Week 6 DVOA Ratings
by Aaron Schatz
In a statistical metric based on play-by-play analysis rather than wins and losses, sometimes a team will rate better for a close loss than another team rates for a close win. That's what happened this week in the Football Outsiders DVOA ratings, and it was enough to knock the undefeated Los Angeles Rams out of the top spot. Instead, the Kansas City Chiefs ascend to No. 1 despite losing their first game to New England on Sunday night. (Perhaps it is more accurate to say that Kansas City doesn't ascend but instead stands in place while the Rams and Chicago Bears drop.)
After Week 6, neither the Chiefs nor the Rams now rank among the dozen best offenses we've ever tracked through this point of the season. The Rams really saw their passing game struggle against Denver this week. The Rams had only 2.6% offensive DVOA for the game after adjusting for Denver's defense, and the split there was -43.1% for passes and 42.8% for runs. It really was a huge ground game for Todd Gurley, even considering that the Broncos are one of the league's worst run defenses (29th in DVOA).
For Kansas City, the interesting split this week was first half vs. second half. In the first half, Kansas City had -32.3% offensive DVOA. In the second half, the Chiefs had 68.1% offensive DVOA.
There are two big differences between the Chiefs and the Rams this year. The first is schedule strength. The Rams are still higher than the Chiefs in non-adjusted VOA, although both teams saw this rating drop after Week 6. However, the Rams saw their DVOA drop as our opponent adjustments got stronger, while Kansas City's DVOA actually went up slightly. Kansas City's past schedule now ranks seventh by average DVOA of opponent, while the Rams' schedule ranks 22nd.
The second difference is special teams. I'm usually reticent to draw attention to special teams early in the season because one or two plays can have such a huge impact on special teams DVOA in just a game or two. But now that we're talking six games, I think it's a bit more safe to talk about how historically good the Kansas City special teams have been this year. Kansas City has had positive special teams value in five of six games, and the other game was -0.6%. Kansas City has more games with special teams DVOA over 25% than the entire rest of the league combined: Week 1 against the Chargers and Week 6 against the Patriots, compared to just the New York Jets in Week 1 against Detroit.
This isn't just about Tyreek Hill. Head over to our special teams ratings table (now sortable!) and you'll see that Kansas City currently ranks first or second in all five aspects of special teams that we measure. Rookie Tremon Smith had the 97-yard kickoff return on Sunday night, Dustin Colquitt leads the league in net punting average, and Harrison Butker hasn't missed a field goal or extra point yet this season. As a result, the Chiefs are on pace to rank among the best special teams we've ever tracked:
|BEST SPECIAL TEAMS DVOA
THROUGH 6 GAMES, 1986-2018
Of course, a caveat goes here: special teams performance is nowhere near as consistent as offense or defense, so it's harder to predict what we'll get from the Chiefs special teams for the next two and a half months. For example, that 2003 Chiefs team with Dante Hall, the team with the best special teams DVOA ever through six games? They do not appear on a similar list of the best special teams over an entire season, because the 2003 Chiefs had below-average special teams in the second half of the year. But a few of the units listed above do finish on the list of the best special teams DVOA ever through an entire season, including the 1986 Saints, 1996 Panthers, 2006 Bears, and 2009 Browns.
Only three teams have ever finished the regular season with special teams DVOA above 10%: the 2002 Saints, 2007 Bears, and 1994 Browns. If the Chiefs can keep this up, they'll become the fourth.
Below the Chiefs and Rams, the Chicago Bears drop to No. 3 after losing to Miami this week. But the big gap we had last week between the top three and the rest of the league is now gone. Baltimore moved up 8.8% DVOA after shutting out Tennessee, and the Los Angeles Chargers moved up 6.6% DVOA after beating Cleveland. A third team was also a big riser this week after blowing out an inferior team: the Seahawks moved up from No. 14 to No. 8 after their big win over the Raiders in London.
Seattle is now up to No. 4 in defensive DVOA. Chicago is still No. 1, and Baltimore is No. 2. So who is No. 3? You'll have to scroll all the way to the bottom of the overall DVOA table for that one. We have to talk about the insane split right now between the Buffalo Bills offense and the Buffalo Bills defense.
The Buffalo Bills defense right now is third in DVOA, narrowly behind Baltimore. The Bills are even better against the pass, second behind Chicago. The offense, on the other hand, is the worst in the league. But this is not a run of the mill worst of the league. We've had plenty of teams before that were run-of-the-mill bad on offense and very good on defense. The Bills are in another galaxy compared to the rest of the offense-fortified NFL of 2018. Right now, the average team is gaining 5.7 yards per play. Arizona is 31st at 4.4 yards per play. Buffalo is way down at 3.7 yards per play. It's even worse if we look at net yards per pass attempt. The NFL average is 6.6, and Arizona is 31st at 5.1. Buffalo is way down with 3.8 net yards per pass attempt. The Bills lead the league with nine interceptions and are dead last with only three passing touchdowns. (Somehow they are not in last place in total offensive touchdowns; Buffalo has seven and Tennessee has only six.)
Put it all together and normalize for the overall offensive environment of the 2018 season, and Buffalo now has the second-worst offense we've ever tracked through six games.
|WORST OFFENSIVE DVOA
THROUGH 6 GAMES, 1986-2018
Now, we could compare Buffalo to other teams that have had a high ranking in offensive DVOA and a really low ranking in defensive DVOA. For example, last year's Arizona Cardinals were 30th on offense and fourth on defense. However, what really matters here is the rating, not the ranking. The Bills are so, so far behind everyone else in offensive DVOA that it's not enough just to say that they rank dead last.
So I've put together this historical table of similar teams through six games. These are teams that had a bottom-five offensive DVOA but a strong defensive DVOA, with the biggest gap between the two, looking at the actual rating rather than just the ranking. You'll notice that the only team with a bigger gap was the team that's ahead of Buffalo (or should I say behind?) for worst offense through six games: the 2004 Miami Dolphins. That team had just four offensive touchdowns through six games. But they also only allowed seven offensive touchdowns through six games. The 2004 Dolphins had four different turnovers returned for touchdowns just in their first six games of the season! So the 2004 Dolphins were like this year's Bills only even more extreme. They were No. 1 in defense through six games as well as dead last in offense.
Most of these teams were terrible. What on earth are the 2015 Broncos doing on here? That team went to the Super Bowl! However, they weren't as good in their early games as you would expect a 6-0 team to be. Denver won five of those six games by just one score. At least their offense significantly improved over the course of the season, while their defense stayed No. 1. They had a lot more success than the rest of the teams on this list.
Is there hope for Buffalo? If Josh Allen is injured and they have to play Nathan Peterman, it would seem not. But maybe veteran Derek Anderson can rescue the Bills from historically awful offensive efficiency. The 2004 Miami Dolphins, at least, didn't suffer with an offense this bad for the whole year. They beat the St. Louis Rams 31-14 in their very next game. Eventually, as you can see in the table, their offense regressed towards the mean a bit. Unfortunately, so did their defense. Most of the teams on this table were bad on offense all year long, even if things got a little better than they were in the first six weeks. The 2015 Broncos are really the exception.
* * * * *
Stats pages should now be updated through Week 6, including playoff odds, the FO Premium DVOA database and snap counts. We're still working on transitioning all our tables over to the new sortable tables but you'll notice this week that spspecial teams are now sortable, which should be super helpful because you can sort by any of the five aspects of special teams that we track. The defense vs. types of receivers table is also now sortable although we need to tweak that one a little more to make it look right.
* * * * *
These are the Football Outsiders team efficiency ratings through six weeks of 2018, measured by our proprietary Defense-adjusted Value Over Average (DVOA) system that breaks down every single play and compares a team's performance to the league average based on situation in order to determine value over average. (Explained further here.)
OFFENSE and DEFENSE DVOA are adjusted to consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.
Because it is early in the season, opponent adjustments are only at 60 percent strength; they will increase 10 percent every week through Week 10. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.
DAVE is a formula which combines our preseason projection with current DVOA to get a more accurate forecast of how a team will play the rest of the season. Right now, the preseason projection makes up 30 percent of DAVE for most teams (40 percent for teams with just five games played).
To save people some time, please use the following format for all complaints:
<team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>
- NON-ADJUSTED TOTAL DVOA does not include the adjustments for opponent strength or the adjustments for weather and altitude in special teams, and only penalizes offenses for lost fumbles rather than all fumbles.
- ESTIMATED WINS uses a statistic known as "Forest Index" that emphasizes consistency as well as DVOA in the most important specific situations: red zone defense, first quarter offense, and performance in the second half when the score is close. It then projects a number of wins adjusted to a league-average schedule and a league-average rate of recovering fumbles. Teams that have had their bye week are projected as if they had played one game per week.
- PAST SCHEDULE lists average DVOA of opponents played this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
- FUTURE SCHEDULE lists average DVOA of opponents still left to play this season, ranked from hardest schedule (#1, most positive) to easiest schedule (#32, most negative). It is not adjusted for which games are home or road.
- VARIANCE measures the statistical variance of the team's weekly DVOA performance. Teams are ranked from most consistent (#1, lowest variance) to least consistent (#32, highest variance).
58 comments, Last at 23 Oct 2018, 7:21pm
#48 by BJR // Oct 18, 2018 - 6:52am
What happened to that team after 1991? Seems unusual for such an historically great team to fade away so quickly, especially pre-free agency. I'm aware Gibbs retired soon after, and Dallas were emerging as a new powerhouse, but there must be other factors?
#50 by Will Allen // Oct 18, 2018 - 8:32am
Jack Kent Cooke was an owner similar to DeBartolo, in his willingness to spend a lot of money, for that era, meaning pre salary cap/ free agency. An owner like that, joined to superior coaching and a good personnel guy (Bobby Beathard) , really gave a team an edge. In particular, a roster could have depth which simply is no longer obtainable in the NFL of today. That's how you have a team which is truly dominant on offense, defense, and special teams. We will never see a team like that again, in all likelihood. The 2007 Patriots came closest since the salary cap fully emerged, which I date to the mid 90s, but that 2007 Patriots team really wasn't dominant on defense, and was kind of a one year anomaly, with prime Moss being on a well below value contract.
Once the salary cap and free agency fully kicked in, after Beathard and Gibbs left, Jack Kent Cooke was just another rich guy who had no clue as to how to build a roster. Then Cooke died. Then Daniel Snyder happened.
#21 by ammek // Oct 17, 2018 - 11:16am
The NFC South offenses are all among the top 12, its defenses all in the bottom 12. I wonder whether that's because they've played each other. FO projections (and indeed everyone else) had the Panthers topsy-turvy. What has happened to their D?
In fact there are quite a few division bunches. The AFC South has bottom-quartile offenses and top-half defenses. (Yep, even the Colts.) The NFC East is just a pile of mediocrity, its only top-ten unit is Washington's legendary special teams. The NFC North is where special teams go to die.
There are seven NFC teams with playoff odds between 37% and 48%, for three berths (the NFC East title and the wildcards). I've no idea who will prevail. The teams all look good enough on paper. There's a very strong home/road split among them: the teams are a combined 16-4-1 at home, but only 5-13-1 on the road (one of the road wins being a London game). If it comes down to head-to-head tiebreakers, it may be significant that Dallas played (and lost to) Carolina and Seattle on the road, while Washington played (and defeated) Carolina and Green Bay at home. Seattle has home games to come against both Green Bay and Minnesota.
#18 by Will Allen // Oct 17, 2018 - 8:34am
Has FO ever done a retrospective SOS ranking? I wonder how frequently what looks like a difficult, or easy schedule, just before week 1, ends up not being the case, by a significant amount. I even suspect that might happen at week 8 sometimes.
#19 by cstoos // Oct 17, 2018 - 9:34am
SOS above is based on DVOA ratings, so it changes weekly with the results. So it is reactionary and not predictive. A good example would be KC's win @ Denver. Broncos looked tough then. Not as tough now. So it likely reduces KC's DVOA from that game according to the future results.
DAVE still includes some pre-season predictions though, so the difference between that and DVOA is sort of like a retrospective SOS ranking. Teams they thought would be good versus teams that are actually good.
#20 by ammek // Oct 17, 2018 - 10:48am
Yes I'd be interested to see that too. I think the schedule adjustments in the preseason projections are pretty worthless. Maybe it's the example you had in mind: Oakland's schedule was projected to be the 9th easiest, and now it's looking like it might be one of the hardest in DVOA history. The Panthers were supposed to have the toughest schedule, but now look like they'll have one of the ten easiest.
On the other hand it's always worth adjusting the AFC South projections for that division's guaranteed mediocrity.
#14 by TomC // Oct 16, 2018 - 11:43pm
I'm usually reticent to draw attention to special teams early in the season...
Except when it's Dave Toub, because Dave Toub is the GOAT.
But a few of the units listed above do finish on the list of the best special teams DVOA ever through an entire season, including the 1986 Saints, 1996 Panthers, 2006 Bears, and 2009 Browns.
Only three teams have ever finished the regular season with special teams DVOA above 10%: the 2002 Saints, 2007 Bears, and 1994 Browns. If the Chiefs can keep this up, they'll become the fourth.
Hmm, 2006 and 2007 Bears. I wonder who coached those ST units...
#57 by Eddo // Oct 19, 2018 - 1:56pm
Hester was awesome, but every facet of the Bears special teams was strong under Toub.
By FO points added rankings:
Kickoffs: +17.2 (led league)
Kick returns: +3.8 (Hester returned only 20 of their 59 kick returns)
Punt returns: +11.7 (Hester returned 47 of 50)
Kick returns: +13.8 (Hester returned 43 of 67)
Punt returns: +18.0 (led league; Hester returned 42 of 44)
#12 by thatoneguy123123 // Oct 16, 2018 - 10:27pm
New Englands offense ranking went down from 5th to 7th after putting up 43 against the Chiefs. They had zero penalties accepted against them, zero punts, and 2 turnovers (1 fumble, 1 TO on downs). Yes, it was against a very poor defense but that was the possibly the best offense performance of the season IMO.
Other teams have had bigger plays, like chiefs did that game. But Patriots performance was nearly perfect.
I understand that pre season rankings are being adjusted each, teams schedule are factors in, etc.
#15 by Bright Blue Shorts // Oct 17, 2018 - 5:34am
New England's DVOA went UP from 11% to 13%. DVOA recognises that they performed better.
Unfortunately for the Patriots ... Green Bay and Tampa Bay's DVOA also went up from 9% to 13+ % to put them higher in the rankings.
#17 by Aaron Brooks G… // Oct 17, 2018 - 8:27am
Yes, it was against a very poor defense but that was the possibly the best offense performance of the season IMO.
How quickly we forget Tampa-New Orleans in week 1. 48-40, also featuring a matchup of two putrid defenses.
#40 by thatoneguy123123 // Oct 17, 2018 - 5:47pm
That game was different. Tampas offense was big plays and minimal running. Saints had 8 penalties accepted against them, Bucs had 7. Peyton Barber had 19 carries for a putrid 69 yards and saint's completely abandoned the run. Bucs had a fumble recovery for a TD and two 50+ yard TD passes. Patriots longest TD was 17 yards. Everything other TD was inside the 5 yard line.
You could replay the Patriots offense performance against Chiefs and they would very similar results. There wasn't 1 or 2 plays that made a huge difference on the Pat's side. Replay that Bucs Saints game and you will get a range of different results.
#8 by dank067 // Oct 16, 2018 - 8:46pm
Appears that DAVE was accidentally assigned in the table here based on the DVOA rankings of each team - looks like you can find the correct DAVE values if you click on 'team efficiency' under the statistics heading at the top of the screen
#13 by TomC // Oct 16, 2018 - 11:40pm
I love the fact that even after ~15 years, this is how the FO staff responds to mistakes being pointed out. (I.e., swiftly, personally, and with "I goofed" instead of some corporate BS or defensiveness.) One of the many little things that make this a great place to virtually hang out.
#5 by Mountain Time … // Oct 16, 2018 - 7:51pm
I've never disagreed with DVOA as strenuously as I do with its rating of Denver. The 10th best offense in the league (bwahahahahaha! Yeah right) splits out as 18th passing/2nd rushing. The splits seem about right generally, but add it all up and I feel they're a lot closer to 18 than to 2.
The defense splits are similar: 8th vs pass, 29th vs run. At least here I can maybe understand how DVOA is underrating them. They've gotten burned on a handful of huge pass plays that were right in the edge of being huge plays going the other way, and obviously DVOA can't know that. This is easily a top-10 unit overall IMO, and I'm shocked the Rams game didn't raise their rating more.
DVOA also seems to be underrating the Vikings by a large margin. Not even the Bills debacle should account for that.
#6 by Otis Taylor89 // Oct 16, 2018 - 8:21pm
DEN has played the toughest schedule after 6 games by a solid margin over the #2 hardest and by large margins over the #3 and #4 toughest schedule. If they played an average strength schedule they probably would have at least two more wins and a good shot at getting into the playoffs.
#23 by deus01 // Oct 17, 2018 - 12:03pm
Is there a variance listed for the OFF/DEF/ST splits somewhere? Denver's offense can be strange in that sometimes they seem to be able to move the ball really well and other times they're held back by Keenum airmailing passes and holding calls. I wouldn't be surprised if the offensive DVOA variance is pretty high and that's contributing to some extent to their relatively high rank.
#30 by Eddo // Oct 17, 2018 - 2:02pm
There is a "VAR" column on the offensive and defensive DVOA tables (go to "Statistics" in the site navigation you can get there).
That said, what you're describing might not show up in offensive variance; I believe that variance is simply the standard deviation of game-by-game DVOA, so if a team is inconsistent within a game, that doesn't come through.
#34 by deus01 // Oct 17, 2018 - 2:52pm
Thanks. Denver actually has one of the lower Offensive variances.
So I guess it's just be a situation of Denver being relatively good at the types of plays that DVOA 'likes'. If Denver misses on their big plays more often than an average team but can make the smaller plays that are considered successes more often then I think our perception of the Denver offense would be a lot worse than its DVOA.
#51 by Eddo // Oct 18, 2018 - 8:56am
Not necessarily. As I said before, the variance measures here are not play by play, but rather game by game. So if Denver had 10 amazing plays in each game mixed with 30 horrendous ones (their in-game variance is high), but they did that every game, their VAR figure would be low.
#42 by reddwarf // Oct 17, 2018 - 9:04pm
I'm as big a Denver fan as it's possible to get, but top 10 defenses do not give up historically bad rushing numbers like they have the last two weeks. I don't know exactly why its collapsed so badly compared to last year, but it has. If anything, 18 feels overly generous at the moment (to me).
I still think the biggest mistake Elway's made isn't related to drafting bad quarterbacks, or even hiring Joseph (hindisght is 20/20 when it comes to who the best new head coach would be). But letting Wade Phillips walk was just so incredibly and obviously stupid.
#43 by theslothook // Oct 17, 2018 - 10:49pm
I've always been reticent to criticize Elway..Manning revitalized the franchise, but the title was a testament to Elway.
But three moves he did strike me as poor even without the benefit of hindsight. Hiring Vance Joseph and not keeping Wade, you correctly mentioned, but the chubb pick was also wrong. Qb remains a need and he had to know that Case Keenum wasn't a long term solution.
#54 by Mountain Time … // Oct 18, 2018 - 4:34pm
I actually give Elway a pass on Chubb, picking him was fine. I agree that Rosen was the better pick to make, but Chubb isn't so far behind, and he may well be the better player in six years (I expect them both to be good). Now picking Shane Ray, Osweiler, P.Lynch, those were very obviously bad picks even at the time. The Chubb pick was more "possibly non-optimal" than "bad"
#56 by theslothook // Oct 18, 2018 - 11:47pm
I agree and in a way, Denver is really nicely setup to be an elite team. If they had a non awful coach and a QB better than Keenum, I think they would be a serious contender. Unfortunately...Vance Joseph.
I was also very sad to see the Broncos take Bradley Chubb one spot ahead. So very sad.
#2 by theslothook // Oct 16, 2018 - 7:32pm
Last week I mentioned how staggeringly awful the Bills pass offense was. Remarkably(though maybe not so given Peterman was once again on the field) they dropped even further. I haven't watched the bills at all this season, but boy o boy does Tyrod Taylor look like Fran Tarkenton by comparison(yes, even including his sad stint as the Browns week 1 starter).
#7 by JoeyHarringtonsPiano // Oct 16, 2018 - 8:42pm
A pass offense being so bad in today’s era defies logic. Peterman and Allen’s stat lines look like they’re out of the 1940’s. That makes it even impressive that their defense helped them be competitive in 3 out of their 5 games. They remind me of the 1998 Chargers.
#25 by Richie // Oct 17, 2018 - 12:38pm
I can't remember the details, but Testaverde may have done it twice.
In 2005, Testaverde started in week 5 for the Jets, and then in 2007 he started in week 6 for the Panthers. I am pretty sure he signed off the street with the Panthers. And he may have done it for that Jets team as well.
#26 by Bright Blue Shorts // Oct 17, 2018 - 12:39pm
I believe the 1987 replacement players would have been street signings for their week 4 debuts but that's a technicality / aberration rather than genuinely addressing the point being made.
In the old days teams carried three QBs as a matter of course. So highly unlikely to be starting a street signing before week 7.
Injuries would be the most likely reason. Have seen some teams like Cleveland in 88, Raiders in 2003 go through 5 or 6 QBs in a year for that reason.
(The Pats might have started a street signing in week 4 of 2016 if their game hadn't been on Thursday night. Brady suspended, Grop injured and then Brissett injured. But with no time for an incoming guy to learn the system they went with Brissett).
#29 by Travis // Oct 17, 2018 - 1:57pm
(Week 4 for the 2016 Patriots was a regular Sunday game; Brissett had gotten hurt in the TNF game the week before.)
Should Josh Freeman getting thrown to the wolves in Week 7 for the Vikings after being cut after Week 4 by Tampa in 2013 count?
#32 by JoeyHarringtonsPiano // Oct 17, 2018 - 2:39pm
The 1991 Eagles maybe? Randall Cunningham was lost for the season in week 1. Don't remember when the primary backup Jim McMahon got hurt (because of course he did), and don't remember who the original 3rd stringer was (Jeff Kemp, maybe?). But the Eagles eventually started guys named Brad Goebel and Pat Ryan, who must have been street free agents.
#53 by nosoop4u // Oct 18, 2018 - 1:32pm
Josh Freeman started week 7 for the 2013 Vikings a week after being signed. It went poorly (20/53, 190 yds, 0 TD, 1 INT, 2.7 AY/A). He only played one more NFL game for the rest of his career, 14 months later.
Hope for his sake Derek Anderson does better than that.
#36 by jtr // Oct 17, 2018 - 3:27pm
>A pass offense being so bad in today’s era defies logic. Peterman and Allen’s stat lines look like they’re out of the 1940’s.
Peterman's career ANY/A is actually slightly negative, suggesting that it would be about .3 yards per play more efficient for him to just spike the ball every snap than it is for him to whatever the hell it is that he's been doing.
#45 by Lost Ti-Cats Fan // Oct 17, 2018 - 11:13pm
"A pass offense being so bad in today’s era defies logic."
McDermott on nfl.com is quoted as saying re Anderson: "he's worked hard over the last week and a half here to get up to speed"
At first I assumed that was coach-speak, but then I look at those DVOA numbers and wonder if maybe Buffalo's offence is so simple a QB can pick it up in 10 days?