Week 3 DVOA Ratings
by Aaron Schatz
The 2019 NFL season has been very stratified so far. It certainly seems like we've seen some extremes when it comes to easy and hard early schedules, and the result has been more undefeated teams than usual. Last year, there were only three undefeated teams through Week 3. This year, there are eight undefeated teams if we include the 2-0-1 Detroit Lions. All eight of these teams rank in the top 11 of our DVOA ratings along with three 2-1 teams.
Of course, the story of the season is still at the extremes. New England remains No. 1 in DVOA after their 30-16 victory over the New York Jets. They currently have the fourth highest DVOA ever measured through three games. The only teams they trail are three of the greatest teams of all-time: the Super Bowl champion 1996 Packers, the 16-0 2007 Patriots, and the Super Bowl champion (and best team in DVOA history) 1991 Redskins.
Since someone asked on Twitter, I'll point out that the Patriots' DVOA would only be 4.0% higher if we removed the two Jets touchdowns. The muffed punt is not penalized as much as you might expect since muffed punts are recovered more often than most other fumbles, and the pick-six is not penalized as much as you might expect because we treat all interception returns equal based on the length of the pass and the location of the interception, plus the pick-six took place in one of the rare places where DVOA does discount "garbage time" (score gap of 22 points or more, not inside the red zone). Still, if you added 4.0% onto the Patriots' DVOA rating, they would be higher than the 2007 Patriots after three games.
The Dallas Cowboys move into the No. 2 spot after dismantling the Miami Dolphins, even though that game was surprisingly close through the first half. The Cowboys pass the San Francisco 49ers, Kansas City Chiefs, and Baltimore Ravens, who each drop one spot. Despite playing each other with no opponent adjustments applied yet, the Chiefs and Ravens stay above 40% DVOA. That means we have five different teams still above 40% after Week 3, which has only happened three other times (2005, 2007, and 2009).
Of course, the extremes at the top are nothing compared to the extreme at the bottom. Miami is very, very bad. Miami is the worst team DVOA has ever tracked through three weeks. Miami is the fifth-worst offense ever tracked by DVOA through three weeks and the fifth-worst defense ever tracked by DVOA through three weeks. Or three games -- once we get to Week 3, we start running into teams that had a really early bye, so these tables represent the best and worst teams ever through three games rather than three weeks. Teams with asterisks were after Week 4 so their ratings do include some opponent adjustments. The other teams, like the 2019 Patriots and Dolphins, are being measured without any opponent adjustments yet.
| BEST TOTAL DVOA AFTER 3 GAMES, 1986-2019 |
WORST TOTAL DVOA AFTER 3 GAMES, 1986-2019 |
|||||
| Year | Team | DVOA | Year | Team | DVOA | |
| 1996 | GB | 94.4% | x | 2019 | MIA | -109.3% |
| 2007 | NE | 87.1% | x | 2001 | WAS | -92.1% |
| 1991 | WAS | 86.6% | x | 2013 | JAX | -91.0% |
| 2019 | NE | 83.6% | x | 2008 | STL | -85.8% |
| 2007 | PIT | 77.0% | x | 1993 | TB* | -84.0% |
| 2015 | ARI | 76.4% | x | 2000 | CIN* | -82.7% |
| 1990 | CHI | 69.7% | x | 2007 | NO | -74.9% |
| 2005 | CIN | 69.6% | x | 1999 | CLE | -73.0% |
| 2013 | SEA | 69.2% | x | 2008 | DET | -72.2% |
| 2002 | SD | 68.9% | x | 2015 | CHI | -70.2% |
| 2013 | DEN | 68.8% | x | 2009 | CLE | -69.9% |
| 2001 | SD | 68.2% | x | 1996 | TB | -69.8% |
Obviously, with the extreme early schedules leading to some extreme early ratings, I'm sure there's some curiosity about what things would look like with opponent adjustments. After all, the Patriots are dominating but all three of their opponents are 0-3. The Cowboys' opponents are 1-8. This season's early surprise, the 49ers? Their opponents are 1-8. On the other end, Miami's opponents are 8-1. Cincinnati's opponents are 8-1. Pittsburgh's opponents are 8-1.
So I decided to do something I also did after Week 3 a couple of years ago. I created two "what if" ratings that include opponent adjustments. One has opponent adjustments at 30 percent strength, which is theoretically what we might use on the same scale that has us introduce them next week at 40 percent strength. The other version has opponent adjustments at 100 percent strength, but only based on these three weeks of the season.
Two things may stand out about these alternative DVOA ratings. First, the results aren't as different from the regular DVOA ratings as you might expect, even at the extremes. Each of the alternative ratings still has four teams over 40%. The Patriots drop in each one to around 70%. The Dolphins only move up to around -90% in each one. In other words, even if we were including full strength opponent adjustments right now, the Dolphins would be the worst team ever tracked when compared to unadjusted past teams.
Second, you may be surprised to see that some teams will move down with 30% adjustments but then up again with 100% adjustments, or vice versa. That is because the 100% adjustment is not automatically just 3.3x whatever the 30% adjustment is, because of the way creating the adjustments has multiple runthroughs that adjust the defenses, than adjust offenses based on those adjusted defenses, and so on.
| Team | W-L | No Adj | Rk | 30% Adj | Rk | 100% Adj | Rk |
|---|---|---|---|---|---|---|---|
| NE | 3-0 | 83.6% | 1 | 69.5% | 1 | 74.7% | 1 |
| DAL | 3-0 | 62.4% | 2 | 49.6% | 2 | 45.7% | 3 |
| SF | 3-0 | 53.5% | 3 | 47.2% | 3 | 53.4% | 2 |
| KC | 3-0 | 41.2% | 4 | 41.7% | 4 | 42.0% | 4 |
| BAL | 2-1 | 40.1% | 5 | 33.1% | 5 | 34.7% | 5 |
| GB | 3-0 | 24.3% | 6 | 25.5% | 6 | 28.4% | 6 |
| SEA | 2-1 | 20.5% | 7 | 11.6% | 8 | 18.0% | 7 |
| LAR | 3-0 | 17.1% | 8 | 14.1% | 7 | 12.0% | 9 |
| BUF | 3-0 | 11.5% | 9 | 2.7% | 15 | 3.3% | 15 |
| MIN | 2-1 | 9.5% | 10 | 10.6% | 9 | 15.1% | 8 |
| DET | 2-0-1 | 9.5% | 11 | 4.6% | 12 | -1.6% | 17 |
| HOU | 2-1 | 6.1% | 12 | 2.2% | 16 | 5.6% | 14 |
| TEN | 1-2 | 5.7% | 13 | 4.5% | 13 | 6.6% | 12 |
| CHI | 2-1 | 5.6% | 14 | 3.0% | 14 | 1.5% | 16 |
| CAR | 1-2 | 5.6% | 15 | 5.5% | 11 | 7.3% | 11 |
| JAX | 1-2 | 2.1% | 16 | 5.6% | 10 | 11.2% | 10 |
| CLE | 1-2 | -2.0% | 17 | -4.2% | 18 | -3.4% | 20 |
| PHI | 1-2 | -4.5% | 18 | -5.4% | 20 | -3.0% | 18 |
| IND | 2-1 | -4.9% | 19 | -4.9% | 19 | -3.1% | 19 |
| TB | 1-2 | -5.6% | 20 | -2.7% | 17 | 5.9% | 13 |
| NO | 2-1 | -10.9% | 21 | -6.4% | 21 | -13.1% | 22 |
| ATL | 1-2 | -13.1% | 22 | -10.9% | 23 | -9.3% | 21 |
| WAS | 0-3 | -14.3% | 23 | -9.9% | 22 | -13.7% | 23 |
| LAC | 1-2 | -19.5% | 24 | -17.4% | 24 | -20.9% | 24 |
| DEN | 0-3 | -26.5% | 25 | -26.1% | 27 | -24.8% | 26 |
| NYG | 1-2 | -29.7% | 26 | -23.8% | 25 | -28.8% | 27 |
| ARI | 0-2-1 | -29.8% | 27 | -26.0% | 26 | -30.9% | 28 |
| NYJ | 0-3 | -32.5% | 28 | -26.4% | 28 | -40.6% | 30 |
| OAK | 1-2 | -35.9% | 29 | -31.0% | 29 | -21.0% | 25 |
| PIT | 0-3 | -48.5% | 30 | -34.8% | 30 | -37.0% | 29 |
| CIN | 0-3 | -50.6% | 31 | -44.5% | 31 | -55.4% | 31 |
| MIA | 0-3 | -109.3% | 32 | -94.0% | 32 | -94.8% | 32 |
As you might expect, the teams that take the biggest hit when we introduce opponent adjustments include the Patriots and Cowboys, because they were so far ahead of the rest of the league anyway. You might be surprised to see that with 100% adjustments, the San Francisco 49ers move ahead of the Cowboys. The 49ers are playing really well right now. They stomped two bad teams in the first two weeks, and while their game was close with the Steelers, part of that was very poor fumble recovery luck. The 49ers have recovered just 1 of 5 fumbles on offense and 2 of 6 on defense.
Taking a big hit in the ratings and also in the rankings are the Detroit Lions and the Buffalo Bills. Detroit falls from 11th to 12th with 30% adjustments and then down to 17th with 100% adjustments. Buffalo drops from ninth to 15th with both sets of adjustments. I will admit that those rankings feel a bit more in line with how these teams have played this year, suggesting that maybe we do need to incorporate opponent adjustments even after only three weeks of the season.
Also taking an unexpceted fall in the 100% adjustments are the New York Jets! They go up in the 30% adjustments, then way down in the 100% adjustments. I apologize for not having the time this afternoon to go through all the adjustments to figure out why this is the case, but besides the Patriots their opponents are ranked 15th (Buffalo) and 20th (Cleveland) in the 100% adjusted version of the ratings. That still seems like it shouldn't result in their rating taking such a big adjustment, but their run-pass ratios must somehow match where the Bills and Browns are at their worst (and the limited number of places where the Patriots aren't extremely good this season).
Can it possibly get worse, Jets fans? Actually, it will probably get better, because Sam Darnold will be coming back healthy to face the easiest remaining schedule in the league by both current DVOA ratings and DAVE ratings.
At the bottom of the ratings, the biggest gainer is Oakland, which moves up from 29th to 25th in the 100% adjusted ratings. Pittsburgh moves up from 30th to 29th. Miami moves up in rating, although they are still horrific. The big movers in the rankings are the other two Florida teams. Tampa Bay goes from 20th in the unadjusted ratings to 13th in the 100% adjusted verison. Jacksonville goes from 15th in the unadjusted ratings to 10th in the 100% adjusted version. The Jaguars have played three above-average opponents this year: they only lost to Kansas City by 14, they basically tied Houston (but lost because they went for 2 and failed), and they beat Tennessee handily.
And, what the heck, we'll run one more big table here. Let's run a schedule strength table a week early, only this is based on DAVE rather than on DVOA. These DAVE ratings assume Drew Brees returns Week 12 and Sam Darnold returns Week 5, and ignores the small difference I'm currently using between Kyle Allen and Cam Newton. (Past DAVE is based on past quarterbacks, so it counts Houston as playing Brees but counts the Browns and Patriots as playing Jets backups.)
Things are going to get a lot easier for the Jets, Steelers, and Jaguars, but a lot harder for the Chicago Bears. You can discuss further in the comment thread.
| Team | Past Avg. DAVE |
Rk | Future Avg. DAVE |
Rk |
|---|---|---|---|---|
| CHI | -6.7% | 26 | 5.4% | 1 |
| DEN | -0.5% | 16 | 4.0% | 2 |
| SF | -15.6% | 29 | 3.9% | 3 |
| MIN | -3.0% | 21 | 3.7% | 4 |
| CAR | -3.6% | 23 | 3.3% | 5 |
| SEA | -10.7% | 27 | 3.3% | 6 |
| GB | -2.0% | 19 | 3.1% | 7 |
| DAL | -30.3% | 31 | 3.0% | 8 |
| ATL | 1.4% | 15 | 2.8% | 9 |
| KC | -1.9% | 18 | 2.8% | 10 |
| HOU | 6.0% | 9 | 2.2% | 11 |
| ARI | 7.3% | 8 | 1.7% | 12 |
| TB | -2.1% | 20 | 1.6% | 13 |
| LAR | 1.5% | 14 | 1.3% | 14 |
| OAK | 4.5% | 10 | 0.6% | 15 |
| DET | -1.8% | 17 | 0.5% | 16 |
| TEN | -4.0% | 24 | 0.3% | 17 |
| BAL | -15.4% | 28 | 0.1% | 18 |
| NO | 12.2% | 4 | -0.5% | 19 |
| WAS | 10.9% | 6 | -0.8% | 20 |
| PHI | -4.6% | 25 | -1.0% | 21 |
| NYG | 3.3% | 11 | -1.4% | 22 |
| BUF | -19.0% | 30 | -2.9% | 23 |
| CIN | 8.5% | 7 | -3.5% | 24 |
| NE | -33.2% | 32 | -4.3% | 25 |
| IND | 2.9% | 12 | -4.5% | 26 |
| MIA | 25.8% | 1 | -4.7% | 27 |
| CLE | -3.4% | 22 | -4.8% | 28 |
| LAC | 2.4% | 13 | -4.8% | 29 |
| JAX | 12.4% | 3 | -5.1% | 30 |
| PIT | 22.0% | 2 | -6.2% | 31 |
| NYJ | 11.5% | 5 | -9.5% | 32 |
* * * * *
All stats pages should now be updated through Week 3, including snap counts and playoff odds, and the FO Premium DVOA database. The Matchups view in the DVOA database has been updated so that it now displays 2019 ratings instead of 2018 ratings.
Another exciting announcement: we should be introducing our first Football Outsiders stars for Madden 20 Ultimate Team next week! They will feature outstanding September performers.
* * * * *
These are the Football Outsiders team efficiency ratings through three weeks 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 70 percent of DAVE. In part because of the extreme schedules this year leading to some extreme early ratings, I've decided to increase the amount that I incorporate the preseason projection into DAVE for the next few weeks. (It was 60 percent for a few years, then 65 percent, and now is 70 percent.)
DAVE ratings represent what we expect a team's rating to be at the end of the season, so for certain teams the offensive ratings are adjusted to split the difference between two different quarterbacks who are expected to start for the remainder of the year (Carolina, New Orleans, New York Jets).
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 | OFF. DVOA |
OFF. RANK |
DEF. DVOA |
DEF. RANK |
S.T. DVOA |
S.T. RANK |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | NE | 83.6% | 1 | 37.6% | 1 | 3-0 | 25.5% | 4 | -61.0% | 1 | -2.9% | 22 |
| 2 | DAL | 62.4% | 5 | 23.3% | 3 | 3-0 | 54.1% | 1 | -7.2% | 14 | 1.1% | 10 |
| 3 | SF | 53.5% | 2 | 15.7% | 6 | 3-0 | 12.3% | 7 | -42.2% | 2 | -1.0% | 14 |
| 4 | KC | 41.2% | 3 | 23.9% | 2 | 3-0 | 46.1% | 2 | 5.1% | 21 | 0.1% | 11 |
| 5 | BAL | 40.1% | 4 | 16.4% | 4 | 2-1 | 36.1% | 3 | 4.1% | 19 | 8.1% | 2 |
| 6 | GB | 24.3% | 10 | 10.2% | 9 | 3-0 | -5.0% | 20 | -27.5% | 3 | 1.9% | 8 |
| 7 | SEA | 20.5% | 7 | 12.8% | 7 | 2-1 | 20.3% | 5 | -3.4% | 15 | -3.2% | 23 |
| 8 | LAR | 17.1% | 6 | 16.3% | 5 | 3-0 | -0.7% | 16 | -22.9% | 4 | -5.1% | 28 |
| 9 | BUF | 11.5% | 12 | -3.0% | 19 | 3-0 | -3.2% | 18 | -18.9% | 6 | -4.2% | 26 |
| 10 | MIN | 9.5% | 14 | 4.5% | 15 | 2-1 | 5.7% | 11 | -7.9% | 13 | -4.1% | 25 |
| 11 | DET | 9.5% | 11 | 5.2% | 13 | 2-0-1 | 1.6% | 13 | -10.0% | 10 | -2.1% | 18 |
| 12 | HOU | 6.1% | 15 | 7.4% | 10 | 2-1 | 7.1% | 9 | 0.4% | 17 | -0.5% | 13 |
| 13 | TEN | 5.7% | 8 | 5.8% | 12 | 1-2 | -1.1% | 17 | -8.2% | 12 | -1.3% | 16 |
| 14 | CHI | 5.6% | 16 | 4.4% | 16 | 2-1 | -12.8% | 24 | -19.9% | 5 | -1.5% | 17 |
| 15 | CAR | 5.6% | 24 | 1.9% | 17 | 1-2 | -7.5% | 21 | -10.1% | 9 | 3.1% | 5 |
| 16 | JAX | 2.1% | 25 | -6.3% | 22 | 1-2 | 0.6% | 14 | 2.5% | 18 | 3.9% | 4 |
| 17 | CLE | -2.0% | 21 | -0.1% | 18 | 1-2 | -24.1% | 28 | -16.2% | 8 | 5.9% | 3 |
| 18 | PHI | -4.5% | 13 | 5.1% | 14 | 1-2 | 5.7% | 10 | 4.7% | 20 | -5.5% | 30 |
| 19 | IND | -4.9% | 9 | -5.5% | 21 | 2-1 | 19.9% | 6 | 16.5% | 25 | -8.3% | 32 |
| 20 | TB | -5.6% | 18 | -10.3% | 23 | 1-2 | -16.3% | 26 | -18.9% | 7 | -8.2% | 31 |
| 21 | NO | -10.9% | 28 | 11.0% | 8 | 2-1 | 0.4% | 15 | 21.7% | 28 | 10.4% | 1 |
| 22 | ATL | -13.1% | 27 | -3.3% | 20 | 1-2 | -8.5% | 23 | 0.2% | 16 | -4.5% | 27 |
| 23 | WAS | -14.3% | 17 | -15.6% | 26 | 0-3 | 4.7% | 12 | 21.4% | 27 | 2.4% | 7 |
| 24 | LAC | -19.5% | 19 | 6.4% | 11 | 1-2 | 9.9% | 8 | 24.1% | 30 | -5.3% | 29 |
| 25 | DEN | -26.5% | 22 | -14.8% | 25 | 0-3 | -8.1% | 22 | 15.9% | 23 | -2.5% | 19 |
| 26 | NYG | -29.7% | 29 | -22.1% | 30 | 1-2 | -3.3% | 19 | 29.0% | 31 | 2.6% | 6 |
| 27 | ARI | -29.8% | 26 | -16.8% | 28 | 0-2-1 | -17.8% | 27 | 13.8% | 22 | 1.8% | 9 |
| 28 | NYJ | -32.5% | 20 | -18.4% | 29 | 0-3 | -40.6% | 31 | -9.2% | 11 | -1.0% | 15 |
| 29 | OAK | -35.9% | 23 | -15.9% | 27 | 1-2 | -14.0% | 25 | 21.9% | 29 | 0.0% | 12 |
| 30 | PIT | -48.5% | 30 | -14.0% | 24 | 0-3 | -24.7% | 29 | 21.3% | 26 | -2.5% | 20 |
| 31 | CIN | -50.6% | 31 | -22.5% | 31 | 0-3 | -31.5% | 30 | 16.4% | 24 | -2.6% | 21 |
| 32 | MIA | -109.3% | 32 | -53.3% | 32 | 0-3 | -60.6% | 32 | 45.3% | 32 | -3.4% | 24 |
Comments
56 comments, Last at 26 Sep 2019, 1:52pm
#2 by BlueStarDude // Sep 24, 2019 - 7:47pm
It looks like it will take a huge hit once opponent adjustments kick in, but was wondering if you would chart the Dallas offense this week.
They have a 54.1% offensive DVOA through week 3. If I’m looking at the table you did last year correctly, 54.1% would make this the third best offensive performance at this point of the season (behind ’99 Washington and ’98 Denver).
Last year the Chiefs’ 50.6% offensive performance was enough to get them called a juggernaut. This year Dallas goes unnoticed. Still, it’s understandable: the Chiefs were pretty obviously amazing, while the Cowboys have a lot to prove. But DVOA analysis is supposed to cut through the eyeball test.
#31 by andrew // Sep 25, 2019 - 8:34am
That is part of the reason we don't normally use defensive adjustments so early, as thus far 33% (give or take based on # of plays) a team's rating is based on what they did vs the team you are adjusting for.
Is there (or would it be possible) o do a Defense Adjusted for Everyone Else VOA? I know it would play havoc for rankings because it would mean every team's opponents defenses would be different, but I would be interested to see what that would look like....
#36 by Aaron Schatz // Sep 25, 2019 - 11:11am
Yes, I had other things to concentrate on this week, so I stuck to just the all-time total DVOA best and worst list. But you are correct. The Cowboys have the No. 3 offense ever through Week 3. I promise to update that table next week once the opponent adjustment kicks in (although the 2019 Saints defense has been poor so far, so that's another team that will get the Cowboys offense adjusted downward).
#50 by BlueStarDude // Sep 25, 2019 - 7:28pm
Always easy for me to remember how long since I learned of FO because it was from a NYT Magazine article on the weekend my daughter was born. Will be 16 years pretty soon.
The Fox mailbag (or whatever it was called) was the best though. ;-)
#3 by dank067 // Sep 24, 2019 - 7:55pm
In spite of how bad Jared Goff played in the Super Bowl and a few other games down the stretch last year, I was totally fine with the extension they gave him and thought it was completely reasonable to expect him to continue producing at a high level for them overall.
So... here he sits at 26th among QBs in VOA and YAR through 3 weeks, and sites like PFF aren't much more generous (17th). Rams offense sitting at 16th. And for a guy who you could reasonably defend with "he might struggle to read the defense, but he can make some incredible throws," he has missed a bunch of wide open receivers down the field.
#42 by theslothook // Sep 25, 2019 - 2:09pm
I was in the camp that was leery of handing him a big contract. Having watch him thus far, he's been fairly mediocre. Not awful, but he's definitely struggling with accuracy. I can't tell if it's early jitters or loss of confidence, but that particular hole in his game should improve. It's the rest of the holes he needs to close
#4 by poplar cove // Sep 24, 2019 - 7:57pm
Looks like DAVE has 7 of top 10 teams from NFC and 13 of top 17 as well.
Might be something to keep our eyes on going forward as things seem to be most lopsided I can recall between the two leagues in a long time. Granted it does look like the AFC does have the two best teamsb overall in the NFL right now with the Chiefs and Patriots.
#6 by ssereb // Sep 24, 2019 - 8:12pm
I count the AFC with 4 of the top 10 (NE, KC, BAL, HOU) and 6 of the top 17 (TEN and LAC plus the previous 4). Agreed that the middle and bottom of the AFC is extremely weak, but DAVE's use of the preseason projections is holding down the Jaguars and Bills (though it's also propping up the Chargers). The AFC should have enough decent teams to make the playoffs interesting, though not much more.
#8 by ssereb // Sep 24, 2019 - 8:23pm
I think the NFC will definitely have the better record in interconference games (though maybe not quite as dramatic as the record so far), but as a viewer I'm more interested in the teams at the top than the teams in the middle or at the bottom since nothing could possibly compel me to watch Miami-Washington.
#18 by Richie // Sep 25, 2019 - 12:03am
It's still early to be sure which teams are good or bad, but it looks like most of the good NFC teams have played against the bad AFC teams. I think that 11-3 record might be a bit of a mirage.
The 11 NFC wins are from SEA (2), NO, CHI, DET, SF (2), Rams, GB, MIN and DAL. The top 8 NFC teams by DVOA have all played an AFC team.
The 11 AFC losses are from CIN (2), DEN (2), OAK, MIA. Plus these teams that were expected to be playoff contenders this year: HOU, PIT (2), LAC, CLE.
That leaves the top 4 AFC teams by DVOA that either haven't played against the NFC, or won those game: NE, KC, BAL, BUF.
#5 by thok // Sep 24, 2019 - 8:11pm
I've been looking at Pro-Football-Reference's Simple Rating System, and it's been having issues with opponent adjustments as well. For example, it has Carolina above Dallas, mainly because it's calculations end up attributing Dallas MOV almost entirely due to the extremely weak slate of teams they've played; in particular, SRS doesn't really know what to do with Miami. Hopefully a few more weeks will clear things up as Miami stops playing elite offenses, and we get less matchups between 3-0 teams and 0-3 teams.
#9 by Lost Ti-Cats Fan // Sep 24, 2019 - 8:30pm
"SRS doesn't really know what to do with Miami."
Unless something changes, Miami's going to play havoc with a lot of statistical analysis. A lot of metrics have a base assumption that all opponents are NFL-caliber teams of varying skill sets, not 31 NFL teams and a Pop Warner team accidentally included on the schedule.
Take strength of schedule, for example. You can only beat Miami once per game. A normally bad team at -20% or -30% still has a chance of winning, so tossing them into strength of schedule makes sense. Once you get to Miami's numbers though, whether they're at -100% or -500%, the numbers (and the eye test) say they have no reasonable chance of winning against a "normal" 0% DVOA team. Your schedule probably isn't easier because you play Miami followed by a pair of 0% teams than if you played Arizona, Oakland, and the Jets three weeks in a row, even if they have roughly the same average. Expected win stats, on the other hand, would do a better job of handling an outlier.
Maybe college football stats have a way of dealing with cupcake opponents when it comes to offensive and defensive efficiency. I doubt most NFL sites would have bothered to think about the issue. At least not until this year.
#11 by Cythammer // Sep 24, 2019 - 9:27pm
The Dolphins have played three teams that were in the playoffs last year. All three of those teams look strong this year too. Miami won't look this bad all year because they'll be playing plenty of bad teams going forward.
#38 by Lost Ti-Cats Fan // Sep 25, 2019 - 11:35am
That's a fair point. For the Dolphins to be as bad as I suggested would require that they lose big to other below-average teams, not just get blown out by teams who look to be top tier even when evaluated against non-Miami opponents.
I expect them to get whumped by Keenum-led Washington, the Darnoldson-less Jets, etc., but until those games are played, we won't know. Plus more time for the defense to pick up Flores' schemes should help as the season goes on. (I'd like to suggest the offence could improve, too, but short of a talent-transfusion, I'm not sure it will.)
#19 by Richie // Sep 25, 2019 - 12:05am
SRS says Miami should lose to San Diego by about 29 points this week. Would you be surprised if that happened? I wouldn't.
I love that PFR shows Miami ahead of New York in the standings. (Because they sort ties alphabetically.)
#43 by DraftMan // Sep 25, 2019 - 2:13pm
A bigger surprise than the Chargers successfully managing to pack up all their stuff in a hurry, move the team back to San Diego on such short notice, AND have everyone fly out to Miami in time for a 1:00 game to whoop up on the Dolphins?
#44 by Aaron Brooks G… // Sep 25, 2019 - 2:18pm
I believe their formal name is the "San Diego Chargers of Los Angeles", as LA has refused to let them use the Los Angeles moniker.
Although there is some chatter that Riverside may accept them, resulting in the Riverside California Chargers of Los Angeles, or RC COLA.
#24 by RevBackjoy // Sep 25, 2019 - 2:32am
SRS has a pretty high correlation with DVOA by the end of the year (R^2 around 0.9), but is too wonky during the first few weeks to really mean anything. They'll converge as the season goes on.
Either that, or the Vikings really are 5 points per game better than the Patriots!
#33 by Aaron Brooks G… // Sep 25, 2019 - 8:55am
For some reason, the PFR link to the Miami Dolphins team page keeps redirecting to https://www.sports-reference.com/cfb/. Not sure why.
Currently SRS says the Pats are 42-pt favorites vs Miami. That feels about right.
#12 by Cythammer // Sep 24, 2019 - 9:31pm
I suspected the Bills would still be in the bottom half of the league in DAVE despite their 3-0 record, and indeed, it is so. There are five 1-2 teams ahead of them in DAVE. I think the DAVE projections are about right, but it might not matter for making the playoffs. They've already banked those three wins and with an easy schedule still have a lot of opportunities for wins going forward. If they can make it that would mark two times in three seasons the Bills qualify for the postseason without actually being particularly good.
#28 by Mike B. In Va // Sep 25, 2019 - 5:23am
So far, the eye test says this team is better than the last team to make the playoffs. At least they played the merely awful and not putrid version of the Jets, so that game was harder than it looks on paper. Still, I have doubts as to the team being in the top ten by midseason, yet they may have a top ten record. Odd year.
#13 by TomC // Sep 24, 2019 - 10:04pm
One thing that randomly stands out to me is the Tampa Bay defense at #7, especially after giving up 32 points to Danny Dimes and the Saquon-less Giants. Can anyone say whether this ranking corresponds to reality in any way?
#15 by theTDC // Sep 24, 2019 - 11:01pm
Well they've given up 77 total points, so that's 45 over the next two games, for 35.7 per game. That's actually only slightly below league average, eyeballing the stats. So a few bounces here or there could explain the rating, although I also find it somewhat puzzling myself.
#41 by crw78 // Sep 25, 2019 - 1:38pm
In Week 1, TB outgained SF (295 to 256) and had more yards per play (4.5 to 4.3). Both defenses played well in that game. SF only scored 31 points due to two pick 6's.
In Week 2, TB had a YPP of 4.9 vs. 4.8 for Carolina, and completely shut down Carolina's running game (19 carries for 39 yards).
In Week 3, TB defense didn't play as well (6.4 YPP for NYG vs. 6.8 for TB), but they did force two turnovers and had 4 sacks as well I think.
Take away the 14 points opposing defenses have scored and they've give up 63 points, or 21 per game, so the overall numbers seem about right.
#16 by Mountain Time … // Sep 24, 2019 - 11:51pm
The Broncos are clearly rated too low, because they're obviously going to get their shit together and start kicking ass, just as soon as Lock can play. My years of experience watching football on tv is way better than this. Drew Lock is better than Drew Jones or whatever that kid's name is in new york!!!!
#25 by Swilson1472 // Sep 25, 2019 - 2:43am
I only watched the Broncos play the Raiders. Can someone tell me what is wrong with the Broncos pass rush? I understood a little bit the first 2 weeks playing teams that like to get the ball out fast, but Rodgers is infamous in certain analytics circles for holding the ball too long and taking too many sacks.
#49 by Lost Ti-Cats Fan // Sep 25, 2019 - 6:19pm
I didn't see many Bears games last year, but I did catch part of the Bears-Pats game. Fangio had Mack dropping back into coverage most of the snaps I saw that game, and not rushing.
I presumed that was out of the ordinary, given Mack's overall sack totals last year, and that maybe Mack was hurt or something and not able to rush properly. But maybe Fangio was trying to surprise Brady? (Surprise! Take your time and pick us apart.)
If the same thing's happening with Von Miller, is it possible there's a tell to Fangio's scheme, and offenses can push the end rusher into coverage with certain formations or shifts? Seems unlikely, but then it also seems unlike that the Broncos front seven wouldn't have a sack by now.
#51 by TomC // Sep 25, 2019 - 8:10pm
Yeah, my immediate reaction was to blame Fangio and his very conservative schemes too. I know it sounds crazy given how good the Bears D was last year, but I feel like they could have been even scarier with more aggressive pressure. We'll see how that goes with Pagano.
#55 by reddwarf // Sep 26, 2019 - 1:44pm
I think it's more subtle than scheme, and more about usage, lack of a lead and Miller himself. He generated some pressure (though no sacks) against Green Bay, but was invisible the first two games and he still rushed plenty (if not every snap). If you look at his splits going back years (http://www.nfl.com/player/vonmiller/2495202/situationalstats?season=2019) he's always been both a front-runner (generating most sacks when the team has a lead-not unusual for rushers but a situation Denver simply hasn't been in at all this year) and in the 2nd and 4th quarter (2015 being the only major exception).
Usage wise both the Phillips and Joseph regimes made it a point to rotate Miller a lot. He was not near the team lead in snaps (which is why fans would complain he was on the sidelines during a big play so often). This year Fangio is doing a lot less rotation, and both Chubb and Miller have been on the field the vast majority of snaps. Data at Pro Football Reference says Miller's been on the field for between about 75% to the low 80's of snaps the last few years. So far this year he's at nearly 95%. I wonder if he's simply wearing out a bit as the game goes on, and instead of picking up 2nd/4th quarter sacks against an o-line that has played many more snaps than he has, he doesn't have that "freshness" advantage he took advantage of previously.
Finally, as much as I love Von, the simple truth is even I've seen him take plays off when he gets frustrated, and disappear for streaks when the team overall isn't doing well. Both of which are also happening now.
tl;dr: I don't think it's Fangio's scheme so as much as usage, plus the fact Denver hasn't had any leads, plus Miller being a bit frustrated at times, plus simple bad luck.
#22 by RevBackjoy // Sep 25, 2019 - 1:24am
I don't get how the Vikings are so low, even sans opponent adjustments. Superficially, they seem pretty strong: +10.3 points per game differential, +1.3 yards per play differential (6.4 vs 5.1: 7.6 vs 5.5 passing, 5.6 vs 4.4 rushing), +2 turnover margin... what gives? Did DVOA have an allergic reaction to lefse & lutefisk one Christmas as a youngster? Is it holding Kirk Cousins' Pizza Ranch commercial against the whole team (trigger warning: https://www.youtube.com/watch?v=eblELpZHp9A)?
#26 by Swilson1472 // Sep 25, 2019 - 2:50am
I'd assume it might have to do with some reliance on very large plays that get discounted by DVOA and not enough of the intermediate plays that it loves. That and all things about the way Kirk Cousins has played the past 2 weeks and that commercial. DVOA has access to the internet confirmed.
#30 by Will Allen // Sep 25, 2019 - 7:53am
Also, Zimmer is more accomodating than a lot of head coaches and dcoordinators in trading yards for clock when he is up multiple scores. His defenses are good at making the opposing offense use 7 minutes to get a td when the Vikings are up 27-7. Now, FO has found that even garbage time tds have predictive value so that dings the Vikings DVOA somewhat, but I suspect that the Vikings defense is a bit of an outlier in this regard.
#39 by Thomas_beardown // Sep 25, 2019 - 12:03pm
The NFC North seems ridiculously stacked so far this year. The Bears are the worst team and ranked 16th. If that holds, early opponent adjustments won't really know what to do for a while. Especially with how the NFL likes to schedule divisional games early
#54 by Hoodie_Sleeves // Sep 26, 2019 - 12:02pm
The Patriots defense has historically done similar.
I suspect that a not-insignificant part of garbage time being predictive in DVOA is that it is usually pulling extreme performances to the mean.
Or, it's not really that garbage time is predictive, it's that going up 45-0 on an opponent isn't actually much more meaningful than going up 28-0, and the team giving up 14 points on slow drives at the end of the game helps to dampen that.
#56 by Will Allen // Sep 26, 2019 - 1:52pm
Of course I've noticed that about Darth's teams as well, especially since his minions have so much more frequently developed big leads. Zimmer's not a perfect coach but he does share this trait with Darth, he really gets his team to be very efficient in choking out the opponent without much drama, like a python on a rabbit, once his team does have a big lead. Even a very smart offensive coach like Reid has a tough time avoiding the trap of the 8 minute scoring drive.
#23 by mehllageman56 // Sep 25, 2019 - 2:03am
This Jets fan is hoping things get worse. Because they have to get worse to get better, i.e., get rid of the coaching staff. I have some faith in Joe Douglas going forward, and that is it. So hearing they're going to win a bunch of games against garbage teams like Miami does not make me feel better. Sorry, Aaron, but you probably understand since you were a Pats fan before they turned it around.
#27 by Jetspete // Sep 25, 2019 - 3:16am
If you have faith in Douglas you better hope they start winning. Yes it is completely unfair to judge anything about this year on him, but the Jets have 15 months to prove themselves. Whether or not Trump wins, I have to believe Woody's vacation to London will come to an end by late 2020. And when he returns he will fire everyone to put his own stamp back on the team unless victories on the field force him otherwise.
And why do you think losing would make them fire the coach? Bowles got 4 years with no playoff appearances and in both 2016 and 2017 his players openly quit on the field....AND THEY STILL BROUGHT HIM BACK FOR 2018!!! Jets could find a way to go 0-19 this season and theyd still keep Gase.
#32 by mehllageman56 // Sep 25, 2019 - 8:50am
Woody has never been a fire everyone kind of guy. Even when Tannenbaum took over as GM, they kept the previous GM, shifting him to personel director. The only people you can argue he gave a short leash to were Mangini and Idzik, and they both deserved to go sooner rather than later. Joe Douglas got a six year contract; I think he's pretty safe for a while.
The players didn't openly quit on Bowles in 2017; they were supposed to be atrocious and still had a much better offense with Josh McCown at quarterback. 2016 was a mess, but even then the Jets were able to blow out the Bills in the last game, embarrassing Anthony Lynn in his only game coaching that year.
The Browns may have kept Hue Jackson when they went 0-16, but they still fired the general manager. Since Douglas just got hired with a six year contract, Gase would be the one to go if they got rid of anyone. Perhaps you are correct that they would wait until 2020 and the return of Woody to fire anyone, but if Douglas has any reason to worry about his job he would have every reason to throw Gase under the bus, especially since it's plain to see Gase is terrible at his job.
#52 by Sarsgaard // Sep 26, 2019 - 4:08am
Am I correct in assuming that the 6.0 DAVE subtraction you gave last week for Bridgewater will not simply be 6.0 again because now the season average DAVE for New Orleans incorporates one more Brees week, so the Bridgewater penalty will be higher? I don't know the exact math, but maybe it goes to something like 7.0...