passing defense. The AFC South plays the same teams as the NFC East this year and so these quarterbacks will feel the same effects of facing some great pass defenses. The hit to AFC South pass defense isn't quite as strong, but it will be there.
HELPS: Rushing offense. The AFC South plays the same teams as the NFC East this year, but they didn't last year. That's why the AFC South runners will have an easier schedule in 2003 but the NFC East runners will not, even though they have the same shared opponents.
|Mack, Stacey (JAC 02)||23.4%||4||97||436||31.3%||3||23.8||6|
|Swift, Justin (SFO 02)||-19.1%||15||63||-16.4%||-3.9|
|Mack, Stacey (JAC 02)||12.3%||13||79||3.6%||1.6|
Other important additions: WR Andre Johnson (R1), TE Bennie Joppru (R2), RG Zach Weigert (JAC), OLB Charlie Clemons (NOR), RT Greg Randall (NWE), RB Domanick Davis (R4), RB Tony Hollings (SUPPLEMENTAL DRAFT)
Other important losses: OT Tony Boselli (RETIRED), RT Ryan Young (DAL), LG Cameron Spikes (ARI), OLB Jeff Posey (BUF)
Tight end Billy Miller was the only Houston starter with a positive value in 2002. Based on total value over average (DV+) he was the best tight end in the game. Miller was even better with in the red zone with a +80% DVOA: six passes, three touchdowns. To top it off, he was the only Houston player worth anything in fantasy football.
So will somebody please explain to me why Houston not only added a free agent tight end, ex-49er Justin Swift, but also drafted a tight end in the second round, Michigan's Bennie Joppru? What am I missing here?
Based on 2002 VOA ratings, Houston has the hardest schedule in the league in 2003. Ye gods.
Houston had a very standard receiver set-up, as Corey Bradford was one of the league leaders in passes right and Jabar Gaffney was one of the league leaders in passes left.
Gaffney was the Houston leader in red zone passes, with a measly seven. His DVOA in the red zone was an astonishing -100%. I love expansion football!
If you think Houston started out bad, they got worse; on third down they were abysmal passing and horrid running:
|Player||1st Down||2nd Down||3rd Down|
|D. Carr passing||-4%||-28%||-80%|
|J. Allen running||-28%||+2%||-75%|
|J. Wells running||-40%||-35%||-19%|
David Carr actually had a positive DVOA passing to the middle of the field, however. Oddly, this wasn't because that was the outlet where he used Billy Miller. Here are the three main Houston receivers with DVOA in each direction:
|Receiver||DVOA left||DVOA middle||DVOA right||# Passes left||# Passes middle||# Passes right|
There isn't much to say about the Houston defense. The offense pretty much always put them in horrible situations and, compared to other teams in the same situations, they were pretty much average in every way.
|Williams, Ricky A.||18.0%||11||35||23.2%||1.6|
|Stokley, Brandon (BAL 02)||-70.2%||6||33||-68.7%||-4.8|
|Ismail, Qadry (FA 03)||-17.9%||84||84||476||-17.3%||84||-19.3||89|
|Stokley, Brandon (BAL 02)||-7.1%||65||52||386||-9.0%||70||-4.8||62|
|Davenport, Joe Dean||-28.5%||15||76||-30.0%||-5.4|
|Williams, Ricky A.||115.2%||2||20||71.6%||1.9|
Other important additions: LB Jim Nelson (MIN), TE Dallas Clark (R1), SS Mike Doss (R2), CB Donald Strickland (R3)
Other important losses: MLB Mike Peterson (JAC), LDE Chukie Nwokorie (GNB)
The Colts' passing game was able to turn it on when it counted most: when losing, on third down, and particularly in the second half.
In the first half of games, the Payton Manning had -5% DVOA passing, while the Colts defense allowed +25% VOA to opposing quarterbacks. But in the second half of games, all of a sudden both offense and defense improved dramatically. Manning in the second half had +32% DVOA, while the Cots defense flipped from allowing +25% VOA to -25% VOA.
The difference becomes even more dramatic when you separate the first and second half stats to recognize whether the Colts were losing or winning the game. For example:
|Payton Manning pass DVOA||First Half||Second Half|
|Colts losing by more than a touchdown||-5%||+46%|
|Within 7 points either way||-5%||+21%|
|Colts winning by more than a touchdown||-6%||+14%|
The defense didn't react in quite the same way. They improved in the second half -- but improved much less when the Colts were winning:
|Pass VOA vs. Colts Defense||First Half||Second Half|
|Colts losing by more than a touchdown||+28%||+8%|
|Within 7 points either way||+29%||-20%|
|Colts winning by more than a touchdown||-1%||-49%|
One more thing: The defense didn't have a trend to match this, but Manning was also better on later downs. He had only -3% DVOA passing on first down, but he was +23% DVOA on second down and +35% DVOA on third down.
OK, so all my pride in DVOA takes a big shot when you consider the Indianapolis Colts. How the heck could a team ranked #23 in DVOA finish 10-6? Well, one possibility is that Indianapolis was one of the least-penalized teams in the league, and I may need to increase how penalties are considered in the formula. Then again, Jacksonville had even fewer penalty yards, and they went 6-10 despite a higher DVOA than the Colts.
Marvin Harrison received more passes on the right side of the field than any other receiver. He also received more passes up the middle than any other receiver and more passes between the 40 yard lines than any other receiver. Oddly, of all receivers in football with over 25 passes thrown to them, Harrison was thrown the smallest percentage of passes on the left side of the field
Marvin Harrison is so far ahead of other wide receivers, he leaves them in the dust. He's not #1 when you compare WR performance to average, but one future project is comparing WR performance to a lower level called "replacement level," a term used in baseball analysis. I have a feeling that Harrison's astonishing quality on so many more plays than other wideouts will make him the leader when each good play counts for a little bit more value.
For all the passes Harrison got up the middle, Reggie Wayne actually had the highest percentage of passes thrown up the middle, compared to all receivers with at least 25 passes. But Wayne was actually much better on throws to the sides (+42% DVOA on 36 passes) than on throws to the middle (-5% DVOA on 36 passes).
Indianapolis was one of the worst teams in the league when it came to breaking runs of double-digit yards. Could this be an indicator of Edge's slow recovery from the injury? Probably. James number was only 9%, while backup James Mungro got 23% of his yards from runs of double-digit yards.
There was also a massive difference in adjusted line yards between the two backs. Could James really be this much better to the left, with Mungro this much better to the right?
|adjusted line yards||Left||Middle||Right|
Similar to Atlanta and Tampa in the NFC South, Colt opponents ran up the middle far more often than most NFL teams and ran to the sides far less often -- even though the Colts weren't really any better at defending runs to the sides.
|Mack, Stacey (HOU 03)||23.4%||4||97||436||31.3%||3||23.8||6|
|Stokes, J.J. (SFO 02)||-19.4%||76||55||332||-19.2%||86||-14.2||80|
|Hayes, Donald (NWE 02)||-40.5%||31||149||-38.9%||-16.8|
|Johnson, Patrick (WAS 03)||-21.9%||27||24||196||-21.8%||-6.9|
|Mack, Stacey (HOU 03)||12.3%||13||79||3.6%||1.6|
* Statistics with Jacksonville only
Other important additions: DE Hugh Douglas (PHI), QB Byron Leftwich (R1), MLB Mike Peterson (IND), S Johndale Carty (ATL), S Rashean Mathis (R2), OG Vince Manuwai (R3)
Other important losses: RG Zach Weigert (HOU), K Mike Hollis (NYG)
The Jaguars were the best team in the league at preventing runs of more than 10 yards - although it helps to play in a division with the teams ranked 29, 30, and 31 in getting runs over 10 yards.
Jacksonville threw over 40% of their passes to Jimmy Smith when they were between the opponent's 40 and 20 yard lines, the highest amount in the league.
Here's a strange trend: Marc Brunell is abysmal in the third quarter, but fine in the first, second, and fourth quarters. He had a -42% passing DVOA in the third quarter, +22% DVOA otherwise. Look at the Tennessee comment; Steve McNair has the same strange trend.
Based on VOA, it looks like CB Jason Craft is the hidden star of the Jaguar defense. The Jags allowed only -2% passing VOA on passes to the left (Craft's home), but +32% VOA on passes up the middle and +37% VOA on passes to the right.
Most pass defenses loosen up when the team is winning, but Jacksonville's really loosened up:
Jacksonville winning by a TD or more: +36% VOA
Game within a TD either way: +4% VOA
Jacksonville losing by a TD or more: -24% VOA
Jacksonville's defense becomes much tougher to run on as the game moves towards the fourth quarter:
Q1: +18% VOA
Q2: +7% VOA
Q3: -1% VOA
Q4: -10% VOA
|Green, Mike (CIN 03)||-13.1%||22||71||-14.4%||-2.1|
|Dyson, Kevin (CAR 03)||-17.2%||83||69||479||-16.6%||83||-15.7||83|
|Green, Mike (CIN 03)||38.2%||10||57||30.2%||3.3|
Other important additions: CB Andre Woolfolk (R1), WR Tyrone Calico (R2), RB Chris Brown (R3)
Other important losses: MLB Randall Godfrey (SEA)
My friends and I all joked that Eddie George was running around all year with a big fork in his back, but he actually is middle-of-the-pack when it comes to DVOA. He's a back with a clear weakness and a clear strength. His weakness is breaking long runs -- only Chicago had fewer double-digit runs than the Titans. His strength -- and our stats agree with conventional wisdom here -- is that he can get the tough yards when it matters most, and the Titans' high rank in POWER running (third or fourth down, two or less yards to go) would seem to agree.
Even better is the Tennessee defense's #1 ranking in preventing POWER running success. The Titans were the only team in the NFL to allow first downs or touchdowns in less than half of these situations.
Remember what we wrote in the NFC West comment about teams that run one direction more, even if they get more yards per carry in another direction? Tennessee is one of the few teams that gets it right in terms of knowing which direction to run. They were much better running up the middle than running to the sides - and they ran more of their rushing game up the middle than most NFL teams.
Tennessee probably had the most balanced red zone passing attack in the league. Derrick Mason led the team with 15 red zone passes, but five other receivers had at least eight. When playing defense against McNair in the red zone, it's hard to know which receiver to double team.
We may have finally found proof here for the idea that defensive VOA on pass plays to the middle somewhat reflects quality of safeties. Last year, Tennesee allowed +51% VOA on passes to the middle, but only 0% VOA left and +10% right. Strong safety was a rookie, Tank Williams, while free safety was an ex-49er in his first year with the Titans, Lance Schulters. Let's be honest, the 49ers of 1999-2001 were not known as a great pass defending team.
It was a small sample size, only 19 passes, but Tennessee was the worst team in the league preventing success on passes to fullbacks.
Here's a strange trend: Steve McNair is below average in the third quarter, but fine in the first, second, and fourth quarters. He had a -15% passing DVOA in the third quarter, +23% DVOA otherwise. Look at the Jacksonville comment; Marc Brunell has the same strange trend.
Steve McNair is a better passer when the game is close:
Tennessee winning by a TD or more: -19% DVOA
Game within a TD either way: +29% DVOA
Tennessee losing by a TD or more: -13%