Are the best defenses against play action the best against regular passes too? How much impact does play action really have in an NFL game, and does it correlate from year to year?
18 Oct 2005
by Ned Macey
The cover of Pro Football Prospectus includes a picture of Michael Vick along with the caption: "He's not the reason Atlanta wins." Never was this more true than yesterday.
In a week where the only favored team to lose was Pittsburgh playing without Ben Roethlisberger and Hines Ward, we divert from our customary look at the week's major upset to look at an extremely close call in San Antonio. Atlanta made just enough plays and got just enough help from the officials to escape New Orleans. The small margin of victory in a game matching a supposed Super Bowl hopeful and a team just waiting for a nightmare to end clearly illustrates how two of football's supposed stars, Vick and Deuce McAllister, are among the most overrated players in football.
The Falcons went into San Antonio coming off a depressing loss to the New England Patriots. With their star quarterback back, even at less than 100%, they seemed ready to roll over a Saints team that lost their supposed best offensive player the week before to a season-ending knee injury. Instead, the teams battled in a tense affair where neither side ever led by more than seven points. The Falcons were in the game thanks to a blocked field goal returned for a touchdown and an interception that got them the ball at the Saints 11-yard line. In the end, it looked like they were headed for overtime until a questionable call gave Todd Peterson a second shot at a game-winning field goal which he converted for the win. How did a superior Falcons team get their â€œbestâ€? player back and need an official's flag to beat a Saints team without their â€œbestâ€? player?
To start, we have to look at the deficiency of Mike Vick in the passing game. One week after the unknown Matt Schaub threw for 298 yards and three touchdowns, Vick returned from his injury to go 11-for-23 for 112 yards with one touchdown and one interception. To date, Vick has only completed 55% of his passes for a paltry 6.07 yards per attempt. By our more advanced metrics, which measure each play against the average play in that situation and are further described here, he has a DVOA of -8.6%, good for 25th best in the league. A year ago, Vick had an even worse DVOA of -24.9%, which ranked 36th in the league, right between Ken Dorsey and Eli Manning. None of this suggests that Matt Schaub should be the starter. Outside of Todd Bouman's family, I think people understand that a one game performance by a backup quarterback can tend to be a bit of a fluke. Still, Schaub's showing proves that the Falcons' offense does provide opportunities to make plays in the passing game.
Vick's struggles as he matures are somewhat surprising since when he first took over as the starter, he was as deadly with his arm as he was with his legs. In 2002, his DVOA throwing the ball was 18.5%, good for eighth in the league. Unfortunately, he has not built on this success, and the West Coast system installed by coordinator Greg Knapp has stunted his development. A major part of the problem is that with increased trust from his coaches, Vick has gotten into the habit of holding onto the football, convinced he can make a play with his legs. In 2002, he was sacked 33 times in 454 dropbacks. Last season, he was sacked 46 times in 367 pass attempts. So far this year, Vick's sack rate is heading back towards 2002 levels, so there is some reason for optimism that he can become a valuable passer again.
The obvious response to criticism of Vick's passing is that so much of his value is tied up in his running ability. Such logic has led to criticism that DVOA does not fully appreciate Vick's talents. This critique could not be further from the truth, as Vick's amazing 11-yard scrambles on third-and-9 get full credit in the system. His rushing DVOA is truly impressive, a robust 41.4% so far this season. Since he runs much more than other quarterbacks, it is useful to translate this number into DPAR, which measures points above a replacement player. Vick has contributed a DPAR of 12.3, easily best in the league. Such dominance was apparent a year ago, when he posted a 25.9% DVOA, third in the league, and 29.3 DPAR, more than twice the total registered by the second most prolific scrambler, Daunte Culpepper.
The problem for Vick, however, is that he is asked to pass the ball a lot more than he is asked to run it. DPAR is also calculated for passing, and by that measure, Vick has contributed only 2.1 points above replacement. This total is much improved from a year ago, when he was a below replacement level passer with a DPAR of -18.5. If you combine his rushing and passing DPAR from a season ago, his total value was 10.8. That combined total tied him with Josh McCown for the 25th most productive quarterback in football. If Vick can keep his passing just above replacement level, as he has done to this point, his running prowess will likely keep him within the top 20 quarterbacks. With 32 teams in the league, that makes him an average starting quarterback.
Vick's proponents argue that statistics undersell his true â€œintangibleâ€? abilities, but a quarterback's team success is dependent on the talent around him. Supporters focus on Vick's impressive record as a starter. Following Marc Bulger's loss on Monday Night, Vick now has the third best winning percentage among active quarterbacks with at least 35 starts. If you need proof that winning percentage does not tell the whole story, the seventh best record among active quarterbacks belongs to the immortal Jay Fiedler. The fickle nature of wins is also demonstrated by Bulger, who after starting his career 18-3 has gone 10-11 in his last 21 starts. Did he forget how to play quarterback, or did his defense collapse around him?
Since Vick took over as the full-time starter in 2002, the Falcons are 3-11 in games that Vick has not played. Most of these games, of course, were in 2003, when Doug Johnson was Vick's replacement. What this record proves more than anything is that Mike Vick is a lot better than Doug Johnson (he of the -22.8% DVOA that season and no scrambling ability). I think we can all agree on that.
Last year, the Falcons' offense ranked 24th in the league with a DVOA of -6.2%. Their overall DVOA was 2.9% due to an average defense and exceptional special teams. Their 11-5 record was very much a reflection of their poor level of competition. They played only three teams all season that finished with a winning record. Even in the playoffs, they made it to the NFC Championship game by beating an 8-8 Rams team. In the disastrous 2003 season, not only did they have the Doug Johnson experience, but their defense ranked 26th in the league. Their running backs, leaving aside the threat of Vick, were roughly equivalent to 2004. The duo of T.J. Duckett and Warrick Dunn combined for 27.2 DPAR on 323 runs in 2003, not much behind the 30.3 they amassed in 369 runs a year ago. With a tougher schedule and the return of the 2003 version of the defense, the Falcons will actually have to rely on Vick to make a repeat run at the division crown.
The division crown seems unattainable to the Saints, but their season is not ruined by the season-ending injury to McAllister. To be a valuable running back in the NFL, you either need to provide substantial value every time you touch the ball or be able to carry the ball extensively at an average level. The great backs - like LaDainian Tomlinson and Edgerrin James - do both. Deuce McAllister does neither. We use DVOA to measure the first of these skills, since it calculates a player's value on a per play basis. Through the first five games of this year, McAllister came in with a below average -2.6%. That total is sadly better than his -6.2% of a year ago that ranked 36th in the league. To credit the overall value of the volume of McAllister's carries, we use DPAR, by which measure Deuce has a mediocre 4.6 so far this year, again an improvement on last year where he totaled 8.9 in 14 games, good for 33rd in the league.
McAllister has long been a running back that we at Football Outsiders do not really value because of his boom or bust style. We keep track of a running back's success rate (explained in detail here), by which standard McAllister has always been terrible. In 2003, he ranked 45th in this measure out of 53 backs. Last season, it was a not much better 36th out of 52. McAllister has always been a back who routinely gets stuffed at the line of scrimmage, leaving error prone quarterback Aaron Brooks in numerous third-and-long situations. Occasionally, McAllister would break a long run, thus creating some value for his team. Sadly, since his overuse in 2003, those long runs were coming less and less frequently. In 2003, he had a run of at least 20 yards on 4.6% of his runs. Last year, that number was down to 1.9%, and in the early going this year, it was at 1.1%. This development has made a player who was probably overrated even back in 2003 among the least productive every-down backs in football.
The Saints organization clearly does not understand the difference between a good back and a 1000 yard rusher, as they brought in the itinerant Antowain Smith to be McAllister's primary backup. Smith posted a DVOA of -16.3% a year ago in Tennessee, and even in winning Super Bowls with the Patriots, he posted a DVOA of -5.8% in 2003 and -11.7% in 2001.
Even with his limited skills, Smith gained 88 yards on 12 carries against a suspect Atlanta run defense that ranks 30th in the league according to DVOA. Smith's production was matched by the pedestrian Aaron Stecker, who in his first action of the season rushed 18 times for 85 yards. Neither Smith nor Stecker (DVOA of -19.7% a year ago) is a quality back, but even so, the drop-off from McAllister is far from severe. McAllister has averaged only 3.9 yards per rush since the start of the 2004 season, and it seems reasonable to imagine the same sort of production from the tandem of Smith and Stecker. At least with those two, offensive coordinator Mike Sheppard will know he does not have a great running back and will stop pounding the back ineffectively into the line to establish the run. In the five games McAllister played, Sheppard's attempts to establish him in the first quarter produced a total of 71 yards on 26 carries.
At 2-4, the Saints are a long shot to get back into the playoff race, but the problem is not Deuce McAllister's injury. The other 52 members of the team are just not very good. They currently rank 27th in overall DVOA and have a below-average offense and defense. For an added bonus, they have the second worst DVOA on special teams. The offense exploded for the first time all season against Atlanta thanks to the success of Smith and Stecker. Such a performance will not be consistently repeated, but at least Saints fans can hope that the preseason idea to pound a running back into the line for no gain has been shelved. The Saints are in for a long season, but the absence of Deuce McAllister will have very little impact on their final record.
In the rejuvenated NFC (well at least rejuvenated NFC East and South), playoff teams will likely need 10 wins. With a defense that has completely regressed from a year ago, more pressure will fall upon Michael Vick's shoulders to get the Falcons there. He retains the strong support of the running game, but the overall offensive DVOA of -2.9% a year ago will not get it done this year. So far, the offense has delivered with a 19.1% DVOA that ranks 5th in the league. The return of the 2002 Vick would make the Falcons a serious threat in the NFC. If the Vick the Falcons get the rest of the season is Sunday's version, their early offensive success will not be maintained. In that case, Vick's winning percentage -- and his reputation -- will take a hit.
Each Tuesday in Any Given Sunday, Ned Macey looks at the biggest (near) 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.
70 comments, Last at 31 Oct 2005, 6:14pm by Jason