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?
11 Oct 2005
by Ned Macey
If a 33-10 score can be not as close as it appears, that was the case in Dallas on Sunday. The Philadelphia Eagles absolutely laid an egg, not gaining a first down until the last play of the first quarter, at which point the Cowboys had run up 17 points. Through four weeks, the Cowboys had managed a 2-2 record by beating a San Diego team without Antonio Gates and by overcoming a 12-point deficit in the fourth quarter against woeful San Francisco. It was hard to imagine such a team beating the high-flying Philadelphia Eagles, who had put up more than 100 points over their past three games. The Cowboys not only won, but they dominated the Eagles, controlling both sides of the line of scrimmage. One bad game is no reason to write off the Eagles, but they head into their bye week with an injured quarterback, a bad habit of slow starts, and a usually reliable special teams unit in disarray.
For the Cowboys, the whole season seems to rely on whether or not they can keep Drew Bledsoe upright. As Philadelphia learned, all the Pro Bowl members of the secondary in the world do not mean much if Bledsoe has time to throw. Not only did the Cowboys prevent any sacks, but Bledsoe was almost never pressured. In his three years in Buffalo, the Bills were always one of the worst teams at preventing sacks according to our adjusted sack rate metric (which measures sacks per pass play adjusted for down, distance, and opponent). The man likes to hold the football so he can make big plays. Coming into the game, the Cowboys' offensive line ranked a mediocre 20th in adjusted sack rate. Holding the Eagles defense that ranked seventh before the game to zero sacks is an impressive feat, albeit one that defies logic.
A crucial component to the Cowboys' early season success is that, when he has time to throw, Bledsoe has multiple quality options. A year ago, an injury to Terry Glenn left the Cowboys with Quincy Morgan as the second receiver in Dallas. Needless to say, that did not work out too well. He had a DVOA of -21.7%, which ranked 77th out of the 84 receivers who were targeted on 50 or more passes. The lack of a second receiver on the outside left Vinny Testaverde throwing over and over to Keyshawn Johnson and tight end Jason Witten. Both players performed well; Johnson was the 20th most productive receiver, while Witten was the third most productive tight end according to DPAR. (DVOA, which measures value per play, and DPAR, which measures total value, and the rest of the Football Outsiders' innovative stats are explained here.)
Despite good numbers, neither Johnson nor Witten is able to stretch the field, something that Glenn and second year player Patrick Crayton do quite well. Glenn is averaging an exceptional 22.5 yards per catch, and 19 of his 23 catches have netted first downs or touchdowns. By our advanced measurements, he has been the third most productive receiver in football so far this year. In Dallas' first drive, Bledsoe went to Glenn on all three of his completions, which netted 49 yards and a touchdown. Two possessions later, when the Cowboys started on the Eagles 38-yard line, Bledsoe hit Glenn again for a touchdown. When the Eagles adjusted to the deep threat posed by Glenn, Bledsoe was able to successfully work underneath, hitting Witten or Johnson for three straight completions that got the Cowboys to the Eagles 11-yard line on their next possession (at which point not the Eagles defense but penalties stalled the drive).
Defensively, the Cowboys effectively blitzed McNabb, who clearly is hampered by his sports hernia injury. Of the Cowboys' four sacks, two came from defensive backs. With McNabb unable to scramble â€“ not counting kneel downs, he has only five rushes on the year â€“ teams are less likely to get into trouble by sending aggressive blitz packages. Any idea that McNabb needs to run to be effective is ludicrous, but if injuries make him about as mobile as Peyton Manning in the pocket, it does free up defensive coordinators to take more chances. Going into Sunday's game, the Eagles had done an excellent job of protecting their injured quarterback, so this may be an aberration, but the willingness of the Cowboys to send defensive backs is something sure to be copied by future opponents.
With McNabb hurting, and the Eagles attempting a total of 20 rushes in the last two games, the general consensus is that the Eagles need to run more. Given the makeup of the Eagles, such conventional wisdom is too simplistic. Against the Cowboys, additional running plays would have been sound because according to DVOA, they have the eighth best pass defense but only the 24th best run defense. No matter what your offensive philosophy, you have to be mindful of your opponents' strengths and weaknesses. Their other division rivals, however, are actually stronger against the run than the pass, so the Eagles will not be playing to their opponents' strengths on a weekly basis. The Eagles have succeeded as a pass-first offense for the last five years, and one game against a team particularly suited to defend the pass is not a reason for them to change their offensive strategy.
The Eagles may not need to run more overall, but they should try and run more in the first quarter. A year ago, the Eagles ran 70 times on 197 first quarter plays, just under 36% of their offensive plays. That total still represents a pass-heavy offense, but one that includes just enough running to be extremely effective. Last season, the Eagles were the best first quarter offense in football, with a DVOA of 58.4%.
This year the Eagles are struggling mightily in the first quarter, with a DVOA of -16.4%, only 24th in the league. To date, they have run on only 20% of first quarter plays. In 2004, they were by far at their best in the first quarter; in 2005, it is the only quarter where the Eagles have a below average offense. In later quarters, the Eagles rushing attempts start to normalize (by their standards anyway). In quarters two through four, they are actually more potent than a year ago, but because of the extreme fall-off from the first quarter, their overall offensive efficiency is roughly the same as last season.
One reason the Eagles cannot run more is that they do not have the personnel to feature a conventional rushing attack. The Eagles unwisely thought that the perpetually injured Correll Buckhalter would be healthy. Buckhalter got hurt in the preseason, again, leaving the Eagles without a conventional back. Brian Westbrook is an extremely valuable player, but he is just not a player you can pound into the line 25 times a game.
A year ago, the Eagles had the ancient Dorsey Levens on their roster, and he was extremely effective in his defined role. Remember, DVOA is a measurement of the relative success of plays, and while Levens was probably about the 50th best NFL running back, he was more productive in those plays where he carried the ball than all but two other backs in football. Andy Reid apparently did not notice this himself, giving Levens one measly carry in the Super Bowl. This year, with Buckhalter hurt again and Levens a year older, the Eagles brought in Dolphins cast-off Lamar Gordon. Taking out the San Francisco game where Gordon got garbage time touches, he has a total of seven rushing attempts in the four other games. In his very limited opportunities, however, he has done well from a DVOA perspective. Given that Duce Staley was also extraordinarily effective in 2003 in the Levens/Gordon role, it would behoove Reid to consider using Gordon more often. This isn't about "establishing the run." This is about running a variety of plays so that the defense has to think a little bit.
Exacerbating the Eagles' first quarter offensive struggles is a defense that has struggled in the first quarter for several seasons. A year ago, the Eagles ranked 20th in defensive DVOA during the first quarter. This year, they rank 29th. Put offense and defense together, and the Eagles have been outscored 48-14 in the first quarter. While they should run more in the first quarter, and their early-game defensive struggles have been going on for over a year, even an analyst who relies heavily on statistical information has to concede that the Eagles are not â€œgetting upâ€? for games, an explanation I usually find wanting. The Eagles are too good a team to struggle so badly in any quarter to this degree, and the fact that it is the first quarter seems to indicate they are not ready to play from the opening kickoff.
An even bigger problem is Philadelphia's special teams. In our book Pro Football Prospectus 2005, we included a research piece on the strength of the Eagles' special teams. Special teams are notoriously fluky from year to year given the smaller number of plays measured and the possibility for major events (kick returns for touchdowns, fumbles) to skew the data. But since the Eagles started their run in 2000, they have never ranked lower than sixth in our special teams rankings. Last year, the Eagles had a negative special teams rating in only two of their 16 regular-season games. This year, they have posted a negative value in all five games.
The injury to David Akers, the best all-around kicker in the NFL, has hampered their special teams numbers, but their failures have occurred across the board. By our measurements, due to poor field goal kicking and poor coverage and return units, special teams have cost the team 22.6 points worth of field position and field goals, compared to the NFL average.
On Sunday, these weaknesses were apparent in the decisive first quarter. The Cowboys returned the opening kickoff to their own 49-yard line. After a three-and-out, the Eagles punted from their own 11-yard line but only netted 32 yards. After another three-and-out, the Eagles' punt from their own 1-yard line netted only 38 yards. Following a decent kickoff return by the Eagles, they went three and out again. This time, the punt netted a paltry 16 yards. Finally, after another Cowboys score, they returned the kickoff only 12 yards to their own 24-yard line.
Based on the first four games of the season, a Cowboys blowout was impossible to predict. For the Cowboys, however, it was a glimpse of what is possible. A Drew Bledsoe with time is still a very dangerous quarterback. Keeping Bledsoe on his feet, however, is not an easy long-term task, and going into this game, the Cowboys were only average at it. The weak run defense faces Tiki Barber and Shaun Alexander the next two weeks. However, both the Giants and Seahawks present serious opportunities for the Cowboys passing attack. If they can outscore their opponents the next two weeks, the articles about Bledsoe's resurgence will be plentiful. Whether or not he can maintain it for a whole season will be the test that determines whether the Cowboys are playoff bound.
For the Eagles, their inconsistent early season play could be disregarded except for the potentially huge problem of McNabb's injury. For now, McNabb has only sacrificed his running ability, an increasingly insignificant part of his game in any case. If, as reported, the pain will keep getting worse, McNabb may start to see decreased productivity. Conventional wisdom is that the Eagles need to run the ball more to protect McNabb, but short of a few additional carries for Lamar Gordon, there is little the Eagles can do besides jump out to early leads. Before the season, Philadelphia was thought to have one of the league's easiest schedules, but the resurgence of the NFC East makes things a lot more difficult. Despite their problems, the Eagles are too good not to win their division if they can keep McNabb relatively healthy and manage even an average level of play in the first quarter.
Each Tuesday in Any Given Sunday, Ned Macey looks at the biggest upset of the previous weekend. The NFL sells itself on the idea that any team can win any given game, but we use these upsets as a tool to explore what trends and subtle aspects of each team are revealed in a single game.
31 comments, Last at 16 Oct 2005, 6:23am by Anthony Brancato