Here are the Week Six Quarterback and Running Back Rankings
Table One: Quarterback Rankings for Week Six
Table Two: Running Back Rankings for Week Six
And here are a few stories (with an obvious Detroit slant):
Top Rookie Running Backs
When we look at the running backs we see three rookies who are above average in Net Points per Game: Matt Forte, Steve Slaton, and Chris Johnson.
Of course Net Points per Game is just one perspective on a running back’s production. Another perspective would be to look at Net Points per Play.
For the per-game measure I only look at running backs that have carried the ball 10 times a game (or 60 times this season). For the per-play measure I am going to lower the limit to 40 carries for the season. To see why I am doing this, consider the top rookies in Net Points per Play:
- Kevin Smith: 0.189 Net Points per Play, 42 rushing attempts
- Steve Slaton: 0.185 Net Points per Play, 72 rushing attempts
- Darren McFadden: 0.158 Net Points per Play, 59 rushing attempts
- Chris Johnson: 0.155 Net Points per Play, 85 rushing attempts
- Tim Hightower: 0.136 Net Points per Play, 43 rushing attempts
- Matt Forte: 0.130 Net Points per Play, 127 rushing attempts
- Jonathan Stewart: 0.066 Net Points per Play, 71 rushing attempts
Of all rookies with at least 40 rushing attempts this year, Kevin Smith of the Detroit Lions (the team I follow) is the most productive per play. And yet, of these seven backs, Smith has gotten the fewest carries.
Would Smith post a higher Net Points per Game than Forte, Slaton, or Johnson if he got more carries? It’s hard to say. Certainly his per-play measure says the Lions should try and find out. It’s not like this team is winning games with Smith sharing the load with Rudi Johnson (who has posted a 0.160 Net Points per play).
Top Rookie Quarterbacks
So far only two rookie quarterbacks have attempted a pass this year. Matt Ryan of the Atlanta Falcons has been an above average performer. Meanwhile, Joe Flacco has been the least productive starting quarterback in the league.
Dan Orlovsky of the Lions is not a rookie (and therefore shouldn’t be discussed in this section). And he has not attempted enough passes to be ranked. Still, his Net Points per Play (how I rank quarterbacks) is currently -0.023. Such a mark is below Flacco and suggests the Lions should be thinking about their second round draft choice from 2007, Drew Stanton.
Of course, the Lions are not thinking along these lines. This week Orlovsky got most of the reps in practice. Yes, there are good reasons why this team can’t win.
The Big Trade
The Matt Millen era will be known for three things:
- The Lions lost a lot of games
- Millen only hired head coaches with the same last initial as his last initial.
- Millen loved to draft wide receivers
For three consecutive years (2002-2004), the Lions took a wide receiver in the first round. This past week, just days after Millen departed Detroit, the last of these three wide receivers left the Lions. For a collection of draft picks the Lions shipped Roy Williams to Dallas.
Now that these three wide receivers have left Detroit we can look back on what these players did for the Lions.
- Charles Rogers: 36 receptions, 440 yards
- Roy Williams: 262 receptions, 3,884 yards
- Mike Williams: 44 receptions, 539 yards
So these three receivers caught 342 passes for 4,863 yards. Immediately after Rogers was taken in 2002, Andre Johnson was taken by the Houston Texans. Here is what Johnson has done all by himself: 405 receptions, 5,292 yards.
Of course we don’t know if Johnson could have played this well in Detroit. Certainly many players have a hard time playing well for the Lions. But certainly it looks like the Lions could have just made a different selection in 2002 and received the same level of production they eventually got from all of their first picks from 2002 to 2004.
Research Question
When Matt Millen took these three receivers his basic idea was that you cannot double-team three great receivers. Therefore on any given play one of these receivers must be open. Clearly this idea never quite worked like Millen thought it would.
Here is why I think it failed. I think a team cannot have more than one number one receiver. In other words, there is just one ball. If you try and split receptions across a collection of receivers who think they should be number one, all of them will be unhappy and hence less productive. If this is true, the Cowboys are about to find out that adding Roy Williams doesn’t make Terrell Owens (or Roy Williams) a better player.
To test this, it might be good to see how a team distributes its pass attempts across its receivers. I would suspect that teams with a more even distribution do worse. In other words, you are better off with a clear number one receiver.
Let me note that I have collected no data on this question so I am purely speculating. And often when I do this – as noted in The Wages of Wins – it turns out I am wrong. Still, I would be interesting to see what someone might find.
– DJ
The WoW Journal Comments Policy
For more on the Wages of Wins football metrics see
Consistent Inconsistency in Football
Football Outsiders and QB Score
The Value of Player Statistics in the NFL
Wages of Wins Journal 
Vince Gagliano
October 18, 2008
Here’s the idea…
In basketball, man-to-man defenses are employed by less athletic teams, zones by more athletic ones.
Millen’s problem was that, by drafted so many top notch receivers, it allowed opposing secondaries to man up on each opposing receiver, making it more difficult for the quarterback to throw to a wide open man.
Also, of his first-round picks, he had one running back (Kevin Jones) and no tight ends. There was a lack of variety to keep opposing defenses honest. End result: when the pass fails, the team has to go to a run. And the running game failed because the offensive line couldn’t hold back a mob of stamp collectors at an anachronistic post office.
Besides, successful teams don’t always need a #1 option, at least, not a conventional one. Consider the three top wideouts of some successful Super Bowl era teams:
2007 Patriots (16-0):
1. Randy Moss: 98 rec, 1493 yds., 23 TD
2. Wes Welker: 112 rec, 1175 yds., 8 TD
3. Donte Stallworth: 46 rec, 697 yds., 3 TD
1998 Vikings (15-1):
1. Randy Moss: 69 rec, 1313 yds., 12 TD
2. Cris Carter: 78 rec, 1011 yds., 17 TD
3. Andrew Glover: 35 rec, 522 yds., 5 TD
1989 49ers (14-2):
1. Jerry Rice: 82 rec, 1483 yds., 17 TD
2. John Taylor: 60 rec, 1077 yds. 10 TD
3. Tom Rathman: 73 rec, 616 yds., 1 TD
1985 Bears (15-1):
1. Willie Gault: 33 rec, 704 yds., 1 TD
2. Dennis McKinnon: 31 rec, 555 yds., 7 TD
3. Walter Payton: 49 rec, 483 yds., 2 TD
1972 Dolphins (14-0):
1. Paul Warfield: 29 rec, 606 yds., 3 TD
2. Howard Twilley: 20 rec, 364 yds., 3 TD
3. Marlin Briscoe: 16 rec, 279 yds., 4 TD
Of course, this isn’t a complete picture.
Rice is the only guy who was a clear “#1” option on his team. Most of these teams have a tendency to combine a home-run threat with a short-yardage guy.
For the Cowboys, however, Williams isn’t going to get his touches from T.O., but rather Patrick Crayton and Miles Austin. Unlike the Lions, the Cowboys can afford opposing defenses manning up on their wideouts because they have the personality to exploit it.
Unlike the Lions, teams still have to worry about Jason Witten and Marion Barber. However, can they make the system work.
By the way, as a Lions fan/economist, you should be absolutely thrilled with the Williams trade.
Now that Calvin Johnson is “the man” in Detroit, he’ll get the chance to be the new #1 option on the team. The Lions will get five picks in the first three rounds of the NFL draft, including another potential top-5 pick in 2009.
Of course, they need to make sure that their manager is competent enough to take advantage of those picks. But if Bill Parcells can rebuild Miami, somebody can build up Detroit.
Ryan J. Parker
October 18, 2008
To do the research properly you’d have to constrain your data set in some way. Clearly no team would turn down having Randy Moss, Steve Smith, and Terrel Owens for fear of not having enough balls to go around. That just doesn’t make sense.
Therefore, you need to in some way identify teams that have two (or three) top quality receivers and understand how their pass distribution affects their performance. I imagine this sample won’t be that large, in addition to the notion of how you define a “top” receiver. Rarely do rookies perform like top receivers, so simply rating 1st round draft picks in this manner doesn’t make sense to me. So the Lions group has a lot of baggage attached to it.
As the previous poster mentioned, ground balance is another key factor to look at. Moss, Smith, and T.O. are going to have a hard time if the team their on can’t run the ball (see Lions).
With all of that said, I’m interested in any data / findings anyone can come up with.
Vince Gagliano
October 18, 2008
Okay, some tidbits:
* Of all 12 playoff teams last season, 8 (the Patriots, Colts, Cowboys, Giants, Packers, Bucs, and Seahawks) had 1000-yard receivers, with the Chargers and Steelers having two players catch for at least 900.
* However, of those eight, only the Pats and Colts had multiple wideouts who covered the mark. Not coincidentally, they had two of the best records in their conferences last year, combining for a staggering 29-3 mark.
* Indianapolis was the only playoff team with a 1000-yard wideout and a 1000-yard rusher.
* The season before, Dallas had two 1000-yard receivers, but also went 2-4 in their division. The NFC East has had three playoff teams in each of the past two seasons.
* In 2005, Indianapolis had two 1000-yard receivers while going 14-2. Also, note that the ’98 Vikings and ’89 Niners also had two 1000-yarders.
Clearly, there are signs that, if you’re a top team, a #1 is good, but two are better. Unless you’re the Bengals.
Ty at Bucks Diary
October 18, 2008
DB:
Supporting your hypothesis: the Packers midseason acquisition of Chargers “superstar” WR John Jefferson way back in 1981.
At the time, the Packers already had an established number one receiver in James Lofton.
It appears from profootballreference that JJ’s yards per catch after the trade were about the same as they were in San Diego (initially… they declined after year 3), but his catches per game and touchdowns per game went way, way down. And he didn’t last all that long in Green Bay.
Meanwhile Lofton’s receptions per game actually declined. But his touchdowns went up, as did his yards per reception (slightly). Whether those increases can be linked to the presence of Jefferson, I don’t know.
But my general memory is the experiment wasn’t quite as lethal as everyone dreamed it would be, and it was pretty short-lived, probably because of the reason you cited.
John W. Davis
October 20, 2008
The Lions make me sad.
Vince Gagliano
October 20, 2008
By the way, it looks like the Lions are already getting the better of the deal.
If the Cowboys continue to underachieve, what would have been a high 20’s-low 30’s pick into a mid first-rounder. Same goes for the third and sixth-round picks.
Tball
October 20, 2008
How can you subscribe to a theory of not having enough balls to go around for top NFL receivers, but discard the same theories in basketball?
A lot of the analysis above is looking to yardage, but the question went to receptions, which would make Wes Welker the #1 option on the Pats last season. Also, maybe you look at the ‘target’ stat instead of receptions, because receivers may be more sensitive to how often the ball is thrown their way than their receptions. Owens had the ball handed or thrown to him a third of the Cowboys offensive plays a few weeks back and complained (so maybe #1 receivers complain and you can’t help it with personnel decisions).
Here’s a quick study of the 2007 season identifying each team’s passing yardage and the percentage of team receptions the top wide receiver (by receptions) had (* indicates the reception leader was a RB):
NE – 4731 yds; Welker 27.8%
GB – 4334 yds; Driver 21.4%
NO – 4314 yds; Colston 22.3%
Dal – 4105 yds; (TE) Witten 28.1%
AZ – 4065 yds; Fitzgerald 28.1%
Ind – 4033 yds; Wayne 29.3%
Cin – 4012 yds; Housh 30.0%
Sea – 3964 yds; Engram 25.3%
Det – 3878 yds; McDonald 21.5%
Phi – 3755 yds; *Curtis 22.0%
Hou – 3751 yds; Walter 18.8%
Cle – 3726 yds; (TE) Winslow 26.9%
Den – 3584 yds; Marshall 31.3%
Was – 3463 yds; Cooley 20.7%
Chi – 3362 yds; Berrian 21.7%
TB – 3357 yds; Hilliard 19.6%
Jax – 3328 yds; Wilford 15.6%
Atl – 3296 yds; White 24.7%
StL – 3233 yds; Holt 27.9%
KC – 3181 yds; (TE) Gonzalez 29.6%
NYG – 3154 yds; Burress 23.2%
Pit – 3071 yds; Ward 25.2%
Bal – 3035 yds; Mason 30.2%
Mia – 3031 yds; Booker 15.7%
NYJ – 3014 yds; Cotchery 26.5%
SD – 3005 yds; (TE) Gates 26.7%
Ten – 2878 yds; Gage 18.1%
Min – 2745 yds; Wade 21.7%
Car – 2735 yds; Smith 30.5%
Buf – 2634 yds; Evans 20.9%
Oak – 2631 yds; Curry 21.2%
SF – 2320 yds; *(TE) Davis 19.0%
I don’t see a requirement for a lead dog receiver, although the less effective passing games seem characterized by a leading receiver (WR or TE) catching less than 20% of the team’s pass completions and/or catching less than 70 passes. Having multiple receivers with 70+ receptions and catching 20% of the team’s completions seemed to be a common component of the top passing teams (e.g., NE, Cin, Cle, Dal, etc)
More numbers are needed if the theory is the second WR has to be clearly secondary for the #1 WR ego to be adequately stroked (e.g., Winslow and Edwards were only two receptions apart in 2007 while Housh had 19 receptions over Cinco Ocho, which you cannot tell from this table). As Moss, Owens, and Edwards failed to lead their team in receptions, much less lead by a margin, I don’t think studying the data in this manner would be useful.
JKL
October 22, 2008
The link in my name is to a post I wrote entitled “When Should the Lions have given up on Joey Harrington?” that should be close to your heart. I looked at historical successes at QB and what numbers they put up in their first opportunities to start, as rookies and non-rookies.
I disagree with your position on Orlovsky. Honestly, there are lots of valid reasons to question the Lions organization. Giving 2+ seasons to a QB like Kitna, at his age, when he was clearly not a Rich Gannon like success and is the QB version of an “innings-eater” and compiler, when you had both Orlovsky and Stanton is an error to start this season or to end last season.
But Stanton and Orlovsky are both young. I think we can assume that if Orlovsky is getting the reps, he is the better option at this point. Why would they otherwise not go to Stanton, the local product, when they are only about a year apart in age, and neither has any prior experience?
But turning back, you assume that because Orlovsky’s numbers are bad, that Stanton is a better option after two games. Orlovsky certainly hasn’t been great, but if we look at his very small sample size of numbers, they are not out of line with alot of successful quarterbacks of the past who debuted by age 25. Most struggled with completion percentage and turnovers, but were at least respectable in yards per attempt. In other words, they were maddeningly inconsistent, but made some big plays. The guys who presented the opposite, guys who were not necessarily prone to turnovers, but simply didn’t produce big plays, are the ones who don’t pan out. Rick Mirer, for example, had one of the lowest interception rates of any QB under age 25.
That’s along way of saying that you would be wrong to exclude Orlovsky so soon. But if you want, feel free to trade him to KC for Tyler Thigpen (you want to see a QB you can rule out, there’s your guy). I don’t know if Orlovsky or Stanton (or either/both) will pan out and become long term starters in the NFL, but the numbers to date don’t dictate that decision. The Lions are not proving they are a bad organization by giving Orlovsky more starts, they are a bad organization because they didn’t recognize the need to do so sooner, while a certain GM was in charge.
dberri
October 22, 2008
JKL,
Thanks for the pointer to your post. I have found the same thing. Teams tend to stick with high draft picks too long.
And you are right about Orlovsky. It is too early to say. My sense as a Lions fan is that I do not see Orlovsky ever developing into anything special. And Stanton — in the pre-season — showed some flashes.
Of course, my sense as a Lions fan has no scientific validity.