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

Week 1 DVOA Ratings

by Aaron Schatz

You love them when your team is high! You hate them when your team is low! Once again, the famous Football Outsiders DVOA and DAVE ratings return for 2019.

Football Outsiders readers are familiar with DVOA, which we use all year round. By now, most Football Outsiders readers are also familiar with DAVE, which we only use for teams during the first two months of the season. DAVE is our rating that combines the preseason projection with the results of early games to give us a better prediction of how each team will rank at the end of the year. For those who don't know the story, this metric is called DAVE as a reaction to criticism that our stats are too much alphabet soup. I mean, who can argue with a guy named Dave? (Technically, it stands for "DVOA Adjusted for Variation Early.") In this week's DAVE ratings, the preseason projection counts for 90 percent, and the current DVOA counts for 10 percent. The value of the preseason projection changes each week until we are only using current-year data after Week 9.

Of course, though I'm calling our main metric DVOA here, it is actually VOA because there are no opponent adjustments right now. We do not apply opponent adjustments until after Week 4, so in Weeks 1-3 DVOA and VOA are the same thing. Please don't get all nitpicky about it. Most readers know what's up, and if you don't, I just told you!

It will surprise nobody that the Baltimore Ravens are on top of our ratings after just one week. They wrecked the Dolphins so bad it made Aquaman cry. Baltimore is in the top three for offense, defense, and special teams after one week, and they end up with highest Week 1 VOA rating in 20 years.

Best Week 1 VOA Ratings, 1986-2019
Year Team VOA Opponent Score
1999 PIT 177.6% CLE 43-0
1989 CLE1 154.7% PIT 51-0
1991 WAS 148.7% DET 45-0
1990 CHI 123.9% SEA 17-0
2019 BAL 119.3% MIA 59-10
1998 SEA 114.8% PHI 38-0
2005 PIT 106.8% TEN 34-7
1999 JAX 106.6% SF 41-3
1996 GB 106.4% TB 34-3
1997 DAL 104.9% PIT 37-7

That 17-0 game stands out among these top Week 1 games with a surprisingly high VOA rating. It's mostly defense, as the Bears held Seattle to 3.0 yards per play with three turnovers.

What's interesting is that the Miami Dolphins did not end up on a similar list of the worst Week 1 games in DVOA history. (VOA for one team is not just the opposite of the rating for their opponent, because certain adjustments are slightly different for offense and defense while plays like false starts and aborted snaps only count on one side of the ball.) You may remember that we ran that list last year, because two Week 1 performances in 2018 rated as worse than Miami's terrible Week 1 this year. Miami's VOA of -94.7% doesn't end up anywhere near as bad as last year's Buffalo Bills (-130.6% for a 47-3 loss to Baltimore) or Detroit Lions (-123.4% for a 48-17 loss to the Jets).

I haven't heard anyone even mention how badly the Ravens beat up on Buffalo in Week 1 last year. That's two straight seasons they've done this!

Miami may not come out as bad as last year's Bills did, but they had a preseason projection worse than last year's Bills so their DAVE rating is even worse. Last year, Buffalo had a DAVE of -34.0% after one week and came out 0-16 in 1.3% of our playoff odds simulations. This year, Miami has a DAVE of -37.0% after one week and comes out 0-16 in 1.7% of our playoff odds simulations. The difference between Miami and the No. 31 team, the New York Giants, is larger than the difference between the Giants and the No. 14 team, the Minnesota Vikings.

New England is No. 2 in VOA after Week 1, making them easily No. 1 in DAVE. However, the Patriots are not as strong as the Dolphins are weak. They go 16-0 in only 0.6% of our simulations. The team they beat, the Steelers, finish 31st in VOA but they're still 12th in DAVE because of their strong preseason projection.

Tennessee finishes No. 3 in VOA, and that plus losses by the rest of their division is enough to make the Titans new favorites in the AFC South. The Titans' VOA is much more positive than Cleveland's VOA was negative for the same game. Dallas and -- surprise! -- Oakland finish out the top five.

Only one game in Week 1 came out with the loser scoring a higher VOA rating than the winner: Washington (eighth) against Philadelphia (15th). This is an interesting one because the Eagles had the much higher success rate (50 percent to 40 percent) but Washington had more average yardage (6.98 compared to 6.24 for the Eagles). Some of the difference comes because the Eagles averaged just 8.2 yards to go on their plays while the Redskins averaged 10.3 yards to go. And part of the issue there was penalties, where Washington had 12 and Philadelphia only six. Washington had a lot of holding calls: five, including one on a punt. That's something that needs to go into the DVOA system next time I get to do a big overhaul on it.

New Orleans and Houston come out just 17th and 18th after that amazing Monday night game, but those ratings will almost assuredly improve once we start putting in opponent adjustments.

Despite choking a win away and settling for a tie, Detroit ends up with a higher rating than Arizona for their Week 1 contest. Detroit had an average gain of 6.0 yards compared to just 4.7 yards on average for the Cardinals. Turnovers were also an issue here: Arizona had an interception but no fumbles, while Detroit had no interceptions but fumbled three times and lost two of them. With the Lions putting up a slightly above-average VOA rating in Week 1 to go with their slightly above-average preseason projection, we still have the NFC North as a very tight playoff race. But the Packers and Vikings moved ahead of the Lions as division favorites thanks to winning their games and having higher Week 1 ratings than Detroit.

* * * * *

Football Outsiders playoff odds are updated through Week 1. The annual stats pages are now updated with 2019 data, although some of that data can be kind of sketchy after just one week (in particular the offensive line and defensive line pages). We're finishing up the update on the snap counts database and the premium DVOA database right now and that should be up by the end of this evening.

The new 2019 season also means a lot of new content on Football Outsiders, especially fantasy football content with our new senior analyst Scott Spratt. You'll want to check out the brand new FO Fantasy Podcast twice each week (click here for Tuesday's edition) and look for it now on the iTunes podcast app. There are also new articles on Tuesday (Win the Wire, about waiver pickups) and Thursday (Start and Sit).

* * * * *

These are the Football Outsiders team efficiency ratings through one week of 2019, 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 90 percent of DAVE. The DAVE rating for Jacksonville incorporates a lower projection to account for backup Gardner Minshew playing for at least the next two months.

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
TOTAL
DAVE
RANK W-L OFF.
DVOA
OFF.
RANK
DEF.
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
1 BAL 119.3% 16.4% 3 1-0 55.2% 2 -56.8% 2 7.2% 3
2 NE 72.7% 24.0% 1 1-0 38.7% 5 -29.3% 7 4.8% 10
3 TEN 66.2% 10.6% 8 1-0 23.6% 10 -37.3% 4 5.2% 5
4 DAL 60.6% 10.7% 7 1-0 59.0% 1 0.4% 17 2.0% 14
5 OAK 50.4% -2.8% 19 1-0 33.8% 6 -7.0% 15 9.6% 2
6 SF 31.9% 1.5% 16 1-0 -25.9% 24 -58.7% 1 -0.9% 19
7 KC 22.3% 15.9% 4 1-0 51.5% 3 30.2% 25 0.9% 18
8 WAS 21.4% -13.6% 28 0-1 39.5% 4 19.2% 22 1.1% 16
9 MIN 12.8% 2.2% 14 1-0 -4.9% 17 -22.2% 11 -4.4% 24
10 GB 12.5% 3.8% 13 1-0 -32.8% 27 -40.3% 3 5.0% 7
11 SEA 10.9% 8.5% 9 1-0 -16.6% 20 -22.4% 10 5.1% 6
12 BUF 4.2% -9.1% 24 1-0 -21.7% 23 -24.8% 9 1.1% 17
13 DET 3.4% 2.2% 15 0-0-1 -3.7% 15 -24.8% 8 -17.7% 30
14 LAC 2.6% 14.8% 5 1-0 22.9% 11 22.0% 23 1.7% 15
15 PHI 0.2% 7.1% 10 1-0 32.4% 8 37.0% 29 4.9% 8
16 LAR -1.3% 13.1% 6 1-0 5.0% 14 -15.8% 14 -22.1% 32
17 NO -4.0% 18.7% 2 1-0 22.0% 12 30.8% 26 4.7% 11
18 HOU -5.3% 5.5% 11 0-1 19.1% 13 22.9% 24 -1.5% 21
19 JAX -5.8% -10.7% 25 0-1 32.7% 7 40.6% 30 2.1% 13
20 CHI -7.7% 1.5% 17 0-1 -42.0% 29 -29.5% 6 4.8% 9
21 IND -8.2% -7.2% 23 0-1 30.8% 9 17.3% 21 -21.7% 31
22 NYJ -9.7% -5.5% 21 0-1 -20.0% 22 -19.1% 12 -8.8% 28
23 CAR -10.6% 0.1% 18 0-1 -7.4% 18 -0.9% 16 -4.1% 23
24 CIN -10.6% -11.7% 26 0-1 -19.9% 21 -18.4% 13 -9.1% 29
25 ARI -20.4% -13.4% 27 0-0-1 -27.0% 25 3.4% 18 10.0% 1
26 CLE -39.8% -4.5% 20 0-1 -31.9% 26 13.8% 20 5.8% 4
27 DEN -44.3% -14.4% 29 0-1 -3.8% 16 32.8% 27 -7.7% 26
28 TB -44.4% -16.7% 30 0-1 -65.7% 32 -29.9% 5 -8.6% 27
29 ATL -55.8% -5.9% 22 0-1 -42.6% 30 7.1% 19 -6.0% 25
30 NYG -73.6% -16.9% 31 0-1 -14.1% 19 58.4% 32 -1.0% 20
31 PIT -83.8% 3.9% 12 0-1 -47.1% 31 33.4% 28 -3.3% 22
32 MIA -94.7% -37.0% 32 0-1 -41.8% 28 56.1% 31 3.3% 12

Comments

74 comments, Last at 16 Sep 2019, 3:12pm

1 I wonder why DVOA liked the…

I wonder why DVOA liked the Spreadskins more than Eagles this week, when the Eagles won pretty convincingly despite going down 17-0 in the first half

12 IIRC there is no "garbage…

IIRC there is no "garbage time" adjustment for DVOA. The FO guys tried it, but it always ended up making the system less accurate. It's a bit counterintuitive, but garbage time performance is actually a useful predictor of future performance.

32 That last touchdown drive…

That last touchdown drive wasn't garbage time. No way. What if they scored that touchdown with ~30 seconds left to go? Because they could've - they were on the 20 yard line, and the Redskins scoring a 20 yard touchdown due to a corner who slipped or some other fluke play is perfectly possible. Or a minute left to go, if the Skins hadn't had that holding penalty and instead had thrown the 18-yard pass on 4th and 12 at that point?

The point is that Philly had no guarantee that they'd be able to keep them out of the end zone until they had 6 seconds left, so all of those early plays are still bad. The fact that things worked out just means the late part of that drive was good, but the early part wasn't. For the early part of that drive, Washington was absolutely increasing their win probability non-negligibly.

29 It wasn't that convincing: a…

It wasn't that convincing: a convincing win is one in which your opponent has literally no chance to win at the end of the game, rather than just no realistic chance.

Also part of the reason it seems so convincing is because on the last two drives the Eagles traded scoring differential for ease of victory (and injury risk, as well), and most fans basically discount that as garbage time, and it's not, really: it obviously would've been better for the Eagles to not let the Redskins even have a chance at victory at the end, however small, but given the main weakness on the Eagles defense (the secondary) it made a *lot* more sense to just burn the clock.

44 They still were playing with…

They still were playing with starters on defense, and they were actively trying to disrupt the plays. They got the Redskins to 4th down twice, after all. Of course they didn't think it was garbage time. A prevent defense is absolutely not the same thing as garbage time.

I mean, yes, the Eagles obviously traded scoring differential for clock on that last drive, but it's not like the Redskins weren't improving their (small) win percentage by actually completing some of those plays, and it's not like the Eagles weren't *trying* to stop them.

61 I basically agree with you,…

I basically agree with you, in that you can't just say "they didn't pull their starters, so it's not garbage time," especially it's not exactly like teams have a "starting team" and "backup team" anyway. Not that simple.

That being said, if you look at the NE/Pittsburgh game (and by the way, it was up 27, not up 30), I don't think New England considered the 4th quarter drive garbage time at all. That was a clock-killing drive - it started with over 12 minutes to go in the 4th quarter - so it's not garbage time. They had a goal (to kill clock) and they succeeded, so why would you ignore or discount the fact that Pittsburgh couldn't stop them?

Now the very last drive for New England was definitely garbage time for them: run the ball each down, don't care what the result is, just punt it back. There it makes sense to ignore that drive, because New England literally didn't care about the result. I don't think Pittsburgh ever gave up on offense at all, for whatever reason, but hell if I understand their coach's intentions after that field goal.

I think people just use "garbage time" too loosely: there's a difference between when a team's goal shifts from scoring to clock-killing and when a team literally has no game goal at all for a drive (see the Giants Daniel Jones drive, or Robert Griffin/Josh Rosen drives in the Baltimore/Miami game) and they're just using it for future evaluation. It doesn't make sense to ignore clock-management drives (note that DVOA compares teams in similar situations, so the fact that teams shift to clock-management in the 4th is handled that way). It does make sense to ignore drives where a team no longer cares about the result of the drive at all.

63 Yup, totally agree, which is…

Yup, totally agree, which is why I think it's not terribly surprising that there's no easy way to exclude those plays without losing accuracy. I think you can usually tell watching the game just from the urgency of the players and watching the defensive scheme, but from the play by play you're probably screwed.

64 Even meaningless drives have…

Even meaningless drives have meaning.

I've seen drives where the game's result was a foregone conclusion, but both coaches opted to treat the drive as if it were meaningful, at least as far as play calls go. So long as everyone knows what's going on, it's an excellent chance to get reps at game speed and in game situations.

You wouldn't break out some secret trick play. But you would work on situational awareness, and execution.

In the Pittsburgh game, the Patriots called a defensive time out in the waning seconds to get into the correct defensive alignment for the time remaining, acting as if a score would matter. I would expect there was a discussion later about the proper defense for 6 seconds left defending at your own 26. Or maybe they called the timeout because the clock had ticked down past the one-play-left threshold. Either way, both teams were (mostly) treating it as if it was a meaningful drive.

Which means that it's good data for DVOA.

66 In the NE/Pittsburgh game…

In the NE/Pittsburgh game the only drive that I think was completely throwaway was the Patriots last drive. I think that Pittsburgh was treating it as meaningful through the whole game.

"Even meaningless drives have meaning."

That's not the threshold, though. You need to make sure you're measuring drives/plays such that the goal for all teams is the same, in the same situation. For the Giants on Sunday, by the late point in the game, winning wasn't really an option anymore, so seeing Jones in at QB for a few plays was more valuable, long term, than improving their tiny win percentage. But other teams might have different goals in that situation. So you can't compare the Giants performance on those plays to other teams, and expect it to necessarily have predictive value.

If DVOA didn't do "apples to apples" situational comparison, for instance, it wouldn't make sense to include the Redskins last drive in the Eagles game to the baseline value for a drive in the first quarter. The two situations are different. But instead, you only compare the results of the plays in that drive to a baseline from plays in the same situation: that is, trailing by 2 scores late in the 4th quarter.

Obviously some drives still have meaning outside the context of the game, but without knowing what the team was trying to do, you can't compare that to the right baseline.

70 PFR's win probability…

PFR's win probability calculator also says that you'd have to be a total idiot to kick the field goal when the Steelers did, so clearly the Steelers weren't playing logically, and the Patriots responded in kind.

Plus the entire point of what I'm saying is that you can't just figure out when teams are going to treat the rest of the game as a foregone conclusion just by the box score. Some teams decide it's important to get starters out to minimize risk, others for whatever reason insist on moral victories and continuing to play hard.

It's not even like coaches do the same thing all the time, either. You can't say "well, Belichick never pulls his starters in garbage time" - yes, he has, and sometimes questionably, such as back in 2009 when he pulled Brady, Moss and Welker when down 21 points with 5 minutes remaining. 

However, in this specific instance, for some reason, Belichick has some belief that it's possible for teams to come back from a 27-point deficit with about 12 minutes remaining. How do I know this? Because he said so in January, when they faced the exact same situation.

"“We’re trying to score until I thought the game got to the point where they didn’t really have enough possessions to catch us,” he said about his rationale behind the decision to not pull Brady and other key starters from the contest. “Look, we’ve seen Peyton Manning come down from 21 points in four minutes and win so it’s not over until it’s over in this league. I know I’m in the minority on that but until it’s over, you’re trying to score."

Which again, just reinforces my point. Up 27 with 12 minutes remaining, Belichick isn't treating it like garbage time, he wants to score because he's apparently afraid of Manning's ghost.

4 Denver is clearly ranked too…

Denver is clearly ranked too high because there is no way 5 teams were worse. I'll accept the Fish Tank as lower, but no one else. Also, Grooden sucks.

5 Cardinals Special Teams

The Card's special teams results caused me to go to PFR to see how they did it; Turns out they didn't really have an outlier KR or PR, they did it with 4 made FG (none especially long) and 8 good punts netting them 66 yards of field position vs. the opponents 8 average punts, and good kick/punt coverage.

On a side note, can anyone who watched this game explain to me how they had time for 13/14 drives a side in this game (vs. roughly 10 typical). There was still somehow time for 55 rushing plays to run the clock.

20 “can anyone who watched this…

“can anyone who watched this game explain to me how they had time for 13/14 drives a side in this game (vs. roughly 10 typical). There was still somehow time for 55 rushing plays to run the clock.””

A)Each team had two overtime drives.

B)Most of the drives were very short.  A lot of the drives in this game were either 3/4 and out (at one point in the 3rd quarter both sides traded 3 and outs twice), or ended in quick turnovers after 1-2 plays.

23 I should have noted

I should have noted I was referring to 13/14 drives in regular time, they actually got 15 & 16 drives respectively inclusive of overtime. Looking at the drive chart it looks like this is just a quirk of a massive number of 3-outs with all of the Card's long drives (by number of plays) coming in the 4th quarter where they were playing fast. The deeper question here was that the cardinals had 10 straight drives resulting in 11 or less yards to begin the game, the drives of 42yd to FG, 68yd to FG, 70yd to TD, 55yd to TD, as the final four drives of regular time. What changed in the game to alter the dynamic that dramatically?

30 changes in drive success

Obligatory "not a Cards fan, didn't see game" disclaimer.

I would opine that a rookie head coach and a rookie QB, who both have some public, vocal doubters, finally got past some nervousness, uncertainty, etc. I believe I read a quote where the coach said that basically. I would also guess that the HC/OC/playcaller (I thought I read that the HC is calling the plays?) started figuring out which plays were working and which ones weren't.

36 Ah, now I see your question…

In reply to by sbond101

Ah, now I see your question.  I wish I had concrete explanation, but I didn't see one other than "the Cardinals offense just started playing better."  I do think fatigue on the part of Detroit's defense had something to do with it.  Early on, the Cardinals were playing at a fast pace, and kept failing.  Later on, they continued to play at a fast pace, but started having more success.

If you want to point to single turning point, to me it was Larry Fitzgerald making an awesome diving 41 yard catch to convert a 3rd and 14, when Arizona was down 24-6.  It only led to a field goal, but it cut the score to a two possession game.  The Cardinals just looked more confident after that, and Murray seemed to settle down.  I know that narrative doesn't really fit into advanced stats, but it really appeared that way to me as I watched the game in real time.  After the Lions went 3 and out on the subsequent drive, I said "uh oh" to myself.

22 You run the ball and go 3 and out

and there are lots of drives in the games. 1st downs are the number one deteriminant of drive time, not rushing. Of course if you run a lot and get first downs, then that runs the clock the fastest.

6 All San Francisco needs is…

All San Francisco needs is two pick-sixes a week for the rest of the season, and they might hang on to that #1 defensive VOA.

35 Templated

SAN FRANCISCO is clearly ranked TOO HIGH because Jameis Winston is prone to interceptions in a way that can't be captured in a simple efficiency number. PFF's preseason secondary grades are way better than this. Jhon Lunch is the wrost GM in FROOTBALL!! Every pcik in the first three ruonds for three years has been a BUST!!

7 Ravens Week 1 dominance

It's actually three years in a row the Ravens have balled out in Week 1.

In 2017, they won 20-0 against Cincinnati, and were #2 in VOA after Week 1 that year at 82.9% (The Rams who killed the Luck-less Colts were #1)

8 following complaint template to the letter!

<PATRIOTS> is clearly ranked <TOO LOW> because <TB&BB ARE GOATZ>. <MY PERSONAL BELIEF> is way better than this. <AB& J GORDON ARE CRAYCRAY BUTT AS LONNG AS DEY CATTCH TUCHDOWNNS FROM TOMM BRADYZ I LIKE DEM ALLOT>

11 Part of the reason Baltimore…

Part of the reason Baltimore won so handily in week 1 of 2018 was the Nathan Peterman was completely useless.

The good news for Miami is that they don't have a starting QB as bad as Nathan Peterman. The bad news for Miami is that they were roughly as bad as a team whose starting QB was Nathan Peterman.

13 «Citation Needed» Joking…

«Citation Needed»

Joking aside, Miami has huge downside potential going forward, given that they're in the midst of Operation Fish Tank.  Buffalo was trying to win games after The Nathan Peterman Experience was canceled.

14 Given that DVOA has no opponent adjustments for next week...

...I wonder how ugly NE @ MIA can end up in the final tally.
If MIA keeps that kind of performance and overall level of effort up, I wouldn't take something like -170% off the table.
Reports are saying that multiple players are actively seeking trades. That's always a good sign after week 1.

53 Last I saw the game was a 19…

Last I saw the game was a 19 point spread.  If we swing 3 points each side for home field, that would make it a 25 point spread in New England.  That doesn't quite match the 26.5 points the Broncos were favored over Jacksonville in 2013.