Eight weeks ago, the Vikings were undefeated and ranked No. 1 in DVOA. Now they're 6-6 and probably done. What went wrong?
23 Aug 2004
by Al Bogdan
When Michael Vick broke his fibula in August, Atlanta still remained a somewhat popular sleeper pick to make the Super Bowl. Doug Johnson had looked impressive in limited time in 2002 and could be expected to fill in admirably until Vick was scheduled to return in Week 6. The Falcons also had what was supposed to be a very good defense returning to the field. Although Vick received a good share of the accolades after Atlanta won a playoff game in 2002, it was the Atlanta defense that carried the team through the season. Although Atlanta was #23 in yards allowed in 20002, the Falcons allowed the eighth fewest points in the NFL and ranked #12 in defensive DVOA (#10 against the pass).
As we know, however, Atlanta's 2003 reality was much different than this rosy outlook would have suggested. Johnson wasn't as good as advertised, but what really sent the Falcons into the basement of the NFC South was the total collapse of the Atlanta defense. The Falcons dropped to #30 in points allowed, dead last in yards allowed and #25 in defensive DVOA.
Atlanta's offensive decline should have been expected, but there weren't many signs of an impending defensive meltdown. Expecting Vick to return in Week 6 proved to be ridiculously optimistic. Doug Johnson's promising 2002 campaign was really just one fluke game against the Giants. On defense, however, the only significant off-season loss looked to be CB Ashley Ambrose. The Falcons managed to replace him with the competent Tyrone Williams.
So that got us thinking -- could Vick's injury have something to do with the Falcon defensive decline? At first blush, the idea may seem ridiculous. Two things on a football field couldn't appear less related than the performance of a team's quarterback and its defense. But, as the saying goes, the best defense is a good offense. A good quarterback will run an efficient and effective offense, taking some pressure off of his team's defense. If the quarterback can sustain longer drives, that means fewer minutes per game that the defense has to be on the field getting tired. If a quarterback goes down suddenly to an injury, his team's defense will now be relied upon more to carry the team. Maybe the defense crumbles under this increased burden, or is just exposed as being overrated because of their team's offensive success.
To examine whether there is any relationship between a quality quarterback missing significant time to injury and his team's defensive performance, a table of quarterback seasons was created, using the wonderful Pro-Football-Reference.com statistical database. An attempt was made to gather the appropriate population of quarterbacks for this study objectively, so no one would have to scour old newspapers searching for injury information for every quarterback that has played in the last twenty years.
The table consisted of all quarterbacks since 1983 who led their team in pass attempts and had a passer rating over 75.0 in year X-1, but who threw 50% or fewer passing attempts in year X. Setting a minimum passer rating would hopefully eliminate a number of players who lost their jobs in year X+1 because of poor performance and not because of injury. Sixty-eight quarterback seasons met these criteria. The player's team's defensive rank in both points and yards allowed in year X was compared to its rank in year X-1. The offensive rankings of the team were also compared, as sort of a control group to ensure that the study's population was accurate. You'd expect that when a quality starting quarterback goes down to injury, his team's offensive performance would decrease as well. The results can be seen below.
|QB Rating >75|
|Year X-1||Year X||Difference|
|Avg. Rank Pts Allowed||13.3||15.5||-2.2|
|Avg. Rank Yds Allowed||14.4||14.6||-0.2|
|Avg. Rank Pts Scored||12.8||16.1||-3.3|
|Avg. Rank Yds Gained||13.4||15.6||-2.2|
This initial study indicates that there may be some relationship with a starting quarterback's "injury" and his team's defensive performance, at least with respect to points allowed. On average, when a quarterback throws 50% or fewer pass attempts than he did the year before, and he had led his team in attempts and had a passer rating higher than 75.0 in the previous season, his team's points allowed ranking dropped an average of 2.2 spots.
One would expect, though, that a team's offensive performance would be more affected by the unexpected loss of its quarterback than is indicated by the table. In this study, a team's ranking in points scored dropped only three slots on average suffering a quarterback injury. One would expect the drop off to be more severe than a mere three places. Maybe the player population needs to be adjusted.
Research was done to find out exactly which of the 68 quarterback seasons used for the original study actually missed significant time because of injury. Somewhat surprisingly, only 23 of the decreases in playing time could be attributed to injuries. These seasons are listed below.
|Injured Starting Quarterbacks|
|Terry Bradshaw||1983||PIT||Erik Kramer||1996||CHI|
|Jim McMahon||1984||CHI||Warren Moon||1996||MIN|
|Bill Kenney||1984||KAN||Brad Johnson||1998||MIN|
|Vince Ferragamo||1984||RAM||Jeff George||1998||OAK|
|Steve Bartkowski||1985||ATL||Warren Moon||1998||SEA|
|Jim McMahon||1986||CHI||Vinny Testaverde||1999||NYJ|
|Danny White||1986||DAL||Steve Young||1999||SFO|
|Don Majkowski||1990||GNB||Drew Bledsoe||2001||NWE|
|Randall Cunningham||1991||PHI||Kurt Warner||2002||STL|
|Bernie Kosar||1992||CLE||Mark Brunell||2003||JAC|
|Dan Marino||1993||MIA||Michael Vick||2003||ATL|
When the study was performed using just these 23 seasons, the results were drastically different.
|Injured Starting Quarterbacks|
|Year X-1||Year X||Difference|
|Avg. Rank Pts Allowed||13.9||14.8||-0.9|
|Avg. Rank Yds Allowed||15.7||15.0||-0.7|
|Avg. Rank Pts Scored||10.0||17.1||-7.1|
|Avg. Rank Yds Gained||9.1||16.0||-6.9|
The relationship between offensive performance and a quarterback injury is much closer to what one would expect in the second study. Since 1983, when a quarterback has a passer rating over 75 and leads his team in passing attempts in one season but throws 50% or fewer attempts the next year because of injury, his team drops on average seven places in both points scored and yards gained. If the Patriots weren't lucky enough to have Tom Brady coming off the bench in 2001, and if the Jaguars didn't have a healthy Fred Taylor gaining yards in 2003, this relationship would be even stronger than the chart indicates.
Limiting the population to players who were actually injured, though, appears to remove any relationship between a quarterback's unexpected missed playing time and his team's defensive performance. The 23 teams were about the same defensively from one year to the next, dropping on average less than one rank in both points and yards allowed.
So what exactly did we learn from this study? When a quality starting quarterback misses significant time because of injury, his team's offense declines, as one would expect. A quarterback injury doesn't appear to have much, if any, of an effect on his team's defense's performance. The numbers seem to agree with what common sense would tell us. We'll just have to find someone else to blame Atlanta's defensive struggles on, like Wade Phillips.