Possibly the closest Super Bowl matchup in history also poses the question: how much does it mean when certain aspects of an NFL team improve dramatically in the second half of the season?
28 Jun 2007
by Bill Barnwell
The additions made by the Patriots at wide receiver this off-season have been a frequently discussed topic, both by the usual suspects and here at FO. It's a fascinating move, really -- part throwing stuff at the wall, part success-cycle capitalization -- and it can be taken as either a rejection or acknowledgment of the fungibility of wide receivers.
Wes Welker, one of the new wideouts, is a player who pretty much everyone at Football Outsiders, writer or reader, has championed for a couple of years now. He's a player who does a lot of things that help a team win games: He returns punts and kicks, he can be a gunner, he's a solid blocker, and as a receiver, he runs good routes, has very good hands, and has a nose for the first down marker. (He can even kick in emergencies.)
As a result, Welker is a player who has succeeded even in the dire passing attack of the 2005-2006 Miami Dolphins, and he would naturally be expected to improve in 2007 for a number of reasons. First and foremost, Welker's only 26; entering his fourth season, he could be expected to improve with experience. Second, Tom Brady is way better than anyone Miami has put behind center over the past two seasons. Welker will also have a better supporting cast around him to open up space underneath, with defenses concerned about the speed of Donte' Stallworth and Randy Moss. Indeed, those are three valid reasons to expect a reasonable bump in Wes Welker's numbers this upcoming season.
So then, why would anyone suggest that Welker's numbers will stagnate or even decline this year? Well, for each of the reasons that listed above, actually.
Perhaps due to the research done by Bill James in analyzing player career patterns in baseball, many people think that the career paths in football peak around age 27 and slowly decline. Well, football is different. The table below lists the average DPAR recorded by wide receivers that were targeted over 20 times at ages 25-35 from 1996 through 2006.
|All WR/TE||DPAR Averages By Yards per Catch, Age 25|
|Age||Avg. DPAR||9-12 YPC||12-15 YPC||15-18 YPC||18+ YPC|
As you can see, wide receiver DPAR remains consistent, with the DPAR of older wide receivers actually surpassing that of younger ones. However, keep in mind that most wide receivers are out of the league by that point; while 191 receivers recorded 20+-target seasons at 25, only 42 did at age 32. The players still around at that age tend to be nifty. This shouldn't be taken as an indication that wide receivers are not affected by age-related decline, because many receivers decline from "good enough to be in the league" to "not good enough to be in the league."
The splits on the right side of the table list the DPAR for players based on their yards per reception at age 25, Welker's age in 2006, when he recorded 9.6 DPAR despite averaging only 10.3 yards per reception. The DPAR of those players over the rest of their careers was then tracked to see if there were differences in the aging patterns of players with low yards per catch figures as opposed to those who were more of big-play threats.
The players with below-average yards per catch at age 25 (9-12) experience stunted DPAR over most of their career; they get to be about average around 29 and 30, when players hit their peak, but there's very little upside to these players. To make another baseball comparison, this aging curve looks similar to the one for young starting pitchers who don't strike anybody out.
The list of players who had such low yards per reception figures at the age of 25 includes two hopeful figures: Hines Ward and Joe Jurevicius. Marty Booker and Arnaz Battle represent a second tier of passable wide receivers. The rest of the group is an operative list of guys whose careers stagnated or never developed: Alvis Whitted, Isaac Byrd, E.G. Green, Jacquez Green, Larry Foster, Tai Streets, Peter Warrick, Rod Gardner, Antonio Chatman, Randy Hymes, and Charlie Adams. In other words, Welker will have to develop against the odds.
Yes, Tom Brady's a much better quarterback than the Miami mess, even if Miami's quarterback were actually an amalgation of Gus Frerotte, Daunte Culpepper, and Joey Harrington and had six arms, six legs, and about four healthy shoulders. He's a much more accurate quarterback; Brady has completed right around 62 percent of his passes, while the combined completion percentage of primary Miami starters Frerotte and Harrington hovers around 55 percent, a figure which would be scarily low if it were not for Welker himself. Welker caught 67 percent of the passes thrown to him in 2006, and 56 percent in 2005, both above average for his team.
Catching 67 percent of the passes thrown to you is a pretty gaudy figure; while part of that success is due to Welker's low yards per attempt and shorter-than-average routes, it's also a sign that Welker has good hands. The problem with using that to forecast future success is that guys simply don't catch 90 percent of the passes thrown to them. Since 1996, 67 receivers have caught 67 percent of the passes thrown to them. (Numerical coincidences are fun!) These receivers averaged a catch rate of 62.6 percent the year after; if you widen that range out to include all receivers who caught between 65 to 70 percent of throws, a range Welker sits smack dab in the middle of, those receivers caught 61.4 percent of throws in their direction the year after -- textbook regression to the mean.
Ah, but those receivers weren't going from Joey Harrington to Tom Brady, you say. That's true. Will that actually improve Welker? Hard to say. 116 receivers over the last 11 years moved to a new team with an incumbent quarterback. The correlation between the difference in a receiver's catch rate the year before and after the move to the difference in his old and new quarterback's completion percentage the year before making the move (thereby measuring the completion percentages of the quarterback before the arrival of the new wideout) is relatively slim at .15, so it doesn't have much predictive value in telling us whether Welker will catch more of the balls coming his direction with Brady in 2007. The average receiver moving to a new team saw his catch rate rise by less than one percent.
In other words, the more dominant variable in catch rate when it comes to Welker is his own high catch rate as opposed to the completion percentage of his new quarterback. While Welker is likely to again run short routes and show off his good hands, expect him to catch somewhere between 60 and 63 percent of the passes intended for him in 2007.
This year, Welker has more competition for passes than he did in Miami -- the area once occupied by the fearsome Derek Hagan is now occupied by Reche Caldwell or Kelley Washington, while Donte' Stallworth is actually the player that people think Chris Chambers is. Randy Moss and Ben Watson are probably upgrades on Marty Booker and Randy McMichael, respectively, even if Moss is not the superstar of 1998 that some Patriots fans think they are getting.
There's one problem with this situation. Even if the Patriots improved across the board at receiver (and they have), they only get to play with one ball. That presents a serious problem for those people who are forecasting career years for multiple Patriots wideouts in 2007. Instead, by looking at how the Patriots passing offense has been used in the past four seasons, several trends appear that serve as a guide to what we might expect in this upcoming campaign.
First, let's make several assumptions. Let's peg the Patriots wide receivers to be Stallworth, Welker, Moss, Caldwell, and two players from the Washington/Troy Brown/Chad Jackson/Jabar Gaffney quartet. The tight ends and running backs will be as they appear at the moment, and we'll assume an injury-free season for the key players.
|Patriots Targets By Position, 2003-2006|
Note that all figures may not add up to 100 percent or be consistent with official numbers due to rounding and omission of players who were targeted fewer than ten times.
The targeted by position splits make enough sense; in 2006, the absence of wide receivers and the presence of Ben Watson caused an increased focus on the tight ends; in 2003, Kevin Faulk and Larry Centers were targeted a total of 100 times. The similarities between the 2004 and 2005 seasons is pretty amazing when you consider that Deion Branch only played half a season in 2005. The averages there seem like a pretty good guideline to how the Patriots offense will shake out in 2007; Kyle Brady will see the ball less often than Daniel Graham did, and Kevin Faulk will see some of his touches usurped by Welker and Sammy Morris. Since the Patriots will be winning to run a whole lot in 2007, that pass attempts total seems just about right.
|Patriots Target Percentage By Role, 2003-2006|
Again, if we look at how Tom Brady split his passes over the previous four years, they're pretty consistent across the board. Last year, the secondary wideouts lost a good amount of targets that were directed towards Ben Watson, but Reche Caldwell was still thrown 102 passes, many of which he could actually see clearly.
The question here is who necessarily will fill the roles above at wideout. Wes Welker is commonly thought of as a replacement for Troy Brown; Brown was the WR2 in 2003 (60 targets) and 2006 (76 targets), but was the WR4 in 2004 (29 targets as he played a lot of defensive back) and WR3 in 2005 (59 targets) when the Patriots had Deion Branch and David Givens to work with. With the presence of Stallworth and Moss as likely receivers on the line, Welker's probably going to see a lot of time in the slot. That would peg him as somewhere between a WR2 and WR3 for the Patriots in 2007.
With all that in mind, let's try and project the Patriots wide receivers' performance in 2007. We'll base their catch percentages and yards per catch on their average for the last three seasons, giving the new arrivals a generous (based upon the data in Question 2) two percent increase in catch percentage for playing with Tom Brady in a solid, well-rounded offense. We'll say that Moss and Stallworth are WR1A and 1B, and that Faulk and Morris will be RB1A and RB1B. (It should also be noted that RB1 is referring to their usage in the passing game, not on the overall depth chart.) We'll give the touches Daniel Graham saw last year to David Thomas, and give most of Kyle Brady's touches to Ben Watson since Brady will be used almost entirely for blocking.
|Expected Patriots Performance, 2007|
Throw in about 10 catches on 20 throws from assorted flotsam and jetsam and you have a vintage Tom Brady peak season: 327 completions on 547 attempts, and a 60 percent completion percentage (owing to the deeper routes run by Stallworth and Moss), with around 4,200 yards passing.
(Ed. note: We should point out here that these projections are done solely in the context of this article, and are not the same as the KUBIAK fantasy football projections for these players that will be published in PFP 2007 and available on the site in the next week or two.)
Obviously, the above projection assumes that the Patriots offense will remain injury-free in 2007, an extremely unlikely proposition. An injury to Brady or Maroney would be catastrophic, but a much more reasonable expectation would be to slice 2-4 games off of the projections of Moss and/or Stallworth, both of whom have proven themselves injury-prone. Even if you gave 20 percent of each of their targets to Welker, getting to 95 attempts -- a figure Troy Brown hasn't come close to since the pre-Deion Branch days -- would peg his 2006 production as his peak performance, not the precursor to it. It's not that Welker won't be valuable or a worthwhile use of a roster spot; instead, Welker has become, in a sense, a victim of his own success.
68 comments, Last at 12 Jun 2012, 4:10pm by Anonymous2012