website statistics
When writing about these rankings, I usually like to see where things differ from the mainstream public perception.  The mob mentality, made ever more powerful by social media and other forms of digital connectivity, lends itself to extreme (and often false) opinions.

Nonetheless, this predictive model is not necessarily perfect, so there are times when we need to step back and examine particularly puzzling rankings.  That's the case with the top two teams in the rankings this week: One is almost universally considered the best team in its conference, whereas most see the other as a likely 7-9 or 8-8 retread.  Let's take a deeper dive into these two teams, which should highlight a meaningful truth about this year's hierarchy.

The Top Dogs

One of these seems to make more sense than the other.

- It's been a while since the NFL has seen a truly legendary team exhibit all-around dominance.  It's a little too early to declare any team as likely to fulfill that distinction, but apart from a three-quarter blip at CenturyLink Field, the Denver Broncos might be one of the best teams in recent NFL history.

That's a vague phrase that needs more clarification, but the first-half evidence points in that direction.  After their 42-17 dismantling of San Francisco, the Broncos are now the top-ranked offense and defense in these efficiency rankings.  Their 0.68 Gross Winning Percentage, a measure of their true long-term winning percentage, is absurd.  The next highest mark is 0.61—for reference, the second-place team is as close to the 12th-place team as it is to the Broncos.

The Cowboys may be a half-game ahead in the standings, but that's merely due to bye week semantics.  Most advanced measures of team performance are publishing at around the same time as this, but pick your favorite measure of team efficiency—the Broncos seem likely to top just about everything, whether it be DVOA, Elo, Pythagorean win total, etc.

Peyton Manning may be the engine that drives the Orange Crush train, but even with a league-average quarterback, it would not be difficult to imagine Denver as a top-10 outfit.  After a rough start, the running game has shimmied up to near the league-average mark, with Montee Ball still to return in a few weeks.  The offensive skill position corps also holds a legitimate claim as the best in the league; in fact, the Broncos' offensive design typically asks Manning to simply get his receivers and tight ends the ball in space or in one-on-ones and allow their freakish athletic gifts to take over.

However, it's the defense that has made the Broncos a potentially historic squad, something other analysts have taken note of.  Denver ranks second in pass defense efficiency and ninth in run defense success rate, the only team that can lay claim to a top-10 ranking in both categories.  Veteran free-agent signings are often a recipe for disaster, but the trio of DeMarcus Ware, Aqib Talib and T.J. Ward has made as positive an impact as one could hope.  Injuries created plenty of defensive issues for the Broncos by the end of the 2013 campaign—lack of pass-rushing depth, coverage safety/linebacker, etc.—but those signings have essentially eradicated those worries.

If Denver can keep this pace up, they might begin to resemble a team many consider the most dominant in recent history, the 2007 Patriots.  Those Pats were undefeated at this time, of course, and they dazzled with a relentless and furious attack that made them a literally off-the-charts squad.  As you might expect, this year's Broncos don't really stack up yet (note that the Patriots graph is only for the first seven weeks of the 2007 season):

But those Patriots peaked in Week 10 with a ruthless 56-10 victory over the Buffalo Bills.  The Patriots still led the league from Week 11 on with a plus-107 point differential, but they regressed to a team that was only about 1.75 standard deviations above the mean in that category, rather than the otherworldly three standard deviations they were from Weeks 1-10.  That squad finished with a Pythagorean win expectancy of "only" 13.8, suggesting that they were rather fortunate to finish 16-0.

Denver is not anywhere near that territory at the moment; in fact, the Broncos' plus-11.3 point differential per game margin is less than that of either Indianapolis or Baltimore.  One could question the level of competition as well—whereas the 07 Pats blew out a future 13-win team and the team they would later face in the AFC Championship Game, Denver's best win is over the banged-up Niners, who may or may not even reach the postseason, given their schedule and the strength of the NFC.

Comparing these Broncos to that historic Pats squad is probably presumptuous at this point, at least based on the numbers.  But the eerily similar roster construction creates parallels that are hard to ignore—a prolific passing game led by one of the all-time great quarterbacks, an aging but well-rounded defense and an average running game.  In truth, the main difference probably stems from the fact that the Pats would have continued scoring against the helpless Niners defense, rather than sending in the Brock Osweiler brigade.

Unlike those Patriots, though, these Broncos are gaining steam, with a plus-60 point differential over their last three weeks.  Part of that stems from injury luck, as Denver's most damaging long-term injury has probably been to linebacker Danny Trevathan, who has been ably replaced by (the other) Brandon Marshall.  Unlike the other AFC contenders, the Broncos have yet to lose their Robert Mathis, Jerod Mayo or Nick Hardwick type of cog to a season-ending catastrophe.

Still, if injury luck is the best evidence to arguing for a Denver regression, then that essentially equates to throwing up one's hands and admitting that there is nothing really wrong at all.  The Broncos offense could always afford to drop off significantly from its 2013 heights and still remain the best unit in the league.  Now that the defense has become a similarly well-rounded unit, the Broncos look likely to further separate themselves from the pack over the next two months.

- The 3-3 Miami Dolphins may be third in their own division, but they sit in second place through seven weeks in these rankings.  This is obviously a curious placement, so let's dig deeper to see why AFA sees Miami as an NFL penthouse-caliber of a team, and whether or not we should expect their future performance to more closely reflect this ranking.

If the Broncos represent the clear-cut dominant team, then the Dolphins illustrate how the next tier of teams are fairly closely clustered together.  The Fins are not propped up by an artificially opponent-strength adjustment, as their 0.49 opponent GWP suggests that they have made hay against a below-average schedule.  Nor have the Dolphins goosed their own GWP with a bunch of junk-time scores.  Miami has run 47 offensive plays with a two-possession fourth-quarter lead this year, fourth-most in the league, but they have somehow not scored a single point in those situations.  That would suggest some legitimacy in their wins, as they are staking out to leads and playing their best while the game is still in doubt.

When I wrote about the Dolphins in the Week 4 version of this column, I noted how an elite defense was propping up their ranking.  That remains true, as Miami possesses the second-ranked defense in terms of efficiency, led by the best pass defense in the league.  However, while All-Pro level talents like Cameron Wake and Brent Grimes are safe propositions to bank on, I also argued that Miami's offensive inconsistencies made for an unsustainable long-term formula.  Adequate quarterback play is arguably the only indispensable ingredient in a winning team, and Ryan Tannehill needed to move the Dolphins out of the extreme "Good Defense/Bad Offense" EPA quadrant to take Miami seriously.

For the season, Tannehill still ranks just 26th in EPA per play and 31st in adjusted yards per attempt.  However, those low rankings stem mostly from a poor start, as it is clear that he has shown significant improvement in the three games since I last talked about the Dolphins:

This is essentially night and day.  Among players with at least 90 passing attempts over their past three games, Tannehill's 8.34 adjusted yards per attempt ranks seventh.  Conversely, over the first three weeks, Tannehill ranked 20th out of 21 qualifiers in that stat under the same minimum pass-attempt criteria.  It's been a timely turnaround for a quarterback whose job security was precarious just a month ago.

All together, that has left Tannehill with a seasonal mark of 6.4 AY/A, almost exactly in line with his career averages.  These types of brief streaks aren't exactly new for Tannehill—he has had five previous three-game streaks of an AY/A of at least 6.26, which is the run he is currently on.  That includes four last year alone, one of which culminated in arguably Tannehill's best career performance in a Week 15 win over the Patriots.

Of course, the Dolphins would lay eggs in their final two contests, and Tannehill's ghastly 2.55 AY/A mark over that fortnight was the worst in the league.  Tannehill has tantalized with his talent repeatedly, to the point that pro-Tannehill evidence looks like noise amid widely varying signals.

Here's where it's also worth noting that Miami has given up just five sacks over the past three games (12th-best in the league), as opposed to the nine they gave up over the first three games (third-worst).  Every quarterback is better with more time, but it has been magnified with Tannehill.  As a predominantly one-read quarterback, Tannehill has yet to demonstrate the acumen to progress through his reads while maintaining good pocket presence, a discouraging truth given his elite athleticism.  Bill Lazor's system works fine if misdirection consistently opens up that read, as Nick Folles will happily attest to, but much like Foles, Tannehill lacks the ability to stretch defenses past their first wave of coverage defense.

In facing Oakland, Green Bay and Chicago, Tannehill has feasted on a pair of bottom-10 defenses, while playing an above-average game against the sixth-ranked Packers defense.  This does not really prove as much as his early-season struggles, when he faced three defenses ranked in the upper-half of the efficiency rankings.  From Weeks 9-15, Miami faces seven upper-half defenses.  That will tell us much more about the long-term legitimacy of both Tannehill and the Dolphins.

Biggest Movers

One game this week may have revealed much about the AFC hierarchy.

- Apart from drafting Andrew Luck, the Indianapolis Colts have not found many keepers in building their roster over the past three years.  The ill-fated Trent Richardson trade came after a series of long-term contracts for marginal veteran free-agents, which has hamstrung the roster with oodles of non-damaging but relatively mediocre talent.  After an 0-2 start, some wondered if the Colts were even AFC South favorites anymore.

Since then, the Colts have ripped off five straight wins and now possess the second-best point differential in the league.  Two of those victories came over JV punching bags Jacksonville and Tennessee, but the Colts also posted a pair of impressive wins over fellow AFC contenders Baltimore and Cincinnati.  Their dominance over the previously second-ranked Bengals propelled Indy all the way up to third in these rankings, a 12-spot jump that made them this week's biggest riser.

These rankings have typically hated the Colts because of their reliance on close-game luck (no pun intended).  Indy is just 1-2 in one-possession games this season, but they had gone 14-2 in those games over Luck's first two seasons.  That was the second-best two-year stretch in NFL history, topped only by Luck's Indianapolis predecessor from 2008-09.

However, the Colts' 2014 claim to legitimacy rests not solely on Luck's right arm, but also a defense that is puzzlingly ranked seventh in overall efficiency.  The Colts mostly preserved the same 2013 defense that finished 16th in DVOA, except that they lost an All-Pro season from Robert Mathis and replaced the uninspiring middle linebacker platoon of Pat Angerer and Kelvin Sheppard with the uninspiring middle linebacker platoon of D'Qwell Jackson and Josh McNary.

And yet, this motley group of defenders has far outperformed expectations, holding three top-half offenses over the past three weeks to just 13.7 points per game.  That's deflated a bit by last week's shutout against the A.J. Green-less Bengals, but we cannot simply dismiss that performance, considering that the Bengals had averaged 6.34 yards per play leading up the game, second-best in the league.

The Colts still possess a below-average run defense with a 57 percent success rate, but their pass defense sits eighth in EPA.  Vontae Davis has finally translated his talent into consistent results, producing 18.2 EPA per game, 10th-most among all cornerbacks.  Free safety Mike Adams, signed off the scrap heap, has also helped stabilize the back end.  Indeed, in terms of EPA/P, the Colts pass defense ranks second only to Detroit's shutdown unit:

Indianapolis figures to remain at those lofty heights the rest of the season.  The Colts face five top-half passing attacks over their final nine games, but two of those are the Browns and Washington, neither of whom possess confidence-inspiring quarterbacks.  Playing in the putrid AFC South, the Colts will likely be the first team to clinch a playoff spot.  However, apart from a couple showcase games against New England and Dallas, there really won't be many opportunities to confirm this defense's legitimacy until the postseason.

- After emerging as the top-ranked team for the first month, the Cincinnati Bengals have since plummeted over the past three weeks.  Cincy represents the other side of the coin from Indy's thorough beating, as the Bengals nosedived from second to 12th place, easily this week's biggest drop.

Though an Andy Dalton-led offense will likely always exhibit plenty of variance, it's the defense that has earned deserved criticism for its sudden regression.  Long one of the league's most consistent and underrated units, the Bengals have allowed 6.0 yards per play over the last three games, eighth-worst in the league.  This comes after allowing 4.9 yards per play in the first three weeks, seventh-best in the league. 

It's not really hard to spot the culprit.  A toddler learning shapes and directions for the first time could probably spot something awry with this trend:

There are some encouraging long-term signs even in that ugly decline, believe it or not.  The Bengals per-play pass defense numbers are not as bad as that chart would imply.  They have conceded 6.94 yards per pass attempt during their three-game winless streak, which ranks a below-average but not heinous 19th.  Thus, Cincinnati still ranks sixth overall in pass defense efficiency on the season, a ranking that does not stem entirely from their strong three-game start.

The issue for the Bengals has been poor situational and big-play defense.  Despite that tenable per-play rate, opponents have picked up a first down on 40.2 percent of their pass attempts, fifth-highest in the league over that span.  Most distressingly, the Bengals have just two combined turnovers and sacks in the last three weeks, tied with the Seahawks for the lowest total in that stretch.

Small sample size plays a role in that, but so do a couple key injuries.  Top coverage linebackers Vontaze Burfict and Emmanuel Lamur have missed time, with Burfict's repeated concussion symptoms representing a scary issue that figures to threaten his season.  Burfict's long-term loss would represent a significant reduction in speed, a worrisome proposition given that the Bengals are mostly reliant on plodding technicians at all three levels of the defense.

Apart from Geno Atkins, who has appeared less explosive in his return from ACL surgery, the Bengals defense lacks a definitive playmaker.  Speed is increasingly important as offenses begin to incorporate more spread concepts in an effort to exploit the spaces available in any coverage scheme.  League-wide, passes have gotten shorter as offenses emphasize yards after the catch.

The Bengals pass defense is unlikely to suddenly plummet into Falcons territory, as there is still a meaningful amount of talent and depth here, even without Burfict.  Still, if this is really a league-average overall defense, then Cincy's fortunes will ride on the week-to-week whims of the Dalton-led offense.  For a team that harbored legitimate Super Bowl aspirations after September, that would likely keep the Bengals firmly beneath that first-round glass ceiling they cannot seem to break.

Pooch Punts

- Other than the Dolphins' ranking, perhaps the most curious takeaway from this week's ranking is seeing Washington with the third-ranked passing efficiency.  Considering that they are already on their third quarterback, and likely to take a new one at the top of the draft this spring, it is clear that something does not quite add up.

In actuality, the trio of Robert Griffin III, Kirk Cousins and Colt McCoy has posted some impressive per-play numbers.  Washington as a team ranks third in net yards per pass attempt, behind only the prolific Denver-San Diego duo.  However, Washington ranks out as a below-average pass offense in terms of EPA, despite those lofty yardage totals:

Of course, turnovers have doomed the position, as Washington quarterbacks are on pace for 27.4 turnovers this season.  For perspective, only Eli Manning and Joe Flacco exceeded that number last season.  These efficiency rankings give less weight to turnovers, since they are flukier and do not always hold significant predictive value.  However, given the histories of the Washington quarterbacks, turnovers represent the main reason why the organization will likely have its 17th different starting quarterback since 2000 next season.

- The Minnesota Vikings are the new owners of the ignominious 32nd-ranked overall offense.  Poor Teddy Bridgewater has been under constant duress, and as such, the Vikings have posted the second-least efficient passing game.  Without Adrian Peterson, Minnesota's league-average run success rate is no longer enough to keep the Vikings afloat.

Bridgewater has shown flashes at times, though most of those came at the expense of the 32nd-ranked Atlanta defense three weeks ago.  In seeing Bridgewater take a pummeling every week, it is difficult not to have flashbacks to David Carr, who was famously scarred by a league-record 76 sacks in 2002.

It's obviously far too soon to speculate about any long-term ramifications for Minnesota's first-rounder, but there is also legitimate reason for concern.  Bridgewater's 62 sack percentage index ranks second-worst among all quarterbacks with at least three starts this season, ahead of only Chad Henne.  If sustained over the entire season, that would represent the fourth-worst mark of since 2002

It's hard to ascribe causative to sack percentage index and eventual success.  Yes, rookies with the worst sack percentage indices generally flamed out, but that often stemmed from a glaring lack of pocket presence (see Gabbert, Blaine).  Joey Harrington and Brandon Weeden enjoyed significantly better protection than Ben Roethlisberger or Russell Wilson, but that never translated into superior careers.

Thus, it's too simplistic to suggest that Bridgewater will fail simply because he is likely to continue taking beatings behind Minnesota's shoddy offensive line.  But if Vikings cannot to improve to a merely below-average pass-blocking team, rather than a historically inept one, it might be best to preserve Minnesota's first-round investment at some point.

- Most fans are familiar with the devaluation of the running back, so unsurprisingly, run defense is also becoming increasingly overrated.  Of the top 10 teams in run defense success rate, only half have cracked the top-15 overall rankings.  Conversely, nine of the top-10 pass offense units and all of the top-10 pass defenses sit in this week's top 15.

Of those top-10 run defenses, only four also possess above-average pass defenses.  As mentioned earlier, only the Dolphins are top 10 in both, likely accounting for their unexpected place as the second-ranked team.

It's hardly surprising that run defense shows a lower correlation with winning.  Interestingly, though, offenses are on pace to run 2,449 rushing plays in the red zone this season, which would indicate an uptick back towards what the seasonal rates were in the mid-2000s:

There's no doubt that passing has far surpassed rushing as the more efficient form of offense, but if passing games are more reliant on space, as discussed earlier, then perhaps the constricted area of the red zone could represent the safe haven that preserves the value of the running game.  This isn't a new problem; Brian has done research suggesting that running is more efficient inside the 10-yard line, and that teams are not calling enough red-zone runs. 

That was four years ago, though, and passing schemes have taken some important leaps forward since then.  My intuition would be that Brian's numbers are still likely reflective of reality, but the research might be worth updating to see if passing progress has closed the gap on the "three yards and a cloud of dust" mantra.

Here are the updated team efficiency rankings after seven weeks.  As always, observations, questions and snide remarks are welcome in the comments section.

1 DEN 1 0.68 0.53 1 1
2 MIA 4 0.61 0.49 14 2
3 IND 15 0.61 0.52 4 7
4 GB 9 0.59 0.51 10 6
5 SEA 3 0.58 0.56 6 16
6 DAL 6 0.57 0.50 3 22
7 DET 7 0.55 0.48 21 3
8 KC 14 0.55 0.56 22 14
9 BAL 11 0.55 0.47 9 9
10 WAS 12 0.54 0.48 5 10
11 SF 5 0.54 0.52 18 5
12 CIN 2 0.54 0.49 8 17
13 SD 10 0.53 0.47 7 15
14 CLE 8 0.51 0.48 11 19
15 NO 13 0.51 0.45 2 28
16 NE 16 0.50 0.49 17 11
17 CHI 18 0.49 0.50 16 26
18 HOU 21 0.49 0.50 13 27
19 CAR 17 0.48 0.50 20 18
20 TEN 19 0.48 0.53 24 23
21 PHI 24 0.48 0.50 25 12
22 PIT 20 0.47 0.47 12 21
23 BUF 23 0.47 0.51 28 4
24 NYJ 25 0.46 0.53 27 13
25 NYG 22 0.46 0.49 23 25
26 ARI 26 0.45 0.52 26 24
27 JAC 31 0.43 0.52 31 20
28 STL 30 0.42 0.48 15 30
29 MIN 27 0.41 0.48 32 8
30 OAK 29 0.38 0.51 29 29
31 ATL 28 0.36 0.47 19 32
32 TB 32 0.33 0.46 30 31

ATL 6.8 36 2.8 1.4 8.2 57 1.8 0.50
ARI 6.2 33 0.5 1.6 7.4 69 3.6 0.39
BAL 7.1 40 2.0 1.2 6.3 65 1.5 0.35
BUF 5.9 36 2.3 2.1 6.2 67 3.1 0.57
CHI 6.3 46 2.7 2.4 6.9 56 3.5 0.44
CAR 6.3 39 1.1 1.7 6.7 60 2.8 0.47
CIN 7.1 39 1.5 1.0 5.9 46 2.7 0.40
CLE 6.9 41 1.0 1.5 6.4 52 3.3 0.45
DAL 7.5 44 2.8 2.5 6.8 59 3.1 0.36
DEN 7.9 39 1.4 1.4 5.3 65 2.0 0.52
DET 6.3 40 2.4 1.5 5.5 62 3.2 0.41
GB 6.8 42 0.9 1.5 5.8 49 4.1 0.34
HOU 7.4 37 3.6 2.8 6.6 50 2.2 0.42
IND 7.3 43 2.3 1.4 6.0 57 2.6 0.47
JAC 5.2 39 4.4 1.3 6.8 66 0.7 0.26
KC 6.0 43 2.1 1.5 6.1 54 1.6 0.33
MIA 5.7 53 2.3 2.2 5.2 63 1.8 0.32
MIN 5.0 41 4.8 0.8 6.1 59 2.6 0.46
NE 6.3 42 0.8 1.5 5.9 51 3.1 0.65
NO 7.0 50 2.7 1.1 7.2 63 1.4 0.38
NYG 6.3 40 2.2 2.3 7.5 69 4.8 0.36
NYJ 4.9 48 2.9 2.6 6.3 64 0.4 0.52
OAK 5.8 37 3.3 1.4 7.3 56 1.6 0.44
PHI 6.4 34 2.9 1.9 6.1 60 1.3 0.41
PIT 6.5 44 1.2 1.5 6.8 62 1.7 0.58
SD 7.9 32 1.3 1.5 5.9 51 1.7 0.45
SF 6.5 40 2.1 0.8 6.0 59 2.8 0.55
SEA 6.4 48 1.1 1.9 6.6 66 1.0 0.53
STL 6.4 45 2.3 2.6 7.6 61 1.7 0.59
TB 5.8 36 3.3 2.5 7.8 62 1.8 0.47
TEN 6.3 40 3.3 1.8 6.7 55 3.3 0.59
WAS 7.5 41 3.5 1.7 6.3 65 1.3 0.61
Avg 6.5 41 2.3 1.7 6.5 59 2.3 0.46