DVOA Analysis
Football Outsiders' revolutionary metrics that break down every single play of the NFL season

Week 2 DVOA Ratings

by Aaron Schatz

The Los Angeles Rams are on top of our DVOA ratings after two weeks, thanks to big wins over Oakland and Arizona. The Kansas City Chiefs and their dominating offense are right behind them; the Chiefs are currently No. 1 in both offense and special teams but dead last in defense. Those teams are then followed by a couple of unexpected 2-0 teams, Miami and Cincinnati, and Jacksonville is No. 5.

Now, you might be saying to yourself, "Wait, the Chiefs beat two good teams, the Steelers and the Chargers. The Rams beat two tomato cans, the Raiders and the Cardinals. How are the Rams No. 1? And how are the Miami Dolphins third when the teams they beat had an injured quarterback (the Titans) and a rookie quarterback (the Jets).

This is where we point out that our ratings will not include opponent adjustments until after Week 4. After all, can we say that when the Chiefs beat the Steelers, they beat a good team? We think the Steelers are still a good team, based on the information we had going into the season. That's the point of Football Outsiders' DAVE ratings, after all, which combine performance early in the season with our preseason projections to get a better idea of how good teams will be going forward. But we don't know for sure which teams are good and bad this season. We need more information. Once we've got four weeks of information, we can start filtering that knowledge into our ratings and gradually correct for opponent strength with those corrections getting stronger each week.

After all, there are some very strange results early in the season, or at least results that look strange. The strangest after Week 1 is probably the dominating early performance of the Miami Dolphins defense. The Dolphins are No. 1 on defense by a fairly big margin. The New York Jets are a surprising No. 2. The Bears have moved up to No. 4. It's hard to tell how "real" this is because we don't know how good the offenses they played really are. When the Jets stopped the Lions, did they stop an average offense or a bad one? The same goes for the Dolphins holding down the Titans, or the Bears holding down the Seahawks. Are these the average-to-good offenses of recent years, or the bad offenses they looked like playing these defenses that have been dominant early. Time will tell.

I did go into the archives to see what happens with defenses that play as well as Miami has played early this season. The Dolphins currently have defensive DVOA of -40.8%. Thirty-seven different defenses since 2002 have put up defensive DVOA of -30% or better after Weeks 1-2, without considering opponent adjustments. These teams ended the season with an average defensive DVOA of -9.5%, and an average defensive rank of 8.5. However, that's dragged down a bit by a few defenses where the first couple weeks were particularly a big mirage, including the 2012 Eagles and 2014 Redskins. The median defensive rank of these 37 teams was fifth at the end of the year. Only seven of the 37 teams ended the year with a defensive DVOA above 0% (i.e. worse than average).

If we want to look at the Dolphins and Jets together, I'll tell you that 32 teams since 2002 ranked first or second in defensive DVOA after Week 2, without considering opponent adjustments. The average rank of these teams at the end of the year was 7.5, and the median rank was fourth. So there's a good chance that the Miami Dolphins defense has, in fact, improved significantly from last year's finish of 28th in defensive DVOA.

My guess is that the general consensus is that Miami will be dragged down a bit once the opponent adjustments go in, and I would agree with that. Miami is a surprising 13th in the DAVE ratings right now. When you combine that with a 2-0 record and a schedule that we rated among the league's easiest before the season, you end up with Miami now ranking FIFTH in our playoff odds and a shocking SIXTH in Super Bowl odds. Yes, this is probably too high. Miami's true playoff odds are likely lower than we are showing them right now. But they are also likely higher than conventional wisdom thinks they are right now. And the Dolphins have a nice winnable game this week with Oakland coming to down. It's not ridiculous to think the Dolphins could start the season 3-0.

Miami's hot start is part of an overall hot start for football teams in Florida, with all three teams at 2-0. Yet only two of them are in the DVOA top five. Tampa Bay is way down at No. 16. Why so mediocre? Well, the answer is that the Bucs have been a very similar team to the Kansas City Chiefs, only not quite as efficient on offense despite Ryan Fitzpatrick's big numbers. You know how I mentioned that the Chiefs are in first place in offense and last place in defense? Well, Tampa Bay is in second place in offense and second-to-last place in defense. But while their defensive DVOA is pretty close to Kansas City's, their offensive DVOA is 20 points lower because Tampa Bay is dead last right in run offense. The Bucs are also much lower than Kansas City on special teams. Tampa Bay actually makes sense as an average team because their wins have both been very close, each one by a touchdown. The Bucs also have had excellent early fumble luck, recovering all four of the fumbles they've caused on defense. And there's an element of waiting for the opponent adjustments here, since we think the Eagles and Steelers are good opponents to beat. That's not reflected in the ratings yet, and we'll have to see where the Steelers go since their locker room seems to be currently melting down.

The other team where the record and the DVOA look very different is Houston. Despite starting the season 0-2, the Texans are 11th in DVOA. That's without opponent adjustments, so there's no consideration for the fact that their first loss came to the New England Patriots. Instead, what's going on here is that the Texans absolutely whooped the Tennessee Titans in DVOA this week despite actually losing the game. Houston's DVOA for this game is 39.2%, with the Titans at -44.4%. The Texans gained 7.0 net yards per play, and held the Titans to only 5.0. Their turnover margin of -1 doesn't come close to making up for that yardage difference. The Texans just had a number of drives with a lot of yards that didn't end in points. They had drives of 58 and 38 yards end in punts, and another 38-yard drive ended in an interception. They also got 31 yards on the meaningless last play from midfield but ran out of time to get up, spike the ball, and get the field goal team on. (That one play matters more in the yardage average than in DVOA.) The Texans were also more efficient than the Titans on third downs (5-fo-11 vs. 5-of-15). They lost the game because of one big play (the fake punt touchdown), two Tennessee fourth-down conversions, and some very bad game and clock management. They got stuck twice with fourth-and-long in no man's land, and punted both times: fourth-and-14 on the Tennessee 38 and fourth-and-8 on the Tennessee 37. But overall when it came to getting yardage, the Texans were better than you think they were. Those yardage gains aren't usually going to end up with fourth-and-long in no man's land in future games. Thus, the higher DVOA rating.

Finally, I wanted to point out something about the DAVE ratings that's fairly important. I've seen a couple of references to DAVE on the Internet, usually on Twitter, that say DAVE ratings are stupid because they combine 2018 regular-season performance with preseason performance. Nope, although I admit, that would be pretty stupid. We don't do anything to rate preseason performance. What DAVE is including is preseason projection, which includes how well the teams have played over the last couple years, coaching changes, personnel changes, turnover regression, and other factors. That's an important distinction.

* * * * *

The playoff odds are fully updated through Week 2. Snap counts and the FO premium database are updated as well. And we have offensive line and defensive front seven stats posted for the first time this season.

This week we're still experimenting with making the table below sortable plus we've added a "sticky header." Please let us know what you think.

* * * * *

These are the Football Outsiders team efficiency ratings through two 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.

Please note that there are no opponent adjustments in DVOA until after Week 4. (It's still listed as DVOA instead of VOA because I don't feel like going through and changing all the tables manually.) In addition, our second weekly table which includes schedule strength, variation, and Estimated Wins will appear beginning after Week 4.

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 80 percent of DAVE.

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>

RK TEAM TOTAL
DVOA
LAST
WEEK
TOTAL
DAVE
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 LAR 59.8% 3 22.7% 1 2-0 25.4% 4 -25.9% 3 8.4% 3
2 KC 49.7% 6 8.0% 7 2-0 59.4% 1 31.7% 32 22.0% 1
3 MIA 42.2% 9 2.7% 13 2-0 0.6% 18 -40.8% 1 0.8% 11
4 CIN 30.7% 8 3.9% 11 2-0 18.2% 5 -8.6% 12 3.9% 8
5 JAX 24.3% 11 0.4% 16 2-0 12.2% 9 -9.5% 11 2.5% 9
6 BAL 22.4% 2 7.7% 8 1-1 5.3% 16 -20.6% 5 -3.5% 18
7 CHI 19.7% 13 -5.0% 23 1-1 -7.3% 21 -21.3% 4 5.8% 5
8 MIN 18.7% 5 12.9% 4 1-0-1 15.9% 6 -16.7% 7 -13.9% 29
9 NYJ 18.0% 1 -5.6% 25 1-1 -23.1% 27 -26.8% 2 14.3% 2
10 LAC 17.5% 22 2.4% 14 1-1 36.7% 3 3.8% 17 -15.4% 30
11 HOU 15.6% 17 3.8% 12 0-2 0.9% 17 -10.4% 10 4.3% 7
12 PHI 15.4% 4 11.9% 6 1-1 7.4% 13 -15.2% 9 -7.2% 25
13 CAR 7.2% 10 -0.1% 17 1-1 13.3% 8 11.2% 24 5.2% 6
14 DEN 6.4% 7 -4.9% 22 2-0 7.0% 14 -0.5% 16 -1.1% 14
15 DAL 6.4% 24 7.3% 9 1-1 9.1% 11 -7.0% 13 -9.7% 27
16 TB 4.3% 21 -3.8% 21 2-0 39.3% 2 28.6% 31 -6.4% 22
17 WAS 3.5% 15 0.8% 15 1-1 -5.0% 20 -15.5% 8 -7.0% 24
18 NO 1.4% 18 12.9% 5 1-1 15.7% 7 9.9% 21 -4.4% 20
19 NE -6.5% 14 13.1% 3 1-1 6.0% 15 5.8% 19 -6.7% 23
20 OAK -8.6% 29 -5.4% 24 0-2 8.9% 12 15.0% 27 -2.5% 16
21 PIT -9.1% 12 13.2% 2 0-1-1 10.7% 10 11.6% 25 -8.3% 26
22 IND -9.8% 23 -3.3% 20 1-1 -10.6% 22 -2.4% 15 -1.6% 15
23 GB -12.0% 16 5.8% 10 1-0-1 -2.3% 19 6.9% 20 -2.7% 17
24 ATL -20.5% 27 -0.6% 18 1-1 -10.8% 23 10.7% 23 1.0% 10
25 SEA -23.5% 26 -2.3% 19 0-2 -27.8% 29 -4.5% 14 -0.2% 12
26 SF -28.5% 30 -9.6% 28 1-1 -20.1% 25 4.6% 18 -3.8% 19
27 NYG -30.7% 19 -17.8% 30 0-2 -19.4% 24 10.2% 22 -1.1% 13
28 TEN -32.0% 20 -6.6% 26 1-1 -23.8% 28 15.9% 28 7.8% 4
29 CLE -34.3% 25 -9.4% 27 0-1-1 -30.6% 30 -16.8% 6 -20.5% 31
30 DET -63.1% 31 -14.2% 29 0-2 -21.4% 26 20.4% 29 -21.3% 32
31 ARI -70.6% 28 -18.2% 31 0-2 -53.1% 31 12.3% 26 -5.2% 21
32 BUF -94.0% 32 -37.5% 32 0-2 -56.5% 32 27.4% 30 -10.1% 28

Comments

58 comments, Last at 20 Sep 2018, 5:45pm

1 Re: Week 2 DVOA Ratings

Minnesota is clearly ranked too low because their uniforms are so horrid on the outside, you can't see what great players they all are on the inside. Ranking teams by the number of NFC East refugees is way better than this. <333 Cousins, the most uncool starting QB in NFL history!

2 Re: Week 2 DVOA Ratings

I think there's reason to be a little pessimistic about the Chiefs simply because their success on offense seems less sustainable than their failure on defense. Mahomes is not going to be as good as he has been so far every week, especially once teams get some film on him. The defense will likely improve some, but maybe not as much as the offense will fall back to earth.

Funny that past the Rams, the next top 5 teams are from the AFC, which in the preseason was the consensus much weaker conference. Obviously way too early to tell if that will hold up though.

15 Re: Week 2 DVOA Ratings

On the other hand, in the modern rules environment, I believe it is easier for great passing offence + lousy defense to produce positive results, than vice versa.

The Chiefs aren't going to storm the league with that defense, but they ought to make the playoffs now. The it only takes a few kind bounces.

21 Re: Week 2 DVOA Ratings

There have been very few conference crossover games so far, and Detroit and NYG were involved in two of them.

Basically, the AFC may be inflated because they've played 40% of their delta games against historically-inept teams. Buffalo hasn't crossed over yet.

24 Re: Week 2 DVOA Ratings

I think if Berry returns healthy, it will have massive implications for their defensive performance. He is that good. So both offense and defense will likely regress towards the mean. I wouldn't be surprised to see them finish with the #1 offense in DVOA at the end of the season though. So many weapons.

3 Re: Week 2 DVOA Ratings

I liked the analysis with regards to -40% or better and 1st-or-2nd place defenses after two weeks.

Taking it one step further, what's the overall correlation between defensive DVOA after two weeks and season-end DVOA? What about correlation between rank after two and at the end of the season?

4 Re: Week 2 DVOA Ratings

I like the addition of the 'frozen header' to the grid. Kudos for keeping on tinkering and adding additional useful features.

8 Re: Week 2 DVOA Ratings

Eugene R McDuffie

The Chiefs ranking will fall . The defense is not very good at all, they will have to outscore the better teams in the league. The youngster has a very nice arm , very accurate, but if you look at his throws, he locks in on his target before he releases. That will catch up to him sooner than later.

The Rams look like a legit # 1 so far. I like the Cowboys , Jaguars and Dolphins defense more than the Rams right now. Patriots look suspect so far early on. Brady looks his age finally.

6 Re: Week 2 DVOA Ratings

The Vikings ST ranking at 29 with a -13.9% cannot be correct. In fact if the ranking was 32 that would be too high.

Zimmer, "did you see the game?" priceless.

The same can be said for Chiefs D and particularly their D line. Watching the euphemistically named pass rush is a sight to behold. They were dreadful last year and if it is possible they are worse this year.

25 Re: Week 2 DVOA Ratings

They have spent the majority of the season rushing only 3-4 players. They did go a series or two in the second half last week where they actually blitzed, and it worked. Then they stopped again and Ben just played checkers behind the line while he waited for someone to eventually get open.

38 Re: Week 2 DVOA Ratings

The Lions have earned their place at the bottom of the special teams rankings. With a new improved kicker Minnesota should(?) get better on ST, but the Lions general ineptitude is unlikely to improve.

7 Re: Week 2 DVOA Ratings

Somethings wrong with this because you have the Pats' D in the middle third. I expected bottom ten at best, and likely back in the bottom 3 after their matador display on Sunday.

10 Re: Week 2 DVOA Ratings

Yeah I was wondering if this was a historically worst two game start to season kind of bad they've been. Watching them is like watching a Alabama vs East Bumf*ck State or other rando Div II college no one has ever heard of, just a joke and free win for every other team on the schedule

12 Re: Week 2 DVOA Ratings

I did a quick search through previous Week 2 ratings, and it looks like the previous worst was -83.3 by the 2009 Lions. But I know DVOA gets tweaked from time to time, so I don't know if Detroit's rating from 2009 used the same formula as this year.

31 Re: Week 2 DVOA Ratings

Sorry, I should have done that table in the main article. Will do it next week if Buffalo loses again, but in the meantime:

2001 WAS -107.8%
1999 CLE -103.5%
2008 STL -103.3%
1989 PIT -100.5%
1997 SEA -95.5%
2018 BUF -94.0%
1996 TB -91.7%
2005 MIN -89.3%
2003 CHI -88.3%
2007 NO -86.7%

40 Re: Week 2 DVOA Ratings

That redskins team is one of the weirdest teams of all time imo. Those 2 weeks happened to be the end of the Jeff George experiment. Less than 300 yards combined in the first 2 games including 7 turnovers en route to losing 30-3 and 37-0. They then lost their next 3 games in a... slightly less embarrassing fashion I guess (a combined 77-29), resulting in a ridiculous 144-32 point differential through the first 5 games. After that however they somehow won 5 in a row, 8 of their last 11, and outscored their opponents 224-159. Despite the finish, Marty got fired for Spurrier, thus continuing the endless cycle of LOLREDSKINS.

41 Re: Week 2 DVOA Ratings

Right, that was the "LaVarr Arrington" year with Tony Banks as starting QB! Definitely a weird one. It was remarkable what Schotty managed to accomplish with that team

58 Re: Week 2 DVOA Ratings

I was also confused by the 2007 Saints' appearance on this list, given it was Brees/Payton, and they'd been to the NFC title game the previous season. Turns out they were eviscerated by Peyton Manning and the reigning champion Colts in the season opener, and were then blown out on the road again by a Tampa squad that won the NFC South that year - probably the last decent team of the Gruden era.

11 Re: Week 2 DVOA Ratings

With regard to the table design, I love the sortable columns, but I could do without the floating header. It's very distracting.

13 Re: Week 2 DVOA Ratings

Re:Sortable table feedback

1) It wasn't visually obvious that the tables are sortable.

Even though it's mentioned in the text and other posters mentioned it, when I looked last week they didn't seem sortable. I put it down to my laptop / Win10 / Chrome browser not having the capability. Then I happened to click in the header row and a little down arrow appeared and the column sorted. But I accidentally stumbled on it.

2) The header does float but being transparent it's confusing to read when previous rows are visible through it. Noting however that the two coloured DAVE rows do it perfectly.

3) The W-L column is narrow so the records are being displayed vertically "2-" on one row and "0" on the next. Looks particularly bad with teams now having ties and therefore needing three rows to display their win-loss record. Yet every other column in the table has ample space.

4) Quibbling - but I'm not sure the sortable on W-L actually ranks tied teams in the correct place. For example, Pittsburgh and Cleveland with their 1/2-win are ranked below Buffalo and Detroit with their 0 wins.

16 Re: Week 2 DVOA Ratings

Regarding 1: The up-down arrows have been present the whole time for me.
Regarding 2: Works on Chrome (Linux and Win10), Firefox (Linux and Win10), Edge (Win10), and IE (Win10) for me. What browser / OS are you using?
Regarding 3: Works for me with above browser/OS combos.

Inspecting the table in Chrome, I notice that the font used is Helvetica. Issues 1 and 3 seem like they could be caused by your browser using a different font that is wider than standard Helvetica. Helvetica is not available on Linux OR Windows 10 (by default) and is being replaced by Arial (on both and in all browsers). Perhaps your Helvetica is being replaced by a wider font?

Having Helvetica as the font is a CLEAR sign of a Mac user. Not specifying an alternative is just sloppy from Ars, since Apple systems are the only ones that come with Helvetica.

Issue 2 appears to be a bug in your browser. The background color for the table head is given as white and without an explicit opacity, the browser should render it completely opaque. Are we talking an old IE version? Or something completely different?

EDIT: I just realized you wrote that Win10+Chrome. That does make me wonder, since I've been unable to confirm that behavior with that combination. I've had issues previously with some web assets having an incorrect timestamp leading to the use of a cached version for e.g. a single css file that didn't have the expected contents, ultimately resulting in hard-to-reproduce incorrect page display / behavior. The prodigious use of forward and reverse proxies (e.g. CDNs and transparent ISP proxies) means that even flushing the local cache won't always fix the fault. As for the behavior you report, which font is getting subbed for Helvetica? Win10 doesn't ship with it and the default sub in your case is Arial, but you could have a Helvetica font installed that perhaps doesn't match the proportions of the original. I can't test it at the moment, but another possibility is that Windows is applying a font-scaling in some crappy way which is not uncommon for high-DPI displays. I haven't tested it with Win10, but it was a problem with earlier Windows versions.

28 Re: Week 2 DVOA Ratings

I gave it a try in Edge and everything came up good - it really became obvious that the table is sortable with arrows and opaque floating header etc. But won't be using Edge ongoing.

Not sure how I find which font is getting subbed for Helvetica ... it looks like Arial ... but fonts are not my specialist subject so take that with a pinch of salt. I had a look under Settings, Customise Fonts where Standard=TimesNewRoman, Serif=TimesNewRoman, SansSerif=Arial, FixedWidth=Consolas.

It's no big deal, I just wanted to provide feedback in case others are receiving the same issues.

44 Re: Week 2 DVOA Ratings

To avoid local caching issues in Chrome, you can try opening the page in "incognito" mode. This mode uses a separate cache that is cleared on closing all incognito tabs.

To see which font is actually rendered in Chrome select the text, right click, and select "inspect". Under the "Computed" tab in the right-hand side of the window there will be a "Rendered Fonts" part. This will list the font(s) rendered.

14 Re: Week 2 DVOA Ratings

The Chargers are 3rd in offensive DVOA by less than 3%. I wonder how much better their DVOA would be if their receivers had been able to catch in week 1.

46 Re: Week 2 DVOA Ratings

It was only three or four plays and if I'm correct in how DVOA works it wouldn't be that much. Additionally, some of the drops were on drives that ended in TDs regardless, so some good plays would have been thrown out.

47 Re: Week 2 DVOA Ratings

It was only three or four plays and if I'm correct in how DVOA works it wouldn't be that much. Additionally, some of the drops were on drives that ended in TDs regardless, so some good plays would have been thrown out.

17 Re: Week 2 DVOA Ratings

A couple of key plays in the Dolphins/Jets games immediately spring to mind. On the final play of the first half Enunwa was tackled at the one yard line after a long completion, as the clock expired. Then in the second half, Terrelle Pryor inexplicably quit on his route, which led to a well thrown ball by Darnold becoming an interception in the end zone.

Of course it is possible to identify such pivotal plays in most games, but just to emphasize that flipping the outcome of those plays, which were mostly down to bad play from the opposition or good luck, would significantly impact Miami's defensive DVOA. Unless a team is kerb-stomped into oblivion, like the Cardinals were by the Rams on Sunday, I'm trying not to draw many conclusions after two weeks.

18 Re: Week 2 DVOA Ratings

This is correct, although it was Chris Herndon getting tackled at the 1 not Enunwa. Herndon also failed to get OOB a couple playes earlier that cost them a TO and dropped a 20 yard pass on 3rd down that hit him in the numbers later in the game. Not a good game for the rook. Anyway, there are always plays that could have gone either way but in this game it seems like there were more than usual and more pivotal ones. It was a weird game.

33 Re: Week 2 DVOA Ratings

DVOA may have frowned greatly on that play even though it resulted in no points. I remember a 2011 game between the Jets and Patriots, where the Jets were dominated on the scoreboard 30-21 but came out way on top in DVOA, partially due to an Antonio Cromartie interception returned from the Jets goal line to deep in Patriots territory as time expired. Not sure DVOA cares if that play ended the half or not.
Honestly, I think the Jets' and Dolphins' defenses are for real, even if the teams themselves are not.

19 Re: Week 2 DVOA Ratings

Your Ryder Cup ad keeps overwriting the background and making the comments section completely unreadable.

22 Re: Week 2 DVOA Ratings

Cowboys fan here, I'm willing to buy the defensive ranking at 13 even though I'm cautiously optimistic it will actually improve. We're also among the crap kicker teams after jettisoning Bailey to save dollars, and the special teams are struggling.

But there is absolutely no way that offense belongs in the top two thirds of the league, much less the 11 spot. While Prescott has avoided turning the ball over, his solid looking completion percentage of 64.8 is a mirage after he's thrown for a combined 330 yards across both games with a ANY/A of 5.3. For comparison's sake, Andy Dalton's career rate is 6.2. Dallas just looks inept on offense, I definitely expect that ranking to plummet once adjustments come online.

23 Re: Week 2 DVOA Ratings

Dolphins are ranked too high because they're the Dolphins. But I have to admit it looks like they might be better than most of us thought.

26 Re: Week 2 DVOA Ratings

I do find it interesting that so many AFC teams are ranked so highly.. when the prevailing wisdom is/was that the NFC had all the great teams, and the AFC was the weaker Conference.

27 Re: Week 2 DVOA Ratings

Worst 4 teams in the league are the Browns, Lions, Cards, and Bills.

That feels right to me from a historical point of view.