DVOA Analysis

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

Week 4 DVOA Ratings

by Aaron Schatz

It's been a topsy-turvy season all over the NFL, but so far everything is coming up aces for Andy Reid and the Kansas City Chiefs. The NFL's only undefeated team is also No. 1 in DVOA through four weeks, with a huge lead over the rest of the league. Kansas City is at 43.3% DVOA, and no other team is over 30%. Pittsburgh is the highest, at 29.8%.

The Chiefs are a strong team, but not historically strong. Now that opponent adjustments have kicked in and boosted the Patriots' (still last place) defensive DVOA a little bit, no units this year qualify as particularly great or terrible historically. Perhaps the only worthy note along those lines is that the No. 1 Buffalo defense has the worst DVOA ever for a defense ranking No. 1 after four weeks. Still, the Bills were 27th in defensive DVOA last year and are enjoying the wonderful position of first place in the AFC East, so I think they'll take that.

The clearest trend through the first month of 2017 is the colossal gap between the bad teams and the rest of the league. With only one team undefeated and every team showing flaws, it may seem like there are fewer good teams in the NFL than usual. But the opposite is true: suddenly in 2017 the NFL has turned into Lake Wobegon. There are currently 20 teams with above-average DVOA ratings and only 12 teams that are below zero. That's how bad Indianapolis, Cleveland, and to a lesser extent Miami have been this year. None of them is historically among the worst September teams, but because there are three of them, that's enough to offset all of the teams sitting with DVOA ratings above zero. In fact, the gap isn't really 20 positive teams and 12 negative teams; it is really about 24 good or average teams and eight horrendous teams.

Dallas, New England, and Oakland are currently 21-23, all teams that were expected to be better this season and probably will be over the rest of the year. No. 24 is the "best winless team," the Los Angeles Chargers at -5.9% DVOA. And after that, the bottom falls out. There's a gap of almost 12 percentage points between the Chargers and the New York Jets, who climb to No. 25 this week. By the time we get to the Giants at No. 27, we're down around -30% DVOA. The teams at the bottom of the league are really, really bad. Of those bottom eight teams, the best offense belongs to the Giants at 22nd and the only defense better than average is Arizona which ranks 12th.

Another big trend this year is that the NFC has been far more predictable than the AFC, at least if you were trying to predict with the forecast in Football Outsiders Almanac 2017. Remember, we had the Rams as an average team, so their hot start is less of a surprise in our metrics. However, nobody expected the Patriots or Raiders to be this mediocre through four weeks. Nobody expected Buffalo to play so well. Nobody would have guessed that two AFC South teams would be in the top ten after a month, but both Houston and Jacksonville are up there because of one huge win: Jacksonville over Baltimore in Week 3, and Houston over Tennessee in Week 4.

And so, because the top AFC teams in DVOA are so different from the top AFC teams in our preseason forecast, the DAVE ratings are dominated by NFC teams. Ten of the top 13 teams in DAVE are NFC teams. Pittsburgh and Kansas City are first and second, and the Patriots' preseason forecast still has them at sixth in DAVE despite being 22nd in DVOA. But that's it. This also has an effect on our playoff odds report. Pittsburgh and Kansas City are way out ahead of the rest of the league in playoff odds because the NFC teams behind them in DAVE have so much competition within their conference. We currently list Pittsburgh with an 18.8 percent chance to win the Super Bowl and Kansas City with an 18.6 percent chance. No other team is above 7.0 percent.

A big part of the playoff odds report simulation is each team's schedules going forward, and this week is the season debut for the second weekly table that includes past and future schedule ratings. In the preseason we projected the Patriots with the easiest schedule in the league, but the rise of Houston and Kansas City means it has actually been the toughest in the league so far. That gets a lot easier from here on out, with the Patriots' remaining opponents ranked 25th (although only five of those are home games).

[ad placeholder 3]

Other teams whose schedules get much easier after September include Washington, Tennessee, Philadelphia, Houston, and -- I told you everything was coming up aces for Andy Reid -- Kansas City.

Teams whose schedules get much tougher after September include Arizona, Seattle, Pittsburgh, and especially Indianapolis. If the AFC South teams are truly better than expected, then Indianapolis has the toughest remaining schedule in the league over the final three months.

But are they really? Or did Houston and Jacksonville just each have one big game early on? Our usual schedule strength ratings are based on DVOA, and thus only based on performance during the season so far. That may not be the most accurate way to measure the schedule. Most of us, for example, don't expect Dallas and New England to continue to be average football teams for the rest of the year. So I also ran schedule strength using DAVE instead of DVOA. And the results... are actually quite similar to schedule strength using DVOA. Here are all four ratings -- past and future, DAVE and DVOA -- for you to consider and discuss in the comments below. Teams are in order by remaining schedule as ranked by DVOA of opponent.

Past and Remaining NFL Schedules for 2017, DVOA vs. DAVE
TEAM DAVE
PAST
RANK DAVE
FUTURE
RANK DVOA
PAST
RANK DVOA
FUTURE
RANK
IND -6.1% 26 2.3% 6 -11.2% 27 8.7% 1
MIA -6.7% 27 3.9% 3 -3.5% 21 6.7% 2
NYG 5.8% 6 4.0% 2 8.4% 6 6.4% 3
ARI -9.5% 30 3.4% 4 -15.3% 29 6.2% 4
DAL -2.7% 22 4.6% 1 -4.8% 23 5.8% 5
CLE -0.4% 16 2.1% 7 -3.8% 22 4.9% 6
NO -0.2% 15 0.9% 10 -6.4% 25 4.1% 7
OAK -1.0% 17 3.2% 5 4.6% 11 3.6% 8
GB 0.5% 14 1.3% 8 -2.6% 19 2.8% 9
CHI 10.9% 1 0.4% 11 13.2% 4 2.5% 10
TB -9.0% 28 1.0% 9 -17.1% 31 2.3% 11
SEA -10.6% 31 0.3% 13 -17.0% 30 1.5% 12
PIT -12.1% 32 0.2% 14 -18.2% 32 1.0% 13
CIN -4.3% 25 -0.8% 18 -5.8% 24 0.2% 14
ATL 1.9% 12 0.3% 12 4.5% 12 0.1% 15
CAR -1.2% 18 -0.9% 23 -0.3% 17 0.0% 16
TEAM DAVE
PAST
RANK DAVE
FUTURE
RANK DVOA
PAST
RANK DVOA
FUTURE
RANK
NYJ -3.6% 24 0.1% 16 -3.4% 20 -0.3% 17
MIN 9.6% 3 0.2% 15 16.6% 3 -0.5% 18
LACH 2.5% 11 -0.4% 17 5.6% 9 -0.6% 19
LARM -9.4% 29 -1.3% 24 -14.9% 28 -1.4% 20
KC 6.8% 5 -0.9% 21 6.4% 8 -1.9% 21
DET -3.4% 23 -0.9% 20 -7.4% 26 -2.4% 22
DEN 1.6% 13 -0.8% 19 1.6% 15 -2.6% 23
SF 4.0% 10 -1.7% 25 4.1% 13 -3.0% 24
NE 8.0% 4 -3.5% 27 20.0% 1 -3.5% 25
PHI 4.7% 8 -0.9% 22 7.8% 7 -4.2% 26
WAS 10.8% 2 -1.9% 26 17.9% 2 -5.4% 27
HOU 4.4% 9 -4.8% 30 4.9% 10 -7.0% 28
BAL -1.4% 19 -4.0% 29 -1.6% 18 -7.4% 29
BUF -1.5% 20 -4.0% 28 2.2% 14 -8.9% 30
JAC -2.5% 21 -7.1% 31 1.6% 16 -11.0% 31
TEN 4.7% 7 -9.1% 32 8.5% 5 -13.4% 32

* * * * *

Once again this season, we have teamed up with EA Sports to bring Football Outsiders-branded player content to Madden 18. This week, we've got TWO sets of content: our weekly content for Madden Mobile and our monthly content for Madden Ultimate Team on consoles.

These players will get special Madden Mobile items branded as "Powerline, powered by Football Outsiders," beginning at 11am Eastern on Friday. The Football Outsiders stars for Week 4 are:

  • DT Gerald McCoy, TB: 3 run TFL, QB hit.
  • OLB Cory James, OAK: 10 solo tackles including two preventing third-down conversions and two TFL.
  • C Ali Marpet, TB: Helped control the powerful New York Giants pass rush.
  • FS William Parks, DEN: Four tackles to prevent conversions on third or fourth downs.
  • LG Rodger Saffold, LARM: Helped lead Titans RB to 52 yards on nine carries running left; no sacks allowed.

And here are the Football Outsiders September players for Madden Ultimate Team on consoles, which will go live at 10:30am Eastern on Wednesday. All of these players are elites (80 overall or better).

  • QB Jared Goff, LARM: Leads all QB in passing DVOA through Week 4, completing 67 percent of passes for 9.2 yards per attempt.
  • RB Dalvin Cook, MIN: Third among RB with 79 rushing DYAR (4.8 yards per carry, 88.5 yards per game, 2 TD).
  • WR Keenan Allen, LACH: Fourth among WR with 105 receiving DYAR (24-of-40, 334 yards, TD).
  • LT Taylor Lewan, TEN: Titans lead league with lowest adjusted sack rate and rank fourth in adjusted line yards on runs around left end.
  • C John Sullivan, LARM: Only 2 blown blocks through Week 3; Rams rank fourth in adjusted sack rate and fourth in adjusted line yards between the tackles.
  • RG A.J. Cann, JAC: Jaguars are second in adjusted sack rate and sixth in adjusted line yards on runs to the right side.
  • LE DeMarcus Lawrence, DAL: Leads NFL with 8.0 sacks; tied for lead with 12 defeats.
  • MLB Wesley Woodyard, TEN: Leads NFL with 18 run stops (plays that stop a run short of baseline for a successful gain); third among defenders with 11 defeats.
  • CB E.J. Gaines, BUF: Fourth in CB success rate, 14th in yards per pass (through Week 3).
  • FS Budda Baker, ARI: Tied for NFL lead with 7 special teams tackles.
  • K Matt Prater, DET: 8-for-9 on field goals including 4-for-5 from 55 or more yards.
  • P Marquette King, OAK: Leads NFL with average gross 52.1 yards, average net 47.0 yards. Also has 9 punts ending inside 20.

* * * * *

All stats pages should now be updated through Week 4, including snap counts, playoff odds, and the FO Premium DVOA database.

* * * * *

[ad placeholder 4]

These are the Football Outsiders team efficiency ratings through four weeks of 2017, 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. Because it is early in the season, opponent adjustments are only at 40 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 50 percent of DAVE for most teams (60 percent for Miami and Tampa Bay).

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>

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 KC 43.3% 1 22.4% 2 4-0 38.3% 1 -5.4% 11 -0.4% 19
2 PIT 29.8% 2 26.6% 1 3-1 12.4% 7 -19.2% 3 -1.7% 24
3 LARM 23.4% 3 12.8% 3 3-1 14.4% 6 -2.8% 13 6.2% 4
4 WAS 23.1% 6 11.9% 7 2-2 7.7% 14 -16.9% 6 -1.5% 22
5 DET 21.7% 4 11.7% 8 3-1 -3.3% 21 -17.1% 5 7.9% 1
6 BUF 18.4% 7 3.2% 16 3-1 -5.2% 23 -21.2% 1 2.4% 11
7 HOU 16.6% 22 3.3% 15 2-2 -1.8% 19 -17.8% 4 0.6% 18
8 JAC 14.8% 8 2.2% 17 2-2 1.8% 17 -16.9% 7 -3.8% 27
9 NO 13.4% 11 2.2% 18 2-2 19.2% 3 6.4% 20 0.6% 16
10 MIN 11.2% 12 4.2% 12 2-2 17.1% 5 5.3% 19 -0.6% 21
11 GB 11.0% 16 12.3% 5 3-1 9.7% 10 0.7% 15 2.0% 12
12 PHI 10.8% 14 7.2% 9 3-1 9.8% 9 3.0% 17 4.0% 8
13 ATL 10.7% 9 7.1% 10 3-1 12.4% 8 7.0% 22 5.4% 5
14 DEN 9.2% 13 1.3% 20 3-1 3.4% 16 -14.4% 8 -8.6% 29
15 SEA 8.4% 19 12.5% 4 2-2 8.6% 13 -1.5% 14 -1.7% 23
16 CAR 6.8% 17 4.0% 13 3-1 -2.5% 20 -6.6% 10 2.8% 10
TEAM TOTAL
DVOA
LAST
WEEK
TOTAL
DAVE
RANK W-L OFFENSE
DVOA
OFF.
RANK
DEFENSE
DVOA
DEF.
RANK
S.T.
DVOA
S.T.
RANK
17 TEN 3.8% 5 1.2% 21 2-2 18.5% 4 18.0% 31 3.3% 9
18 BAL 3.7% 10 3.6% 14 2-2 -18.0% 26 -21.1% 2 0.6% 17
19 CIN 3.3% 24 2.0% 19 1-3 -15.2% 25 -14.4% 9 4.1% 7
20 TB 1.4% 20 -2.3% 23 2-1 9.6% 12 10.2% 25 2.0% 13
21 DAL -0.2% 15 6.3% 11 2-2 4.4% 15 8.8% 24 4.3% 6
22 NE -2.5% 23 12.0% 6 2-2 27.3% 2 30.9% 32 1.0% 14
23 OAK -5.9% 18 0.8% 22 2-2 1.2% 18 13.7% 27 6.6% 3
24 LACH -5.9% 21 -4.0% 24 0-4 9.6% 11 1.3% 16 -14.3% 32
25 NYJ -17.8% 26 -18.2% 27 2-2 -14.7% 24 4.0% 18 0.8% 15
26 ARI -22.2% 25 -13.4% 26 2-2 -19.0% 27 -4.9% 12 -8.2% 28
27 NYG -29.5% 29 -11.5% 25 0-4 -3.9% 22 14.5% 28 -11.1% 31
28 SF -30.5% 30 -22.4% 30 0-4 -30.2% 31 8.1% 23 7.7% 2
29 CHI -33.0% 28 -19.6% 28 1-3 -23.2% 28 6.9% 21 -2.9% 26
30 MIA -41.0% 27 -20.8% 29 1-2 -23.8% 29 14.7% 29 -2.5% 25
31 IND -52.1% 32 -33.6% 31 1-3 -41.0% 32 10.7% 26 -0.4% 19
32 CLE -54.5% 31 -36.4% 32 0-4 -30.1% 30 15.3% 30 -9.1% 30
  • 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).



TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
1 KC 43.3% 4-0 39.4% 4.0 1 6.4% 8 -1.9% 21 4.2% 5
2 PIT 29.8% 3-1 37.3% 3.1 2 -18.2% 32 1.0% 13 5.0% 8
3 LARM 23.4% 3-1 37.3% 2.7 5 -14.9% 28 -1.4% 20 22.2% 24
4 WAS 23.1% 2-2 18.6% 2.5 10 17.9% 2 -5.4% 27 18.9% 21
5 DET 21.7% 3-1 26.3% 2.9 4 -7.4% 26 -2.4% 22 5.4% 10
6 BUF 18.4% 3-1 24.3% 2.6 7 2.2% 14 -8.9% 30 14.2% 19
7 HOU 16.6% 2-2 13.0% 2.4 11 4.9% 10 -7.0% 28 28.8% 28
8 JAC 14.8% 2-2 20.9% 1.9 19 1.6% 16 -11.0% 31 42.8% 30
9 NO 13.4% 2-2 12.2% 2.7 6 -6.4% 25 4.1% 7 28.7% 27
10 MIN 11.2% 2-2 2.7% 2.2 16 16.6% 3 -0.5% 18 7.6% 15
11 GB 11.0% 3-1 14.2% 2.9 3 -2.6% 19 2.8% 9 7.0% 14
12 PHI 10.8% 3-1 13.6% 2.4 12 7.8% 7 -4.2% 26 5.8% 12
13 ATL 10.7% 3-1 11.4% 2.6 8 4.5% 12 0.1% 15 2.3% 1
14 DEN 9.2% 3-1 12.3% 2.4 13 1.6% 15 -2.6% 23 5.1% 9
15 SEA 8.4% 2-2 19.2% 1.9 22 -17.0% 30 1.5% 12 13.4% 17
16 CAR 6.8% 3-1 8.2% 2.4 14 -0.3% 17 0.0% 16 28.2% 26
TEAM TOTAL
DVOA
W-L NON-ADJ
TOT VOA
ESTIM.
WINS
RANK PAST
SCHED
RANK FUTURE
SCHED
RANK VAR. RANK
17 TEN 3.8% 2-2 1.2% 2.5 9 8.5% 5 -13.4% 32 41.2% 29
18 BAL 3.7% 2-2 3.8% 2.0 18 -1.6% 18 -7.4% 29 55.2% 32
19 CIN 3.3% 1-3 1.8% 1.9 21 -5.8% 24 0.2% 14 43.0% 31
20 TB 1.4% 2-1 8.4% 2.3 15 -17.1% 31 2.3% 11 13.6% 18
21 DAL -0.2% 2-2 1.3% 2.0 17 -4.8% 23 5.8% 5 18.4% 20
22 NE -2.5% 2-2 -2.9% 1.7 23 20.0% 1 -3.5% 25 5.7% 11
23 OAK -5.9% 2-2 1.0% 1.9 20 4.6% 11 3.6% 8 26.6% 25
24 LACH -5.9% 0-4 -6.1% 1.3 25 5.6% 9 -0.6% 19 2.8% 2
25 NYJ -17.8% 2-2 -17.6% 1.6 24 -3.4% 20 -0.3% 17 21.0% 23
26 ARI -22.2% 2-2 -14.0% 0.9 26 -15.3% 29 6.2% 4 6.8% 13
27 NYG -29.5% 0-4 -25.0% 0.6 28 8.4% 6 6.4% 3 3.4% 4
28 SF -30.5% 0-4 -29.3% 0.6 29 4.1% 13 -3.0% 24 4.6% 6
29 CHI -33.0% 1-3 -35.4% 0.6 27 13.2% 4 2.5% 10 4.7% 7
30 MIA -41.0% 1-2 -37.5% 0.2 30 -3.5% 21 6.7% 2 3.2% 3
31 IND -52.1% 1-3 -48.3% 0.0 31 -11.2% 27 8.7% 1 20.8% 22
32 CLE -54.5% 0-4 -52.7% 0.0 32 -3.8% 22 4.9% 6 11.9% 16

Comments

86 comments, Last at 10 Oct 2017, 10:38am

1 Re: Week 4 DVOA Ratings

Week 4 BES Rankings went out this morning - http://besreport.com/week-4-bes-rankings-2017/

Aside from the Saints and Skins, DVOA and BES concur on occupants of the top 10, especially with the Chiefs and Steelers No. 1 and No. 2 respectively. Similar to their big lead in DVOA over the rest of the field, the Chiefs lead the BES by a significant margin as well.

Good to see the Jags get love from DVOA. The BES has them at No. 10. It'd be great if they could string together some wins. Bortles has to take that next step though. Also, the BES and DVOA both have the Jags defense ranked 7th.

2 Playoff odds

What, no Jerome Bettis Bowl? Did you know he's from Detroit? (At some point we'll have to start explaining that joke.)

3 Re: Week 4 DVOA Ratings

Really seems like GB landed a capable punter. Vogel has had some shanks but also has nailed some punts which changed field position to help the cause.

19 Re: Week 4 DVOA Ratings

Amazing to see GB threatening the top 10 on ST. The punter has been mostly excellent, and while I expect we'll see a few more shanks out of the rookie, his future seems very very bright. The kickoff coverage has gone from atrocious to decent, and the punt coverage is phenomenal - you can't double both Janis and Josh Jones on the outside, and neither can be blocked by one guy.

To be able to think the GB ST could actually tilt a close playoff game in GB's favor is a thrilling prospect!

23 Re: Week 4 DVOA Ratings

As a quick follow up, Jeff Janis has carved out a niche for himself as great coverage guy, kickoff return man, and receiver called into service for Hail Mary situations where his terrible route running don't matter (anyone can run straight down the field and after a few seconds begin looking for the ball)

Good for him

37 Re: Week 4 DVOA Ratings

I glad someone else recognizes this. I've been arguing the Jarrett Bush comparison on a couple other forums since last year.

During training camp there have been so many Packers fans predicting Janis to get cut because he's never developed into a complete WR, and of course these predictions were wrong. I don't think many fans appreciate the value in a guy that can contribute significantly to every main ST unit, even if it isn't the really visible positions (kicker, punter, or punt returner, not as much kick returner until the weather gets bad and we see some returnable kicks).

Also saw so many comments that someone else can fill in at gunner without missing a beat, not recognizing that A) it's not that easy to find players with Janis' size and speed, and B) it really helps to have TWO good gunners on your team.

Even if he never becomes a great WR, I'd be really happy to see Janis stick around for many year in his current role. And if injuries do strike the WR corps in the future, I'd still rather have him as a deep backup than your typical WR5 or WR6.

40 Re: Week 4 DVOA Ratings

Given his now extended tenure one would think that Janis would be able to modify (if not fix) the issues that keep him from being on the field as a receiver.

And to step into hazardous non-PC talk, given that his issues tie to mostly mental mistakes (poor routes, wrong routes, not being ready for qb changes) has anyone read anything about Janis ability to learn and retain direction?

Because the speed, the strength, the length, the jumping, the ability to take contact, the ability compete for the ball are all there. By all accounts Rodgers simply doesn't trust him because he's rarely where is supposed to be when is supposed to be somewhere.

And I am like---DUDE, you can FIX that!

Or maybe he can't. Which is kind of baffling

5 Re: Week 4 DVOA Ratings

Wow. The gap between the #31 and #32 defenses is amazing. But entirely earned by the #32 team.

10 Re: Week 4 DVOA Ratings

I'm shocked any part of a Belichick coached team has been that bad that long. Even when he was turning wide receivers into defensive backs things were trending slightly up. If this goes on too long are we going to start to hear comments in the media about his effectiveness starting to fall off?

11 Re: Week 4 DVOA Ratings

I won't be fooled again into thinking the Pats dynasty is over after last year. If this defense progresses towards average(say #20 in DVOA) they are back to being SB contenders.

29 Re: Week 4 DVOA Ratings

This is probably the worst we've seen them, but there have been enough occurrences of bad early season patriots defenses turning into decent late season defenses - so I wouldn't count on things staying this way.

Last year they were in the bottom half of DVOA at this point, with the 26th ranked defense (+12%) and 13th offence (+3.5%). The defense is a bunch worse, but the offense is a bunch better.

Last year, week 4 total DVOA: -5.6%
This year, week 4 total DVOA: -2.5%

(And Yes, I realize Brady missed the first 4, and there was an expected increase in the offense - but I'm not really convinced that 4 games of +30% defense is much more predictive than 4 games of +12% defense - I think everyone can agree that they look terrible now, and they'll probably look better later - question is how much)

57 Re: Week 4 DVOA Ratings

Yeah, I won't believe they are done until they miss the playoffs two years in a row and/or Brady and/or Belichick retire.

Although just "not winning the division" would be an accomplishment that hasn't happened since 2008. And only twice since 2001.

56 Re: Week 4 DVOA Ratings

Wow, I just looked. The Patriots are allowing one more yard per play (7.2) than the next-worst teams (New Orleans and Philadelphia at 6.2).

The Patriots also rank pretty highly in forcing turnovers. They force turnovers on 11% of drives and allow scores on 49% of drives. So they either allow a score or force a turnover 60% of the time. Tennessee is next at 56%. Cincinnati is lowest at 34%.

6 Re: Week 4 DVOA Ratings

It's real, and it's spectacular.

https://www.amazon.com/Why-Daddy-Sad-Sunday-Disappointing/dp/061599234X

7 Re: Week 4 DVOA Ratings

KC's #1 run offense at +41.5% seems pretty impressive. Would that rank near the top historically?

8 Re: Week 4 DVOA Ratings

Hats off to FO for foreseeing the emergence of the Rams. I would have tattooed myself with your brand if you'd predicted that the #1 passing DVOA would belong to Jared Goff as well.

A note about the Packers: the announcers were banging on about the fact that they now lead the all-time series against the Bears for the first time since the 1930s, but what they didn't mention was how frequently, in recent times, the Packers have beaten not only Chicago but also the spread. I wanted to see if there were any other rivalries where one team has done as well against the spread, and there's at least one way of presenting the numbers that produces the answer No.

This is a list of head-to-heads since 1992 ("the Favre-Rodgers Era"), including playoffs, where one team is in the plus column for a net eight games or more against the spread:

Net games ATS – Team vs Oppt – Head-to-head record ATS – Record ATS
+16 – GB vs Chi – 33-17-2 – .654
+15 – NYJ vs Mia – 32-17-2 – .647
+15 – Pit vs Cin – 33-18-1 – .644
+14 – SF vs Rams – 32-18-1 – .637
+13 – Car vs SF – 17-4-0 – .810
+12 – NE vs Ind – 24-12-1 – .662
+11 – Car vs TB – 22-11-0 – .667
+11 – NE vs Buf – 29-18-3 – .610
+11 – Chi vs Min – 31-20-0 – .608
+10 – KC vs Sea – 17-7-0 – .708
+10 – Sea vs SF – 21-11-1 – .652
+10 – Phi vs NYG – 32-22-0 – .593
+9 – Det vs Chi – 29-20-1 – .590
+9 – GB vs Det – 30-21-1 – .587
+8 – Mia vs Den – 9-1-0 – .900
+8 – Oak vs Buf – 10-2-0 – .833
+8 – Was vs Chi – 10-2-0 – .833
+8 – Ind vs Den – 12-4-0 – .750
+8 – Atl vs Det – 12-4-0 – .750
+8 – SD vs Ind – 13-5-0 – .722
+8 – NE vs Pit – 13-5-1 – .711
+8 – NYG vs Was – 28-20-2 – .580
+8 – Pit vs Bal – 31-23-2 – .571

I would be curious to know the Packers' average DVOA against Chicago, and whether it is the highest of any divisional rivalry. (I suspect the Randy Wright / Anthony Dilweg / Don Majkowski era would drag it down considerably, Replay Game notwithstanding.)

51 Re: Week 4 DVOA Ratings

It was a pretty obvious prediction I thought, considering the top-notch coaching staff they hired on both sides of the ball. Weren't they the consensus #1 reader pick to improve too (like Oakland the year before)

9 Re: Week 4 DVOA Ratings

Another consideration on Houston and Jacksonville: both of them produced miserable performances and were blown out the game after their city was struck by a hurricane. This is probably not a coincidence. In Houston's case, it is also probably not a coincidence that their rookie quarterback performed better in his third and fourth games than his first or second. It's possible that both teams are actually likely to be better going forward than their DVOA to date.

15 Re: Week 4 DVOA Ratings

It's like a poor man's version of the late 1980's NFC East! (The Parcells Giants would always beat the Gibbs Redskins, who would always beat the Ryan Eagles, who would always beat the Giants)

30 Re: Week 4 DVOA Ratings

Or like the early 21st-century AFC rock-paper-scissors: Patriots always beat Colts, Colts always beat Broncos, Broncos always beat Patriots.

41 Re: Week 4 DVOA Ratings

That was more of a myth than anything. There never was a real period where that held true unless we cut across years (Shanahan's Broncos beating NE in 2002, NE beating Colts in 2003).

In reality, it was more Pats beat all of them (5-0 vs IND/DEN in 2003-04), or lose to all of them (0-5 in 2005-06).

Now, if we go deeper than wins/losses maybe, as Shanahan's team always played NE close, for instance in '03 NE needed that intentional safety madness to beat Denver with Danny Kanell as the starting QB, but thrashed IND in the playoffs, but it is a bit thin.

What definitely held true was Indy beating Denver all the time (amazingly even after Manning switch sides).

60 Re: Week 4 DVOA Ratings

I think it's more of a playoff thing.

From 2001-2016, in the playoffs:
New England 4-1 vs Indianapolis
Indianapolis 3-0 vs Denver
Denver 3-1 vs New England

64 Re: Week 4 DVOA Ratings

I mean, this rock-paper-scissors thing was more Indy-era Manning, no, which at that point was less stark (given two of the Denver wins over NE were with Manning himself).

I guess Shanny had more success than Manning did pre-2005 against BB's teams.

79 Re: Week 4 DVOA Ratings

As an active Denver fan during the era, myth or not, the 3 way rivalry felt very real, and it highlighted and informed our opinion about what was right and wrong with our team, that we (the Broncos) were generally very good, but were particularly susceptible to defensive lapses against long pass plays on 3rd-and-long which the Colts were better at manufacturing than the Pats were, and with the Pats its was about not losing the field position game, by having just enough non-predictive plays to beat them.

This influenced my feelings (particularly at the time) about DVOA. Where I felt that some of the "random" (non-predictive) events were more significant about the outcome of any given game than what DVOA measured. I felt the Pats were best at the consistent part of the game thus avoiding 3rd-and-long and getting them in consistent 3rd-and-long situations (by aggressive defense) eventually defeated them; the Colts were best at overcoming 3-and-long, but not as good at sustaining drives otherwise; and the Broncos were in the middle. This made it exactly rock-papers-scissors as each advantage/weakness pair generated a similar (intransitive) ordering.

Note I even felt that this carried over to the Baltimore/Denver playoff game the year that Baltimore won the SB. A lapse in 3rd-and-long defense allowed the game tying score and not being able to manufacture the equivalent 3rd-and-long play (PM's final interception) sealed the loss. Slightly different teams, but similar dynamics.

There are people studying such phenomena (e.g. repeated prisoners dilemma) where different strategies have intransitive orderings and one strategy does not consistently beat all others. However, in folk-wisdom I see this mainly reflected in beliefs like I had above, which are roughly equivalent to "informed" superstition. We have some facts and a model that supports our belief, but due to small sample sizes and other effects it is hard to prove them, and when they fail to play out, we invent a slightly more complicated rationale to explain away the anomaly.

59 Re: Week 4 DVOA Ratings

For Jacksonville you can also add that they lost at the Jets after playing in London. The Ravens stunk after playing in London.

The Dolphins stunk after getting hit by a hurricane and then traveling from California to Florida to New York to Florida to London. Will they put up one more stinker this week before finally being able to get in a routine?

83 Re: Week 4 DVOA Ratings

Through week five, the Jags will have played five games in five stadiums in five cities, in addition to the only home game coming six days after a major hurricane. Can't imagine that's ever happened before.

84 Re: Week 4 DVOA Ratings

The 2005 Saints come to mind, of course. Their two home games in the first five were at Giants Stadium and the Alamodome. (Week 6 was also in San Antonio, so their streak ended.)

And, weather aside, I wouldn't be surprised if some team's schedule worked out to five different cities for the first five games. In fact, with the Packers playing some games in Milwaukee for so many years, they'd be a prime candidate.

12 Variance for unit DVOA

Out of curiosity, would it be possible to get the variance of each unit's (off/def/st) DVOA? This early in the season I would expect the variance to be fairly high I'd be interested in seeing how close teams ranks once you take into the account some kind of error measurement.

14 Re: Variance for unit DVOA

I don't think they publish variance numbers for ST, but the variances for offense and defense are listed on the team offense and team defense pages, respectively.

22 Re: Variance for unit DVOA

Thanks, I didn't realize that. I notice that teams like Baltimore, Washington and Jacksonville have high variance on defense so it will be interesting to see if they move lower in the rankings as the weeks go on as games that are outliers have less weight.

16 Re: Week 4 DVOA Ratings

If recent history repeats itself, I expect the Packers' offensive DVOA to crawl up and defensive DVOA to trend down as the season progresses.

17 Re: Week 4 DVOA Ratings

I don't think there's any inherent reason to suspect the Packers' defensive DVOA to worsen. If anything, with two rookies not being dumpster fires and therefore expected to improve, not to mention the return of Mike Daniels, I'd expect it to get better as the season progresses.

31 Re: Week 4 DVOA Ratings

As I've argued on other posts, Dom Capers is not the problem and never has been. The Packers lack top flight talent on defense and have for a long time. I don't expect them to be world-beaters, but I do expect them to stay around league average.

33 Re: Week 4 DVOA Ratings

I think Packer fans have reason to raise questions given the curb stompings in multiple playoff games from various opponents. Not saying that Capers IS the issue. But it's natural that folks point at him when those outcomes keep happening

34 Re: Week 4 DVOA Ratings

It certainly is a natural reaction given some of the performances, I just don't think it is an explanation that actually holds water given the facts. Every defensive coordinator has bad games, and maybe you could lay the 49ers playoff losses from years ago at the feet of Capers as a result of a bad coaching game on his part. But last years' performances were the result of average, bad, and/or injured players at all levels of the defense. There's only so much you can do when your players can't rush the passer, stop the run, or cover against quality competition.

44 Re: Week 4 DVOA Ratings

In those two games where Green Bay held Kaepernick to 450 yards per game, they had Matthews, Woodson, Shields...

It wasn't a defense made up of used gum and misdirection. These were the same guys who won the title in 2010.

53 Re: Week 4 DVOA Ratings

See my reply above. If you want to fire a guy for a few bad games against a bad matchup, be my guest, but it's exactly the kind of reactionary decision that leads to bad teams and coaching carousels. I'll live with an average to good DC who can implement a consistent scheme through the years rather than replacing him and starting anew with someone who, chances are, won't be any better.

20 Re: Week 4 DVOA Ratings

Find it odd that Ali Marpet is listed as a star; he was moved to center from guard this year, and he's had a number of problems snapping the ball over the last two weeks, to the point where Winston has picked a couple balls up off the ground each game due to how bad the snap was in shotgun. Marpet was a really, really good guard, and he still is snapping the ball like a really good guard.