In this week's Varsity Numbers, Bill Connelly revisits some measures and concepts: Adjusted Scores, Covariance, and momentum (or whatever you choose to call it).
14 Feb 2007
by Bill Barnwell
The Bills were the definition of mediocrity in 2006. Finishing 16th in the league in team efficiency, their usually-excellent special teams helped mitigate the difficulties and lumps taken by their young skill position players. While Lee Evans broke out and had a season worthy of being the Bills' #1 receiver, and J.P. Losman showed signs of becoming a competent NFL quarterback, Willis McGahee suffered an injury in midseason and struggled to be the dominant back the Bills hoped he would become when drafting him out of Miami.
McGahee's performance this season was pretty typical of his career so far: his DPAR and DVOA are mediocre, but he had decent traditional numbers because he gets the ball at a pace of about 300 times a season. McGahee was the NFL rushing leader after four weeks, but very few people pointed out that his early success had something to do with his 99 carries, which also led the league. McGahee proceeded to have exactly 99 carries over the next ten weeks, which begs the question of whether a nearly 400-carry pace was too much for him.
The freely available player database at pro-football-reference.com has game-by-game statistics from 1996-2005. The most strenuous four week stretches for a running back over that timeframe would make even Herman Edwards blush. The top ten independent streaks all had 118 carries or more -- Eddie George, who had the most strenuous four-week stretch, had a whopping 128 carries. That's a pace that would put him at 512 carries for the season. And people wonder what happened to him...
Here's the top 10:
As for 2006, by the way, Johnson's worst stretch were his 116 carries in Weeks 11-14; that would put him just off this list's radar.
It's hard to say that the data points above are any sort of conclusive evidence that running backs do suffer from a four-week period of stress; after all, many of these backs were also ridden over the course of the other twelve weeks, as well, which would put them in the 370-carry danger zone that Kenny Loggins spoke of so awfully.
There seemed to be a point during the season at which the common analysis of J.P. Losman went from pity and mockery to hope and expectation. It shouldn't be said that this was something that happened independent of Football Outsiders, since the point in question occurred shortly after Aaron and Mike T. returned from their trip to Mount Jaws. Now that the season's over, we can take Losman's full season statistics and use the similarity scores technique to try and find those quarterbacks who were most similar to Losman at this point of his career.
Based upon one-year similarity scores to Losman's 2006, the five closest matches offer up a mix of hope and disgust, like the pretty girl at the bar talking to an ugly guy. Tim Couch, Tom Brady, Joe Montana, Troy Aikman, and Tony Eason represent a rather large disparity between the Hall of Fame and ... whatever it is Tony Eason does these days. Wake up in a cold sweat thinking about the Bears, maybe.
Losman has the wildman quarterback skill set; if there was physically or visually a closest quarterback to him, it would be Aaron Brooks. He obviously isn't as explosive as Brooks when he runs with the ball, but he has an absolute howitzer of an arm, scrambles way too often at the first sign of pressure, and will occasionally make terrible forced throws into coverage. He does certainly have better mechanics than Vick. It's easy to imagine that these mistakes are easier to correct than the problems of a quarterback like, say, Charlie Frye.
What can the Bills do to help Losman? Well, make sure they hold onto Lee Evans, for one. Evans was third in DVOA and eighth in DPAR this past season, which featured a terrifying six-catch, 205-yard quarter against the Texans. By not getting much performance out of any of their other targets, though, Buffalo enters 2007 one receiver away from having an offense that can strike fear into the secondaries of the AFC East.
The Bills have three key free agents to worry about: Nate Clements, Chris Kelsay, and London Fletcher-Baker. Fletcher-Baker is attracting interest from around the league and likely to leave, while Clements may choose to go following the protracted contract squabbles of last off-season. The Bills will have the cap space to re-sign Clements, but if he doesn't want to stay, he'll have plenty of suitors. Washington, for one, beckons. Kelsay fits the Bills' mold of small, faster defensive linemen and will probably stay with the club.
The Bills' cap space could be put to good use in replacing any of the players who leave on the defensive side along with improving their offensive line; unfortunately, linebacker and offensive line are two of the shallower pools in this free agent market. With this much cap space, though, the Bills could try and make a run at the big tickets of Lance Briggs and/or Adalius Thomas. On the offensive side, Damion McIntosh is someone the Bills have seen on a regular basis and who has experience at both tackle and guard; he would be a solid replacement if the Bills don't re-sign Mike Gandy. One of the many wide receivers on the market would also work while Roscoe Parrish develops -- someone like Ashley Lelie would be good in opening space up for Lee Evans to work.
The Dolphins never seemed to get off the ground in 2006. Daunte Culpepper was expected to provide the Dolphins with a boost and increased skill behind center, but his knee injury limited him to only four games on the season. The Dolphins were 1-3 with Culpepper in the lineup. While the team enjoyed victories over Chicago (ending their undefeated streak in resounding fashion) and New England, and enjoyed a stellar season from Defensive MVP Jason Taylor, their skill position players also struggled. Ronnie Brown was injured and missed a chunk of the season, and Chris Chambers' performance was the worst in the NFL by a drastic margin as per Football Outsiders' DPAR statistics.
Chambers' name has come up a lot at FO over the past 12 months. Why? He's a fascinating player with the skill set to match, and because the gap between his perceived value and his actual value is so high. Even when his repeated struggles are pointed out, many Dolphins fans will offer him excuses about this year's model of Dolphins quarterback, even after the starting quarterback count throwing to Chambers touched about 35 this season. It's worth noting that Chambers had a hand in 74 incompletions this year! That's 18 and a half drives! That's two games worth of nothing but incompletions to Chris Chambers on offense.
One of the talking points about Chambers, especially earlier this season, is that he's a second-half player and would break out in the second half of 2006. He developed this rep after 2005, when he parlayed a big second half into a Pro Bowl berth. If you look at Chambers' career, he has consistently done better in the second half of the season; he's caught more passes in Weeks 9-17 than he has in Weeks 1-8 every one of his six campaigns, averaging 10.83 catches more per season. The average NFL player, meanwhile, averages slightly over 2.7 catches more in Weeks 9-17 than they do in Weeks 1-8, which has much to do with the extra week contained within 9-17. NFL players average about 2.07 catches per week.
Do the defenders of Chambers have a point? Do second half players really exist? There's an easy way to find out -- get out the pro-football-reference database again, and do another mini-study.
Testing the theory took separating the performance of all wide receivers from 1996 through 2005 into Weeks 1-8 and 9-17, taking out players who played fewer than 14 games (since a player performing better or worse in a half of the season due to injury in the other half isn't the argument). 18 guys qualified for the study and had 21 or more receptions in the second half of the season than they did in the first. That year's performance is Split1; the same split in the year after is Split2.
When you remember that the NFL average improvement in Weeks 9-17 for a given year is 2.71, and notice that the only player who appears twice on this list is Muhsin Muhammad, it seems that this idea is a bit of a myth.
Many of Miami's elderly defensive linemen are unrestricted free agents: Jeff Zgonina, Keith Traylor, Vonnie Holliday, and David Bowens are all free to leave Miami if they so desire. Expect one or two of them, notably Holliday, to stay. Starting outside linebacker Donnie Spragan is also unrestricted, but his mediocre play isn't begging for a new contract. Offensive lineman Kendyl Jacox is also a free agent, with Hudson Houck probably getting to make the final say on whether he's worth re-signing.
The Dolphins can't bring in another quarterback, can they? The Daunte Culpepper move looks like a mess at this point, while Joey Harrington's unceremonious benching against the Jets on Christmas Night doesn't bode well for his future, either. This is another potential spot for Damon Huard, who started his career in Miami and could compete with a hopefully-healthy Culpepper for the starting job.
It's an interesting question: Was this Patriots season a success?
It's hard to say, really. On one hand, the team was a second half collapse away from the Super Bowl, where it's hard to imagine them losing to Rex Grossman and the Bears that showed up in Miami. On the other, the team known for their clutch play and veteran presence collapsed in the second half of the AFC Championship Game. Some of the blame can be apportioned around the Patriots team and organization as a whole, but a lot of it has to go down to one of the other stories of the Patriots' season; simply put, the steps that Patriots linebackers lost might be enough for a marathon.
Anyone who watched the Patriots-Colts AFC Championship Game saw the struggles the Patriots had in covering Dallas Clark and Bryan Fletcher. If Ben Utecht hadn't gone down hurt after each catch he made during the playoffs, the Patriots would have struggled covering him, too. On the other hand, Joseph Addai was held to two catches for four yards. What that doesn't note, though, is that Addai dropped a first quarter pass where he ran right by Tedy Bruschi and would have had about a 30-yard gain had he had caught it. But we'll get back to that in a minute.
That's all somewhat surprising considering that in 2006 the Patriots were fourth in the league while covering tight ends and fourth against other wide receivers, but only 22nd against running backs. The spot where the Patriots struggled was in defending other team's #2 wide receivers, where they were 29th. While the Patriots pass defense was terrible across the board in 2005, the previous seasons show similar patterns -- in 2004, they were 29th against #2 wide receivers, and 28th against running backs, but fifth against tight ends; in 2003, 23rd against running backs and second against tight ends.
Now, the Colts offensive scheme (in particular, its use of Clark) is so drastically different and special that it does tend to present itself differently than any other in football, but Clark did only have two catches for 42 yards in the first game. Could Clark's huge game and potentially beneficial matchup against the Patriots defense have been expected? The simplest way to figure that out might be to take a look at who succeeded against the Patriots in 2006.
The best performance against the Patriots in 2006 was by a Colt; it just wasn't Dallas Clark. Marvin Harrison's eight catch, 145-yard game with two touchdowns left a strong impression on Bill Belichick and Dean Pees, whose defense shifted shape to limit Harrison's big-play propensity. Number two (according to standard fantasy points) was Javon Walker, who had 130 yards and two touchdowns, but only three catches.
The rest of the big games aren't really from huge names. Number three was Mike Furrey -- he had nine catches for 123 yards and a score just from going underneath the Patriots linebackers all day and then running past them, in much the same way that Clark did. Also in the top ten was a similar player, Miami's Wes Welker -- yes, beyond the fact that they are both white, they both run similar patterns and work in the middle of the field out of the slot frequently, much like Dallas Clark.
The player who consistently played big against the Patriots was the Jets' Jerricho Cotchery. In three games, he had 16 catches for 291 yards and three touchdowns. Cotchery isn't a burner and even if he was, he wouldn't have the option with Chad Pennington at quarterback. Most of his and Laveranues Coles' work occurs in the middle of the field on slants and curls; Pennington throws the occasional out to keep the defense honest, but his arm simply doesn't allow for it.
The Colts were very smart in how they employed Clark during the playoff game -- they saw the success that Furrey, Welker, Cotchery, and Coles enjoyed and accounted for the fact that a Patriots secondary wounded by Harrison in their first matchup would shift their coverage toward him.
Unfortunately for the Patriots, their linebackers have reached the end of their life cycle. Week after week, it was hard to avoid noticing that Mike Vrabel and Junior Seau were struggling both with coverage and plugging the gaps for run plays. Bruschi was slightly more effective than the other two, and Seau got slightly better as the year went along, but Vrabel's play was a frightening drop down from where he was previously. He was often late getting to gaps and would try and cheat his way to a hole by not taking on offensive linemen or using his superior intuition and experience to guess where a play was going to go, which too often led to overpursuit and Vrabel being out-of-position. It's scary to say for Patriots fans, but their linebackers have gone from being the dynasty's defensive core to a net liability.
The two free agents whose potential absence looms largest are Daniel Graham and Asante Samuel. It's difficult to see Graham returning to the Patriots, who value him but drafted two tight ends in 2006 and already have Ben Watson. Samuel, on the other hand, showed his value to the Patriots in a blistering series of performances in the postseason. Franchising him with a view to a long-term contract seems to be an eminently reasonable proposition at this point.
While the Patriots have also clearly emphasized how fungible they think wide receivers are, the lack of impact Chad Jackson made in his rookie season may worry them about his usefulness for the upcoming campaign. With Troy Brown a free agent and perhaps retiring, the Patriots might reach out and sign someone like Drew Bennett. The Patriots would be well-served to bring in as many as three new linebackers to replace a potentially retiring Bruschi and Seau, plus the out-of-favor Tully Banta-Cain, who will probably not return and may end up in New York under Eric Mangini. They also need to figure out how to get Vrabel back to the outside, where his decline in pass coverage is less of an issue.
The Jets season has to be seen as a major success. Expected to finish last in the AFC East by most observers, an intelligent draft, the coaching of Eric Mangini, the return to health of Chad Pennington, and a weak schedule brought the Jets all the way back to ten wins and a playoff berth. With all those things as positives, there was one very negative aspect of the Jets season, a problem that may haunt them for years to come.
The Jets run defense in 2006 was not exactly one for the faint of heart. To say that, though, is a slight understatement. The Jets run defense offended pornographers, made voyeurs turn away, and repulsed the entire cast of Jackass. It was so awful that even the hipsters in Williamsburg thought the irony was a little strong. It was diabolical enough to make Jimmy Hoffa wriggle around in the end zone. It qualified for disaster relief, placed on the IMDB Bottom 100, and deserved its own Anna Nicole Smith-style reality show dedicated to its utter ineptitude. This week: DeWayne Robertson tries to shed his first lineman! Hilarity ensues!)
The point is, it sucked.
Since the off-season brings hope to every team, let's look back at the DVOA era and see what's happened to the other 32nd-ranked rush defenses in football. Keep in mind when other teams' rushing defense DVOA are quoted that the Jets' rush DVOA was a glorious 16.0%.
So, whither the Jets? It looks like most teams don't recover and become excellent run defenses overnight after being the worst in football, with the lone exception being the flukish '04 Jets rush D. The solutions were also varied: some teams drafted linemen in the first round, while others revamped their entire defense and found it to be no better. The Jets apparently think that Bryan Thomas had a good enough year in the 3-4 to earn a contract extension, which is fine. Robertson, though, is simply not playing to the level that a top-five draft pick should be at this point of his career. While it's hard being a lineman in the 3-4, Robertson neither occupies blockers nor gets penetration on anything amounting to a consistent basis. This causes Jonathan Vilma, the Jets' best player, to get blocked by an offensive lineman directly and renders him impotent to make plays. It was unclear before the year whether the 3-4 was a good fit for the Jets, and it still doesn't really seem to be the case. Eric Mangini's job this off-season will be to alleviate that, one way or another.
The Jets have no notable free agents to speak of and look set to retain the same roster for the 2007 season. Kevan Barlow is someone who might be considered to be a potential salary cap casualty, but since the Jets aren't paying any of his signing bonus, they aren't on the hook for a whole lot to begin with. Justin McCareins, who showed up for camp in poor condition and lost his starting job to Jerricho Cotchery, might also see the unemployment line at some point.
Mangini simply needs to find players who fit his defensive scheme better than the current Jets do. Right now, he has half a defense: Kerry Rhodes and Erik Coleman are a fine set of safeties, Andre Dyson and Justin Miller can hold the fort at corner, Vilma's a world-class linebacker, and Thomas and Shaun Ellis are at least solid pass-rushing ends. That leaves a tackle or two and a pair of linebackers to take care of. Getting a run-stuffer to at least alternate with Robertson is imperative, which is why someone like Ian Scott could find his way to New York. The outside linebacker crop are mostly 4-3 players, which could potentially mean -- don't shoot the messenger, Jets fans -- Cato June in the Green and White in 2007. Sorry.
Next: NFC South by Mike Tanier.
*All projected cap numbers courtesy of www.askthecommish.com. These numbers are "ballpark" and are subject to change. The intention is to give an approximate idea of each team's available resources before free agency and the draft begin.
77 comments, Last at 24 Feb 2007, 11:33pm by Josh