Writers of Pro Football Prospectus 2008

Most Recent FO Features


» Futures: Maurice Hurst

A heart condition discovered at the combine has put the Michigan lineman's career in limbo, but Hurst had the best film of any defensive tackle in this year's draft class.

19 Nov 2004

AFC Midseason Review

by Aaron Schatz (AFC North co-written by Ryan Wilson)

Time for a closer look at the first half of the season through the lens of our advanced Football Outsiders statistics.  Last week we did the NFC, this week we look at each team in the AFC.

I've got a lot of numbers for those who like statistics and a lot of commentary for those who don't.  But before we get to the commentary, I have to give the requisite explanation of the numbers.  For those new to Football Outsiders, DVOA (Defense-adjusted Value Over Average) takes every single play during the season and compares the result to the league average, normalizing for a number of variables including down, distance, location on field, opponent, and so on.  We use this statistic rather than simply yards or first downs, because DVOA rewards progress towards both.  DVOA is further explained here.

Remember that since DVOA represents efficiency at scoring points, defensive DVOA is better when it is negative. In total DVOA, you add in the opposite of defense so that better defense leads to a higher total DVOA.  Trust me, it makes sense.  The average DVOA, according to the baseline established in 2002-2003, is 0%.  However, you'll notice a few more positive ratings than negative ones this year.  As has been reported, this has been a big year for the offenses.  If it continues for the next eight weeks, I'll have to adjust the baselines in the offseason.

There is also a special teams rating that turns field position gained on kicks and punts as well as field goal efficiency into a percentage that can be added to offense and defense.  You can read more about how that works in this article, and you can see this year's special teams performances broken down here. All included numbers are through Week 10.  All statistics are given a rank from #1 (best in league) to #32 (worst in league). If the math confuses you, just ignore the stats and focus on the rank.

Just like last week, we've got DVOA and won-loss record projected for the rest of the season, similar to the complicated preseason projection system.  Projections are based on an analysis of midseason trends from the past four seasons and comparison of past teams with current teams based on first half strengths and weaknesses plus remaining strength of schedule. I run down the projected records after the section on each division, along with predictions for the two wild cards.  The projected DVOA ratings for the AFC are at the bottom of this article.  Unlike in the preseason, there are no subjective changes here (due to major injuries, for example).  Obviously, one bounce of the ball either way means an extra win or loss, and so these predictions are guaranteed to be less than 100 percent accurate.  Such is life in the NFL.

Someone noted last week that the projections are somewhat conservative. That's usually the way complicated regressions end up. In the NFL, where bad teams unexpectedly beat good teams all the time, I don't see a reason to quibble with conservative results. I did spread out the won-loss records a little bit so they were closer to matching the average distribution of wins among the 32 teams over the last few years.

If you find the comments about your favorite team interesting, the best way to help spread the word about Football Outsiders is to link to our articles and discuss them on your favorite team message boards.  I'll even give you some direct links for each division: AFC East, AFC North, AFC South, and AFC West.

These articles appeared in edited form earlier this week in the New York Sun.


The AFC East is a division of contrasts, home to both the defending Super Bowl champions and a team that imploded this season like a 50-car pileup from which nobody can avert their gaze. In between are a team whose strong offense has dragged a poor defense to a likely playoff berth, and team whose poor offense has dragged a strong defense away from one.

NE 8-1 41.1% 1 26.0% 6 10.6% 6 41.0% 4 -13.9% 3 -6.2% 10 -20.7% 3 1.2% 15
NYJ 6-3 14.8% 8 27.3% 5 21.0% 2 34.8% 7 14.0% 25 0.2% 17 26.7% 26 1.6% 14
BUF 3-6 -3.7% 19 -15.4% 29 -9.6% 27 -21.6% 29 -4.2% 9 -12.5% 5 4.6% 19 7.5% 3
MIA 1-8 -28.6% 30 -32.2% 32 -26.6% 32 -36.5% 31 -2.0% 12 -3.2% 15 -0.6% 13 1.7% 11


Thanks to their loss two weeks ago to the Pittsburgh Steelers, the Patriots are no longer considered the team to beat in the AFC, and for that they are thankful. Bill Belichick has never particularly enjoyed the media spotlight, and would clearly prefer to have the media heaping accolades upon the Steelers while the Patriots quietly win games. Nonetheless, according to our breakdown of every play this season, it is New England and not Pittsburgh that stands as the best team in the NFL this year.

In fact, this year's Patriots are better than the team that won a title last season because they are more balanced with the addition of running back Corey Dillon. Last year's running back committee of Antowain Smith and Kevin Faulk averaged 3.6 yards per carry; Dillon is averaging 5.0 yards per carry and is ranked fourth in the league by DPAR (Defense-adjusted Points Above Replacement). Dillon not only gets big chunks of yardage to extend offensive drives, he also forces opposing teams to defend both the pass and the run. That's freed things up for Tom Brady, which is why this has been the quarterback's best season ever. We can now dispense with talk that Brady "wins without great numbers" because this season he has great numbers (like his running back, he is fourth in the league in DPAR).

Conventional wisdom about the importance of a running game is usually overblown, but while a good passing game will always beat a good running game, a team with a good balanced offense will always beat both. Actually, the entire team is balanced. The Patriots are the only team in the NFL that ranks in the top ten of DVOA on rushing offense, rushing defense, passing offense, and passing defense. The only weaknesses of the 2004 Patriots are kickoff coverage and punt returns.

There was some concern early that the Patriots had declined against the run, but they have allowed more than 110 yards on the ground only three times this season, and they rarely allow long runs of double digits. The pass defense, meanwhile, has overcome a score of injuries and continues its strong play despite missing both starting cornerbacks and playing wide receiver Troy Brown as the nickel back.

While the public still believes that the Patriots live on luck, they were profoundly unlucky to lose those cornerbacks, Dillon, and both offensive tackles during what turned out to be the most important game of the year -- a loss to their prime competitor for home field advantage throughout the playoffs. Belichick can mouth the cliches about first needing to win the division and one game at a time and so forth, but it is clear that New England's true competition lives in the AFC North. Over the next seven games, New England must get one more win than Pittsburgh in order to host the AFC Championship game. Otherwise, even if the numbers say they are the better team, and even if the secondary returns to full health, getting past the Steelers and into the Super Bowl will be an uphill battle.

One more note: I did not mention this in the New York Sun, but the Patriots seem to do slightly worse than you would think in a lot of the indicators used by the projection system.  That projects them with only the seventh-highest DVOA in the league going forward.  It won't keep them from getting the first or second seed.


Before the season, the recipe for the Jets seemed clear. The team needed a healthy Chad Pennington to return the offense to the top of the league, while the young defense showed steady improvement. The early part of the season offered an easy slate of opponents that would allow the Jets to build up wins and confidence before things got harder in the second half.

For eight weeks, the recipe was followed to the letter, with the added ingredient of a career year from running back Curtis Martin (he's leading the league in DPAR despite being older than Wilford Brimley). But the last two weeks have changed the outlook for this team, which now has an injured Chad Pennington and two close losses that will make the race for a playoff spot a lot harder than it had to be.

The biggest pitfall the Jets face in the second half is not Pennington's injury, however, but the fact that their defensive improvement has been much smaller than people think, masked by a dramatically easy schedule. According to conventional NFL rankings, the Jets have a top ten defense because they have allowed only 311 yards per game. They have allowed only 5.1 yards per play, the same as New England and Philadelphia. But those numbers don't take into account the fact that the Jets have played the easiest schedule of opposing offenses in the league -- by a huge margin. According to DVOA, the Jets have played only one above-average offense, New England. But they have played twice against the league's worst offense, Miami, and twice against the offense ranked 29th, Buffalo. After taking opponent strength into account, a defense that looks good by conventional numbers still ranks as one of the worst in the league, despite improvement over 2003.

Now things get tougher, especially because of the order of the remaining games and where they will be played. The next two, against Cleveland and Arizona, would be easy victories at home with Pennington at quarterback; on the road with Quincy Carter under center, they become challenging. The Jets also must travel to St. Louis, where the Rams go from overrated to overachievers, and Pittsburgh, where the Steelers crushed two previously undefeated teams. Despite the tribulations that lie ahead, projections based on midseason trends from the past few seasons favor the Jets in the battle for a wild card spot. But in a conference where eight different teams have at least six wins, the room for error is minimal.


The Buffalo Bills are cursed both by history and geography. The former puts them in the remarkably competitive AFC, while the latter forces them to play twice each season against New England. If this team was in the NFC East, with a softer NFC schedule and a slate of opponents devoid of the Patriots, they would probably be a leading playoff contender. Buffalo's combination of quality defense with incompetent offense would certainly fit into an NFC that contains teams like Washington, Chicago, and Tampa Bay.

Buffalo's run defense has been stalwart all season, Sunday night not withstanding. Prior to this week, they had not allowed a 100-yard game to a running back all season. The Buffalo pass defense has improved over the past few weeks as well. The one weakness of the Bills defense is an important one, third downs.  DVOA says the Bills have had the league's fourth-best defense on first and second down, but they are ranked only 24th on third down.

Optimism about the offense is grounded in age rather than on-field performance. Quarterback of the future J.P. Losman still is not healthy enough to consider starting him, and Willis McGahee really hasn't been much better than Travis Henry. He has more yards because he has had more carries, and he has more carries because Buffalo's defense led the team to victories in three of the past five games and McGahee got to run out the clock. McGahee has averaged only 3.61 yards per carry and has been successful on 44% of his runs according to our Running Back Success Rate metric; Henry has averaged 3.45 yards per carry and has been successful on 43% of his runs.

Despite Sunday night's crushing loss to New England, Buffalo still gets a chance to play spoiler -- just not in their own division. Over the next two weeks they host St. Louis and visit Seattle, meaning they get to have some say in determining the winner of the NFC West. After those games are past, it will be time to sit Drew Bledsoe down, end his career as a starter, and begin the J.P. Losman era.


Every NFL fan by now is familiar with the litany of events that have made this season the most trying in Miami Dolphins history: offseason turmoil in the front office, wide receiver David Boston lost to a season-ending injury during training camp, the sudden retirement of running back Ricky Williams followed by his strange attempts to un-retire, the midseason resignation of head coach Dave Wannstedt. And yet, none of these are the biggest reason for Miami's pathetic season.

The task of rebuilding this franchise must begin not with a new quarterback, a new running back, or new receivers, but with the offensive line. Miami has tried two different quarterbacks and approximately 342 different running backs, and the same problems subsist because the linemen remain remarkably subpar. Only Detroit has seen its running backs stopped for zero yards or a loss more often, and only the Giants have seen their quarterbacks sacked more often. Miami's linemen are a mixture of career backups and underachieving youngsters, but the Dolphins need to emphasize the latter. They have seven games remaining to get 2004 first-round pick Vernon Carey some experience and to find out if second-year tackle Wade Smith will ever be able to develop into something better than the turnstile he has been through his first season and a half.

A number of NFL writers are recommending that the Dolphins hire a defensive-oriented head coach despite their offensive problems. But defense is not a problem for this team. The Dolphins had a top five defense last season and continued to have a top five defense in the early part of this year. They no longer have a top five defense not because their defensive players have forgotten how to succeed but because they are plum tired. Over the first five games of the season, the Miami Dolphins gave up only 4.2 yards per play, with five turnovers. Over the next four games they gave up 6.0 yards per play, with only one turnover.

The fact that offensive failure leads to defensive fatigue is also visible if you look at the Dolphins defense by quarter. Over the first nine games of the season, Miami has allowed 4.5 yards per play in the first quarter, 4.9 in the second quarter, 5.1 in the third quarter, and 5.5 in the fourth quarter. They also have caused no fumbles and intercepted no passes in the fourth quarter. By the end of each game, they are simply exhausted.

Believe it or not, there is hope here. If Miami spends its top draft picks and free agent dollars on offensive linemen that give their skill players enough time to be mediocre, they can ride their great defense back to respectability.

PROJECTED ORDER OF FINISH: New England 13-3, New York 10-6 (wild card), Buffalo 6-10, Miami 3-13.


Entering 2004, the AFC North looked like the least exciting division in football, a division which might send an 8-8 team to the playoffs. Instead, though its Ohio teams have been as mediocre as advertised, its other two teams are strong Super Bowl contenders.

PIT 8-1 33.3% 3 14.3% 10 3.6% 12 32.9% 8 -18.2% 2 -5.6% 11 -27.3% 1 0.8% 16
BAL 6-3 22.7% 5 -4.6% 21 0.8% 13 -11.1% 24 -19.9% 1 -15.8% 3 -23.7% 2 7.4% 4
CIN 4-5 0.5% 16 -0.9% 17 -4.1% 17 1.7% 17 1.2% 19 1.7% 20 0.7% 14 2.7% 9
CLE 3-6 -9.4% 23 -3.9% 19 -8.9% 25 2.0% 16 10.6% 24 4.8% 25 16.5% 24 5.1% 6


Philadelphia and New England are each 8-1 thanks to important new players like Terrell Owens and Corey Dillon, but for the most part they are the same as the teams that won so many games last season. What makes the 2004 Steelers so fascinating is that they went into this season needing to improve the running game, bolster the offensive line, get more consistent play from their quarterback, and add depth in the defensive backfield. Every one of those goals has been reached.

Quarterback Ben Roethlisberger has certainly been impressive -- his 7-0 record as a rookie starter is an NFL record -- but talk of him as an MVP candidate is a bit ridiculous. He's not the most important part of the Pittsburgh turnaround, or even the Pittsburgh offense. The pass defense has been the best part of this year's Steelers, and the running game the most important.

The Steelers have run the ball on 341 plays this year, by far the most in the NFL. Kansas City is the only other team above 300. 60 percent of Pittsburgh's offensive plays have been runs, and the league averages 44 percent.

Running the most doesn't necessarily mean running the best, but the Steelers have seen a colossal improvement over last year's rushing performance. Last season the Steelers were last in the NFL with 3.3 yards rushing per carry; this season they have 4.1 yards per carry. Part of that is the arrival of the underrated Duce Staley from Philadelphia, but a huge part of it is a healthy offensive line.

Left tackle Marvel Smith missed most of last season with a neck injury. Center Jeff Hartings played on bad knees. Right guard Kendall Simmons struggled after being diagnosed with diabetes. This season, Smith and Hartings are healthy and strong, combining with Pro Bowl guard Alan Faneca to form an impenetrable left side. On the right, Simmons was lost in training camp but backup Keydrick Vincent has stepped in and been far more effective than the Simmons who battled his health during 2003. Offensive coach Russ Grimm has molded these players into a unit that can make huge holes for Staley and Jerome Bettis as well as give Roethlisberger eons of time in which to find an open receiver.

But the improvement on the offensive line, great as it has been, is dwarfed by the improvement in preventing the pass. Last year, the Steelers had the 18th-best pass defense DVOA in the NFL. This year they are number one. Coordinator Dick LeBeau, creator of the zone blitz, has pressured the ball more this year than in seasons past. That means more sacks and more turnovers. The Steelers already have as many fumble recoveries as they did all of last season, with only three fewer interceptions and seven fewer sacks. In the secondary, the Steelers are the beneficiary of a huge step forward from last year's first round pick, safety Troy Polamalu, and a strong performance by rookie cornerback Ricardo Colclough, this year's second round pick.

New England has been better overall this season, but the Steelers won a huge advantage in the battle for home field advantage in the playoffs by handing the Patriots their only loss, and the projection system thinks they are the better team the rest of the way. Pittsburgh has to be considered the favorite to win the top seed in the AFC, win the AFC title, and -- because their strengths play directly to the weaknesses of the top NFC contenders -- reign as 2005 Super Bowl champions.


Going into 2004, my statistical projections frowned upon Baltimore's chances of repeating as AFC North champion. It was rare for a team to play defense or special teams at the dominant level of the 2003 Ravens for more than one season. The offense, meanwhile, had to expect a decline because past running backs who had run as often as Jamal Lewis had nearly all broken down the following season.

The Ravens still won't repeat as AFC North champion, but the cause is the Steelers and not themselves. Their defense isn't as dominating as last season, but it is once again the best in the NFL. Their special teams are still among the league's top five. And the Ravens are a better team than last year because the offense is no longer among the league's worst.

Kyle Boller has been the subject of much derision, but he does seem to be improving as Baltimore's quarterback. He has registered his top two games of the season over the past three weeks and exhibits a new sense of confidence since he found 6'6" rookie Clarence Moore as his go-to receiver. It helps that Moore's height neutralizes Boller's problem with throwing the ball too high.

Baltimore still has an offense built on the run, and while Jamal Lewis may have come close to the NFL rushing record last year, this year's Baltimore rushing attack has been more effective on a play-by-play basis thanks to improvement on third downs. Ironically, Lewis' suspension for pleading guilty on drug charges may turn out to be the best thing to happen to him professionally. It guaranteed him fewer carries, helping prevent a breakdown from last year's overuse and keeping him fresh in the latter stages of the season.

On defense, the Ravens have been the same disruptive force as years past, with one major difference: safety Ed Reed seems to be taking the title of the NFL's best defensive player away from his teammate Ray Lewis.

If the offensive improvement in Baltimore continues, the Ravens will be the strongest challenger to the AFC dominance of the Steelers and Patriots. But unless a Steeler collapse is in the offing, Baltimore will be limited to a wild card slot and the difficult task of winning three straight road games in order to return to the Super Bowl.


Analysis of the 2004 Bengals has to revolve around strength of schedule. It has affected the way the team's first half performance is misperceived, and it will affect their chances for making a second half run at a playoff spot.

Cincinnati seems to have returned to its inept offensive patterns of years past, but this team's first half schedule was littered with great defenses, including Pittsburgh, Baltimore, Washington, and Miami. By conventional NFL rankings, based on yards gained, Cincinnati comes out as the 23rd best offense in the league. Adjust for strength of schedule, and it turns out the Bengals have been far more ordinary -- in fact, the most run of the mill team in the NFL. According to our DVOA ratings, they rank between 14th and 19th in rushing offense, passing offense, rushing defense, and passing defense.

For the Bengals, that mediocrity is actually quite promising. There has been much gnashing of teeth over the fact that coach Marvin Lewis benched quarterback Jon Kitna, who had led a successful offense in 2003, in favor of last year's top draft pick Carson Palmer. Palmer's struggles have led to calls for Kitna to return, but Palmer's season must be understood in the context of Cincinnati's schedule. Adjust for the defenses he has faced and it becomes clear that Palmer is progressing nicely, albeit at a sub-Roethlisberger rate.

In 2003, the Bengals won six of seven games after the bye and nearly made the playoffs. Since this year's Bengals have three wins in their last four games, fans are optimistic that another second-half surge is in the offing. Unfortunately for those long-suffering fans, it isn't likely. The Bengals have the hardest remaining schedule in the NFL, including the three best teams in the league: Pittsburgh, Philadelphia, and New England. The latter two are on the road, as is another remaining game against Baltimore. This team is not going on a playoff run, and even a return to 8-8 will be difficult. At least the fans can enjoy watching Palmer continue to develop.


Some teams, like New Orleans, have maintained mediocrity by making very few roster changes. The Browns, on the other hand, have maintained mediocrity through constant roster changes, none of which seem to work very well. It can't be exciting as a fan when the peak of your team's season takes place in Week 1.

In Cleveland's opening victory against Baltimore, new quarterback Jeff Garcia looked like the Pro Bowl quarterback he had been in San Francisco, using his elusiveness to create time and connect on two big touchdown passes. Since then, Garcia has struggled with inconsistency, in large part due to a grueling slate of opponents.

No quarterback, not even Carson Palmer, has faced a harder schedule than Garcia. According to DPAR, Garcia's performance has been worth -5.0 points when compared to replacement level if you don't adjust for his strength of schedule. That number skyrockets to 18.8 points above replacement level after the adjustment, higher than Aaron Brooks, Matt Hasselbeck, or Jake Delhomme. It is one thing to draft a quarterback and have him learn from a veteran, but the Browns have no reason not to bring back Garcia as their starter in 2005.

The running game, on the other hand, is a total mess. The Browns expected big things from Lee Suggs after a 186 yard performance in the final game of the 2003 season. Instead, they have learned why sports statisticians harp on the issue of small sample size. DVOA says Suggs has been the worst running back in the league with at least 50 carries this season. Cleveland's other back, William Green, hasn't been much better.

On defense, the Browns have been unimpressive and uninspiring. They have suffered injuries up front, but no more than most teams, and their secondary has struggled mightily. Browns fans can take heart that statistical trends project some defensive improvement over the second half of the season, which would represent a promising step towards a better 2005.

PROJECTED ORDER OF FINISH: Pittsburgh 13-3, Baltimore 10-6 (wild card), Cincinnati 7-9, Cleveland 6-10.


The AFC South was expected to be the most competitive division in the conference this season, with two veteran Super Bowl contenders joined by two young teams that held the promise of breakout seasons. Three of these teams have kept up their side of the bargain, but in Music City they're humming the Crumbling Steve McNair Blues.

IND 6-3 27.0% 4 38.5% 1 8.3% 8 65.4% 1 6.4% 22 -3.9% 13 14.0% 22 -5.1% 30
JAC 6-3 -0.5% 17 1.2% 13 -5.7% 18 6.7% 14 -1.4% 13 -8.3% 8 5.0% 17 -3.1% 26
HOU 4-5 0.6% 15 -0.5% 16 -8.8% 24 7.7% 13 -1.1% 14 0.9% 18 -2.7% 10 0.0% 17
TEN 3-6 -15.0% 27 -12.5% 26 -3.8% 16 -19.2% 27 -2.3% 11 6.8% 26 -9.6% 7 -4.8% 29


We may all tell our children that we saw the greatest season a quarterback ever had, but for the Colts this means bupkis. Peyton Manning's amazing season will once again end in failure if the Colts defense cannot improve significantly over the next seven weeks. Unbalanced teams that are great on offense but terrible on defense, or vice versa, rarely make the Super Bowl.

It may look like the Colts defense is trending towards better performance because of last week's 49-14 tramping of Houston. But that was just one game. Looking at the Colts on a week-by-week basis shows a relatively straight line of poor defense, with two big blips: the Houston win, and a 35-14 victory over Oakland in Week 5. In both games, the Colts had three interceptions. In no other game have they had more than one, and in no other game has their defense been above average. The offensive recipe against the Colts is obvious: double team pass rusher Dwight Freeney, and scoring against the Colts becomes much easier. (The projection system, I should add, does think the Colts are headed for defensive improvement.)

Not only are the Colts behind two other teams in the race for a first round playoff bye, they are locked in a tight race with Jacksonville just to win the division. The Colts must curse the NFL schedule makers, who inexplicably decided to send Indianapolis on the road twice in five days. Chicago and Detroit should be teams the Colts beat easily, but there's a good chance that the weary Colts, with only four days to prepare, could lose to an inferior Lions squad on Thanksgiving. A loss there may mean the difference between the division and a wild card, or even worse, between a wild card and a January tee time.


Only three things in life are certain: death, taxes, and the fact that every year one NFL team will win far more games than their statistics would otherwise indicate. Last year, that team was Carolina, and this year it is Jacksonville. Every single Jacksonville win has been by six points or less, and not once this season have they risen into the top half of the Football Outsiders ratings.

Teams like this pose an important dilemma for the science of football prognostication. Can Jacksonville's success be attributed to "clutch performance" and "an ability to win"? Or is Jacksonville really a mediocre team that has gotten more than its share of lucky breaks? In the NFL, one errant bounce of a loose fumble or a pass off the receiver's fingertips often means the difference between a win and a loss. In a baseball season that lasts 162 games, those random breaks even out. In an NFL season that lasts only 16 games, there's a significant chance that an average team may finish 10-6, or 6-10. An attempt to analyze the NFL in the same way that our partners at Baseball Prospectus analyze baseball must accept the fact that even the most accurate advanced metrics will line up with wins and losses far less often.

Of course, if we spend too much time analyzing how Jacksonville's 6-3 record defies the numbers, we give short shrift to the players who are giving strong performances this year. Jacksonville has been very good defending the run for two years now. Many observers feel defensive tackle Marcus Stroud is the best at his position. That may be a bit of overstatement, but the players behind Stroud -- middle linebacker Mike Peterson and run-stuffing safety Donovin Darius -- are more underrated than Stroud is overrated.

Even if you get the lucky breaks that open the door to a close victory, you need a quarterback with the maturity to lead that game-ending drive, and Byron Leftwich has been unflappable. Leftwich's recent injury worried Jaguars fans, but David Garrard is one of the best backups in the NFL and so far has presented no drop off in quality.

Last season, Carolina defied the numbers to not only finish 11-5 but also come within a few minutes of winning the Super Bowl. But Jacksonville's plan to follow the Carolina storyline has two major obstacles. First, the 2003 Panthers ran away with a playoff spot while the rest of their division floundered, but Jacksonville is one of eight AFC teams with six wins or more so a postseason spot is far from assured. Second, Carolina underwent a dramatic transformation in last year's playoffs, suddenly playing far better than they had during the regular season. Even if the Jaguars can make the playoffs, they can't count on taking a similar leap forward.


With four wins in five games, Houston looked like it was finally shedding its expansion cocoon to take off as a beautiful, playoff-bound butterfly. Then they lost their last two games by a combined score of 80-27. Apparently, the very hungry caterpillar bit off more than he could chew.

But even a 7-9 season would qualify as a positive development for this franchise, especially because of signals that its young stars have finally matured. David Carr stumbled last week with the worst game of any quarterback this season, with three picks and three fumbles. Before that game, however, Carr had led a Houston passing offense that stood among the league's ten best. Andre Johnson has become a force at wide receiver, catching numerous highlight passes and forcing defenses to concentrate coverage on him. That opens up the field for Houston's other receivers, Jabar Gaffney and Derick Armstrong. Though he only has 24 catches for 358 yards, our ratings say that on a per play basis Armstrong has been better than any other receiver in the NFL with at least 15 pass attempts.  Both Gaffney and Armstrong, oddly, have higher DPAR than Johnson.

Houston hoped to be building a set of "triplets" to rival the Aikman-Smith-Irvin combination on the Cowboys of a decade ago, but the role of running back must be recast. It looks like Domanick Davis is the latest example of that unfortunate NFL institution, the one-year running back wonder (see Salaam, Rashaan and Stephens, John). Davis' 4.3 yards per carry average of a year ago has dropped to an abysmal 3.0 yards per carry, and the nadir of his season came when he could garner only 10 yards in 12 carries against a Kansas City defense that has been regularly torched by the worst backs in the NFL.  At least Texans fans can take some solace in the fact that Davis has been much better as a receiver.

The young players on the Houston defense are not as well known as the ones on offense, but they've been maturing too. Houston's defense was among the league's worst early in the season but has played its three best games of the season over the past four weeks (they are the only team to intercept Peyton Manning twice this year) and now ranks as above average.

Does this defensive improvement imply a return to the playoff race over the final seven weeks? Probably not, especially since the Texans have the hardest remaining schedule in this division. But Houston is set up nicely for a strong 2005 campaign, and is an early contender for the role of "suddenly dominant team" played this season by the Steelers.


Of all the dramatic changes of fortune in the NFL this season, none makes less sense than the collapse of the Titans. By the end of last season, the Titans could make a case that they were the best team in football, kept from the Super Bowl because they had to travel to Foxboro instead of hosting the Patriots in Nashville. This season, that great Titans team has disintegrated.

Some decline was expected from the defense, which faced a number of preseason injuries and lost veterans Javon Kearse and Robaire Smith to free agency. But with the exception of the Week 4 fiasco against San Diego, the Titans pass defense has been just as good as last season. Against the run, they've declined a little, but not as much as people think. Official NFL rankings proclaimed them number one against the run last year, but they didn't allow many yards because teams rarely ran against them. This year, teams run against them a lot more, because the Tennessee is playing from behind instead of from ahead. And the culprit there is the offense.

His entire career, Steve McNair has played through injuries, and always at a high level. After an MVP season in 2003, there was no reason to believe that something had changed and McNair had suddenly reached the breaking point. But with one exception, a great game when the Green Bay secondary imploded on Monday Night Football, McNair has been a black hole of offense in 2004. DPAR says that only Craig Krenzel, Jay Fiedler, and Ken Dorsey have subtracted more value from their team's offenses than McNair. The Titans have actually played better since McNair sat down to rest his injuries and Billy Volek took over.

Some of McNair's decline can be explained by the changes in the receiving corps: second receiver Justin McCareins was traded to the Jets, and the man who was supposed to take his place, Tyrone Calico, was injured early in the year. But those changes should have been balanced by the development of a quality running game to take some of the pressure off McNair. Tennessee replaced the broken down Eddie George with Chris Brown, who has cooled off somewhat from a quick start but still stands as a big improvement.

Tennessee will probably be better in the second half, but not by much. Why McNair collapsed is a big question, but the bigger question for the Titans is whether McNair can rebound in 2005. Otherwise, the Titans will need to find a new starting quarterback for the first time since they moved to Tennessee in 1997.

PROJECTED ORDER OF FINISH: Indianapolis 11-5, Jacksonville 10-6, Houston 7-9, Tennessee 7-9.


In the AFC West, touchdowns are more common than lobbyists in the halls of Congress and teams score as easily as Colin Farrell at a sorority mixer. But while the offensive numbers coming out of this division are no surprise, the team that is leading the offensive barrage is. San Diego has this season's most exciting turnaround and Kansas City its most bewildering. Meanwhile, Denver is consistent in its inconsistency, and Oakland despite its better judgment has in fact sent players onto the field for nine games.

SD 6-3 19.4% 6 21.3% 8 4.5% 10 41.2% 5 3.5% 20 -11.5% 6 14.1% 23 1.6% 13
DEN 6-3 10.8% 9 9.6% 11 -3.1% 15 23.5% 11 -2.4% 10 -3.3% 14 -1.4% 12 -1.1% 19
KC 3-6 17.2% 7 30.6% 3 23.6% 1 38.7% 6 7.3% 23 13.4% 29 2.3% 16 -6.2% 31
OAK 3-6 -29.7% 31 -7.8% 25 -7.4% 20 -8.0% 23 20.4% 27 3.8% 22 40.2% 31 -1.6% 20


By now, everyone is familiar with the Drew Brees story. Left as a lame duck after the Chargers drafted his replacement, he has turned his career around with one of greatest contract years in the history of sports. There's no better way to impress your future suitors than to lead the second-highest scoring offense in the league. Brees has an obscene ratio of 18 touchdowns to only three interceptions, and only Peyton Manning and Daunte Culpepper have a higher passer rating than Brees' 108.7. Last year his rating of 67.5 was 28th in the NFL.  By Football Outsiders numbers he's gone from -14.7 DPAR (41st) to 50.4 DPAR (8th)

As great as Brees has been, the real MVP of this offense is tight end Antonio Gates, the former Kent State basketball star who went undrafted out of college. Our breakdown of play-by-play shows that Gates has been worth approximately 26.2 points more to the San Diego offense than a replacement tight end (say, Cam Cleeland). Gates isn't just the most valuable tight end this season -- he has more DPAR than every single wide receiver except Terrell Owens and Hines Ward. Gates is the main reason San Diego has sharply improved its third down conversion rate this season, and that improvement is the main engine behind San Diego's 6-3 record.

The Chargers have turned things around despite a tough year from franchise running back LaDainian Tomlinson, who has struggled with a groin pull and dropped from 5.3 yards to 3.8 yards per carry. He's gone from second in our rushing ratings to 40th.  Backup Jesse Chatman has picked up the slack with a 7.2 yards per carry average, and if the bye week gave Tomlinson the rest he needed to get back to full strength, the Chargers become even more dangerous.

Will defense end San Diego's Cinderella season before they reach the ball? Not necessarily. The Chargers have been one of the NFL's best teams against the run, and they've been improving against the pass as well, with their last three games representing the team's three strongest defensive performances of the season.


The accepted narrative in Denver says that every year coach Mike Shanahan sticks an unknown into his backfield and comes out with a 1000-yard rusher (this year, former fullback Reuben Droughns). But the emergence of a mystery running back isn't the only plot thread that repeats in Denver year after year.

Last year the Broncos rid themselves of quarterback Brian Griese, who despite strong numbers had been inconsistent and unable to lead Denver past the first round of the playoffs. His replacement, Jake Plummer, has strong numbers but has been inconsistent and could not lead Denver past the first round of the playoffs.

Every year the Broncos produce a defense that is above average against both the run and the pass, but outstanding against neither. According to the Football Outsiders DVOA ratings, Denver's defense over the past five seasons has never been better than eighth in the NFL, and never worse than fifteenth.

Every year the Broncos have good placekicking and terrible punting that, because of the effects of altitude, looks like great placekicking and reasonable punting.

Denver fans can be forgiven if they think their team has been running in circles since John Elway's retirement. At least Denver's circles pass through the playoffs most of the time, which is better than the circles that New Orleans has been running in.

Will the second half bring more of the same? The good news is that Droughns has stabilized the ground game after the failures of Quentin Griffin, and the Broncos have an easier forecast schedule than the Chargers. The bad news is that the defense is on a significant downward trend. Over the first six games, Denver allowed the opposition 4.3 yards per play; over the past three games, that jumped to 6.1 yards per play.  The projection system seems to think the Denver defense is going to completely collapse to sub-49er levels -- it is by far the strangest result of the whole projection experiment.  I can't even seem to figure out why the numbers work out this way -- it looks like it is connected to the recent decline, inconsistency being a negative indicator, and the fact that Denver's worst DVOA comes against the pass on second down.  Why does that forecast such failure?  I'm still trying to figure it out.  Just remember that we were also scratching our heads when the projection system said the Chargers would have the sixth-best offense in the league.


It is a bit of an understatement to say that the 2004 season hasn't gone as planned in Kansas City. Break down every play of the season and compare each one to league average, and the Chiefs look like they should be one of the top teams in the NFL. In the Football Outsiders ratings, they rank higher than 15 teams with better records.

An offense like this should be blowing teams away early, but the Chiefs have done this only once, a 56-10 blowout of Atlanta. Instead, their season has featured close game after close game, and when the chips are down the Chiefs have been the victims of poor luck, poor decisions, and poor defense.

The Chiefs lost to Carolina when they drew a celebration penalty, and the Panthers took advantage of improved field position to march for the winning score on the next drive. They lost to Houston by throwing a red zone interception that was returned for a touchdown, and because Kris Brown hit a 50-yard field goal, a successful kick only 55 percent of the time. Lawrence Tynes missed a field goal and an extra point against Jacksonville. They racked up 497 yards against New Orleans but only scored 20 points because of four turnovers.

This year the unstoppable Kansas City offense was supposed to finally be complemented by an adequate defense thanks to the return of Gunther Cunningham, the coordinator of the league-topping Chiefs defenses of the late nineties. It turns out, however, that Cunningham's recipes could not turn lemons into Derrick Thomas-quality lemonade. Pass defense has been satisfactory, but an inability to pursue or tackle turns any run that gets past the front four into a long distance scamper, and only Minnesota has a worse defense when a game is within a touchdown in the second half.

No discussion of Kansas City's woes is complete without a mention of special teams. The Chiefs had the NFL's second best special teams unit in 2003 but the second worst in 2004. Last season, kick returner Dante Hall was the face of a new wave of multipurpose NFL stars. This season, Hall is the face appearing on milk cartons throughout Missouri.

You never know when Kansas City will suddenly flip the switch and begin winning games like their statistics suggest they should. A win Monday night, at home against a New England squad with a battered secondary, would be no shock. But the Chiefs need seven straight wins to even sniff the playoffs, and that isn't going to happen. Kansas City will be left wondering how they ran football's most dominant offense for three years and ended up with a solitary playoff loss to show for it.


On September 26, on national television, Oakland beat Tampa Bay in a game nowhere near as close as the 30-20 score. The Oakland offensive line pushed Tampa's vaunted defensive front at least three yards backwards on every single play. Kerry Collins, rejected by the New York Giants so they could start over with Eli Manning, led the Raiders to four scores and a missed field goal on his first five drives after replacing the injured Rich Gannon. While the legendary Tampa defense let Collins throw at will, the Oakland defense had Tampa quarterback Brad Johnson completely flustered.

When NFL Films produces the 2004 Oakland highlight film, it will consist solely of the broadcast of this game. At that point the Raiders were 2-1 and it looked like the winning Raiders of the past had returned from their 4-12 exile of 2003. The team of September 26 went into hiding immediately following that game and were replaced by a team that gives Al Davis nightmares.

Raiders lost five straight games before a narrow win over a Carolina team destroyed by injuries. Collins has been abysmal, with the lowest passer rating of any quarterback who has started half his team's games. Wide receiver Jerry Porter, expected to be a 2004 breakout, has instead petered out. High profile free agent Warren Sapp has done nothing to disrupt opposing quarterbacks and has a meager half a sack. Pro Bowl cornerback Charles Woodson has no interceptions and seems to have forgotten how to cover receivers. Their kick returns are worst in the league.

Can the Raiders take anything positive from this season? Well, wide receivers Ronald Curry and Doug Gabriel show great potential as deep threats. Offensive lineman Robert Gallery, second overall pick in last year's draft, has been as good as advertised. But it will take a few more standout youngsters before the Raiders can return to respectability.

PROJECTED ORDER OF FINISH: San Diego 10-6, Denver 9-7, Kansas City 8-8, Oakland 4-12.


  • These are projected DVOAs for Weeks 11-17 only, not for the season as a whole.  Projected record is for the whole season.
  • The special teams projections are really, really shaky.
  • Projections are strictly based on statistics and do not include the effect of injuries (cough, Quincy Carter, cough)
  • Strength of schedule is based on the average projected DVOA (not current DVOA) of remaining opponents.  It is not adjusted for home and road games.  It is ranked from hardest (#1) to easiest (#32).
  • Ranks are for all 32 teams, even though this table shows the AFC only.

2 PIT 24.0% 13-3 10.6% 8 -13.5% 2 -0.1% 17 4.0% 7
3 IND 22.0% 11-5 26.3% 2 0.2% 15 -4.1% 30 -2.5% 23
4 SD 21.3% 10-6 10.9% 7 -9.3% 5 1.2% 10 1.7% 10
5 NYJ 19.6% 10-6 19.1% 4 -0.1% 14 0.4% 12 0.8% 13
6 BAL 18.2% 10-6 -0.2% 16 -15.2% 1 3.2% 4 2.1% 8
7 NE 17.2% 13-3 11.1% 6 -5.8% 7 0.3% 14 0.5% 14
9 KC 13.6% 8-8 33.4% 1 14.3% 29 -5.5% 32 -2.8% 25
14 CIN 1.6% 7-9 1.0% 13 1.0% 17 1.6% 7 10.6% 1
15 CLE 0.2% 6-10 -7.6% 23 -4.9% 9 2.8% 6 2.1% 9
17 JAC 0.0% 10-6 0.2% 15 -1.2% 13 -1.5% 23 -3.7% 26
19 BUF -4.2% 6-10 -14.1% 29 -5.0% 8 4.9% 3 -2.4% 22
22 TEN -9.6% 7-9 -3.0% 20 3.8% 21 -2.8% 28 -4.3% 27
27 HOU -13.4% 7-9 -9.4% 27 3.0% 20 -1.0% 20 5.3% 4
28 DEN -13.8% 9-7 2.8% 12 15.6% 31 -0.9% 19 -2.2% 20
30 MIA -27.5% 3-13 -21.5% 32 6.2% 23 0.3% 13 1.0% 12
31 OAK -28.2% 4-12 -8.5% 26 17.7% 32 -2.0% 26 4.5% 6

Posted by: Aaron Schatz on 19 Nov 2004