In this week's Varsity Numbers, Bill Connelly takes a page out of baseball's playbook and attempts to isolate power from efficiency.
05 Apr 2010
by Aaron Schatz
Did the Eagles strengthen a division rival when they traded quarterback Donovan McNabb to the Washington Redskins yesterday? Or is McNabb an aging quarterback in decline, who may no longer be as good as his heir apparent, Kevin Kolb? Actually, both of those statements may be true.
To guess at what the future holds for McNabb in Washington, I looked at the most similar players using FO's similarity scores system. (That system is explained here.) The most similar single seasons belonged to Warren Moon in 1989, Joe Theismann in 1983, and Ken Anderson in 1981. The ten most similar three-year spans to McNabb from 2007-2009 include:
That's a good collection of Hall of Famers, Pro Bowlers, and Jake Delhomme. Even though they were in their early thirties, the players on this list generally still had a few good years left. But when you look at their numbers, you wonder whether McNabb will give the Redskins that much more than they got from Campbell. Let's imagine that McNabb's 2010 season looks like the average performance of those ten similar players, in the year following the three-year span listed above. Those players averaged 13 games, primarily because of injuries to McNair and Marino, but we'll pro-rate the averages to 16 games.
|QBs similar to McNabb||16||296||487||3411||22||14||60.8%||7.00||33||132||1|
|Jason Campbell 2009||16||327||507||3618||20||15||64.5%||7.14||46||236||1|
(Corrected; this originally said "Jason Campbell 2008.")
The "McNabb comparables" end up with a better touchdown-to-interception ratio, but not by much, and the yardage and completion percentage doesn't match what Campbell did last year. Yes, McNabb averaged over eight yards per attempt last year, but that was a yard more than what he averaged in 2007 or 2008, and he doesn't get to bring DeSean Jackson with him to Washington.
(Note: This is a quick-and-dirty, very imperfect way to forecast McNabb's 2010 season. I am hereby acknowledging that it is quick, dirty, and imperfect.)
FO's advanced play-by-play breakdown doesn't judge McNabb's 2009 season as being that impressive, despite that high yards-per-attempt figure. McNabb finished 20th in our DVOA ratings for quarterbacks, his lowest ranking since 2002, although the rating itself was similar to his performance in 2005 and 2007. Kolb had a higher DVOA rating than McNabb, although with a big asterisk, since he only played in three games. McNabb didn't even rank that much higher than Campbell, who came in about league-average and ranked 25th (out of 46 passers).
McNabb's problem in 2009 was third down. Overall, Philadelphia's pass DVOA ranked fourth among all teams on first down, 15th on second down, and 22nd on third down. That's actually a good sign that McNabb will play better with the Redskins. Our research has shown that third-down performance is a lot more variable than first- or second-down performance, so when a quarterback struggles on third down in one season, that actually suggests improvement the next year. After all, with basically the same offense, McNabb had the opposite splits in 2008: He was best on third down and worst on first down.
However, that comparison brings up the main problem with McNabb in Washington: He's not playing with the same offense in 2010. The Washington offensive line is significantly inferior to the Philadelphia offensive line. McNabb will probably have his blindside protected by a rookie left tackle, now that Washington doesn't need to use the fourth overall pick on a quarterback. The receivers aren't going to be as good either. Santana Moss is comparable to DeSean Jackson, but only if we're talking about the Santana Moss of 2003. Jason Avant and Jeremy Maclin have shown a lot more in their young NFL careers than either Devin Thomas or Malcolm Kelly.
Still, it's fairly safe to say that McNabb will be an upgrade over Campbell for Washington. Will Kolb be an upgrade over McNabb for Philadelphia? Well, it's hard for us to predict how good Kolb will be as a regular NFL starter. As noted above, he's played reasonably well, but you can't judge a guy on three games. Do we trust the Eagles' brass? On one hand, the Eagles' opinion of Kolb is a lot more educated than the public opinion. Only the Eagles can judge his knowledge of the playbook and command of the locker room. Most of the passes Kolb has thrown over the last three years have been in practice, and only the Eagles have seen them. On the other hand, plenty of quarterbacks have impressed in practice and melted down under real-game pass pressure.
For Philadelphia, this trade came down to the present vs. the future. When it comes to going after the championship in 2010, McNabb is certainly more of a sure thing. A 34-year-old McNabb is still a better bet than a 26-year-old Kolb with two career starts. But in a couple years, unless the Eagles are just completely wrong about him, a 28-year-old Kolb with two years of starting experience is going to be a much better bet than a 36-year-old McNabb. Kolb, Maclin, Jackson, running back LeSean McCoy, and tight end Brent Celek will all be 26 or younger this season. Those players will all be able to grow and develop together.
As for the Redskins, there's no question this makes them an improved offense for 2010. But McNabb isn't as much of an upgrade as people may think, and his ability to improve things is severely limited unless Washington can fill some of its other holes. And in a couple years, Washington fans may look back on this as yet another example where the Redskins made a big move for a veteran on the downside of his career. The difference is that the Redskins are getting McNabb earlier in his decline compared to those other guys.
(This article also appears today at ESPN.com Insider.)
142 comments, Last at 12 Apr 2010, 3:06pm by Kevin from Philly