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

Week 19 DVOA Ratings

Celebrating Chiefs
Photo: USA Today Sports Images

Once again, it is time for postseason DVOA ratings. As always, the following rules apply:

  • All 32 teams are ranked, whether they made the playoffs or not.
  • Teams are ranked in order of weighted DVOA, not total season DVOA. Since weighted DVOA is meant to lower the strength of older games, these ratings do not include Weeks 1-5, and Weeks 6-11 are somewhat discounted.
  • Only weighted DVOA is listed for offense, defense, and special teams. Total DVOA is also listed, but adding eight games to a 256-game sample doesn't change much.
  • Teams are treated as having a bye week in any week where they did not play. Since most teams haven't played in two weeks, that means some of the ratings for non-playoff teams can start getting a little unreliable. Really, this is only to be used for playoff teams, the other teams are just there for ranking comparison purposes.
  • DVOA, as always, takes a long-term view of an NFL team's performance. That means that the games of the last two weeks are just two games among many, so teams may be listed below other teams that they have beaten in the playoffs.

Three of the final four teams are grouped near the top of the current weighted DVOA ratings, ranked from third to fifth. That includes Tennessee, which is only eighth for the entire season (including playoffs) but moves up to fourth in weighted DVOA in part because only one Marcus Mariota game is still included in the formula. In fact, at this point Tennessee is slightly higher than Kansas City in both offense and defense. That's slightly artificial because of Houston playing backups against Tennessee in Week 17, but overall the weighted DVOA formula believes that Kansas City and Tennessee are roughly equivalent at this point. The difference between the two is that Kansas City has superior special teams and the home-field advantage in the AFC Championship Game.

The surprising team in DVOA is Green Bay. For the full season (with or without the playoffs), Green Bay is down at No. 10 in our ratings. In the weighted ratings, they're down to No. 15 this week, actually dropping three spots despite the victory over Seattle. The reason for the drop has to do with the nuances of the weighted DVOA formula. Essentially, their big Week 7 win over Oakland dropped significantly in weight this week and they weren't good enough in their 28-23 win over Seattle to make up for that drop. The Seattle-Green Bay game gives us another one of those games where the losing team ended up with the higher single-game DVOA, although the Packers were still positive for the game at +10%. Seattle outgained Green Bay 6.3 yards per play to 5.8 yards per play and there were no turnovers in the game. (The difference is smaller, 6.4 to 6.2, if we remove spikes and kneels, but Seattle still has the advantage.)

You may have seen a lot of mentions of Green Bay being the worst 13-3 team of all-time, which they were by Pythagorean projection based on points scored and allowed. They were the second-worst 13-3 team ever according to DVOA, ahead of only the 1999 Indianapolis Colts. Yet you might still be wondering how the Packers are as low as 15th in weighted DVOA, or for that matter tenth in total DVOA. A look at their week-to-week graph might help explain what's going on here.

Packers weekly

What you have with Green Bay is a team that combined small victories with some huge defeats. Including this week, Green Bay had positive DVOA in 12 of 17 games. But only three of those games had DVOA over 30%. By comparison, including the playoffs, Tennessee has six such games while Kansas City and San Francisco each have nine.

On the flip side, Green Bay had three losses with single-game DVOA worse than -30%. Tennessee also had three such losses, but those all came with Marcus Mariota as quarterback early in the season. Neither Kansas City nor San Francisco had a game below -30% all season.

The Week 12 loss to San Francisco, 37-8, really drags down Green Bay's rating for the season. The 49ers outgained the Packers on average 7.5 yards to 2.8 yards in that game. Aaron Rodgers had less than two net yards per pass. Some fans might argue that we need to dull the strength of these kind of outlier games, but on the other hand this terrible performance happened to be against the team the Packers have to play again next Sunday. Not that the 49ers are necessarily beat the Packers by the same score again -- just ask the 2010 Patriots and Jets about that sort of thing -- but there probably is some useful information in the results of that game.

We should also note that for the season as a whole, the Packers were outgained by opponents 5.7 yards per play to 5.4 yards per play, although they did have a +12 turnover margin.

* * * * *

To save people some time, we remind everyone to put their angry troll hatred into the official zlionsfan angry troll hatred Mad Libs form:

<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>

If you are new to our website, you can read the explanation of how DVOA is figured here. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

There are no adjustments here for sitting starters in Week 17, although we do adjust the ratings that we use in the playoff odds report.

You will find DVOA matchup pages for the conference championships on the FO Premium page. Snap counts will be updated later this afternoon with information from the divisional round.

Teams in yellow are still alive in the playoffs. Teams in gray lost this past weekend.

RK TEAM WEI.
DVOA
LAST
WEEK
W-L WEI OFF
DVOA
OFF.
RK
WEI DEF
DVOA
DEF.
RK
WEI S.T.
DVOA
S.T.
RK
TOTAL
DVOA
RK
1 BAL 46.0% 1 14-3 27.7% 4 -18.1% 3 0.2% 20 40.5% 1
2 NO 41.0% 2 13-4 29.8% 2 -5.4% 11 5.8% 3 30.0% 3
3 KC 40.0% 3 13-4 29.1% 3 -4.7% 14 6.2% 1 28.9% 4
4 TEN 33.8% 4 11-7 30.1% 1 -4.9% 12 -1.2% 26 13.9% 8
5 SF 30.9% 5 14-3 10.0% 7 -17.5% 4 3.4% 9 30.5% 2
6 NE 12.6% 6 12-5 -0.8% 14 -10.4% 7 3.0% 11 27.9% 5
7 SEA 11.4% 9 12-6 13.1% 6 3.1% 20 1.4% 16 13.0% 9
8 MIN 11.2% 7 11-7 -3.8% 19 -13.0% 5 1.9% 13 14.7% 7
9 DAL 10.1% 8 8-8 19.9% 5 4.5% 22 -5.3% 30 17.1% 6
10 TB 9.6% 10 7-9 -7.2% 22 -20.9% 2 -4.1% 29 1.5% 14
11 BUF 7.0% 11 10-7 -6.6% 21 -12.3% 6 1.4% 15 1.9% 13
12 ATL 4.9% 16 7-9 3.5% 10 -1.6% 16 -0.2% 21 -5.0% 17
13 PHI 4.1% 15 9-8 -2.1% 18 -4.9% 13 1.3% 17 5.1% 12
14 LAR 4.0% 13 9-7 0.2% 13 -6.5% 9 -2.7% 27 5.5% 11
15 GB 3.3% 12 14-3 2.4% 11 -0.2% 17 0.8% 18 6.9% 10
16 ARI 1.9% 14 5-10-1 6.7% 8 3.9% 21 -0.8% 24 -5.8% 19
17 CHI -2.3% 17 8-8 -8.7% 24 -7.1% 8 -0.7% 23 -2.1% 15
18 LAC -4.5% 19 5-11 5.3% 9 1.1% 18 -8.8% 32 -6.7% 20
19 NYJ -5.0% 20 7-9 -16.6% 28 -6.3% 10 5.2% 5 -15.8% 26
20 IND -7.1% 18 7-9 -8.4% 23 2.3% 19 3.6% 7 -5.0% 16
21 DEN -8.1% 22 7-9 -11.6% 26 -1.7% 15 1.9% 14 -9.0% 22
22 PIT -9.2% 21 8-8 -35.5% 32 -23.1% 1 3.2% 10 -5.5% 18
23 HOU -12.6% 23 11-7 -1.1% 16 16.5% 29 5.0% 6 -8.2% 21
24 NYG -14.4% 26 4-12 -5.6% 20 11.4% 25 2.6% 12 -17.5% 27
25 CLE -14.9% 24 6-10 -1.0% 15 13.4% 26 -0.4% 22 -10.1% 23
26 CIN -15.3% 27 2-14 -11.9% 27 9.5% 24 6.1% 2 -25.2% 29
27 OAK -17.2% 25 7-9 1.1% 12 15.0% 27 -3.3% 28 -11.0% 24
28 WAS -20.8% 30 3-13 -16.8% 29 9.4% 23 5.4% 4 -25.7% 30
29 MIA -21.4% 28 5-11 -1.9% 17 18.5% 31 -0.9% 25 -36.7% 32
30 DET -25.0% 29 3-12-1 -9.5% 25 15.8% 28 0.3% 19 -12.1% 25
31 JAX -34.7% 31 6-10 -17.5% 30 20.7% 32 3.5% 8 -18.0% 28
32 CAR -46.7% 32 5-11 -21.7% 31 17.5% 30 -7.5% 31 -26.9% 31

 

Here are the single-game DVOA ratings for the divisional round. As you might expect, we had three blowouts and one close game; as noted above, Seattle ended up with a slightly higher rating despite losing the game. In case you're curious about the Houston and Kansas City special teams ratings: one of the quirks of the special teams system is that it makes blocked punts a negative for the team that had the punt blocked but not a positive for the team that does the blocking. I've noted this a few times in the past, but given that blocked punts aren't really predictive for either side, they probably should be removed from DVOA entirely or used in a separate rating that measures value in-game but not predictive value for future performance. This is one of those changes I keep wanting to make but never seem to have time for. The negative value of a blocked punt getting returned for a touchdown plus the muffed punt by Tyreek Hill outweighs the positive value of forcing the fumble on the DeAndre Carter kickoff return.

 

DVOA (with opponent adjustments)
TEAM TOT OFF DEF ST
MIN -33% -42% -5% 3%
SF 104% 19% -84% 2%
TEN 69% 27% -41% 1%
BAL -26% -22% 11% 8%
HOU -32% 21% 54% 1%
KC 68% 83% -1% -15%
SEA 21% 33% 12% 1%
GB 10% 22% 11% -1%
VOA (no opponent adjustments)
TEAM TOT OFF DEF ST
MIN -70% -72% 1% 3%
SF 79% 5% -72% 2%
TEN 31% 19% -11% 1%
BAL -28% -19% 17% 8%
HOU -56% 13% 70% 1%
KC 58% 78% 4% -15%
SEA 14% 32% 19% 1%
GB 3% 29% 25% -1%

Late addendum: During the regular season, I posted weekly updates to a table showing how Baltimore was progressing each week compared to the best teams in DVOA history. Now that the Ravens' season is over, we might as well update that table one last time with a list of the top teams in DVOA history including the postseason. Baltimore comes out 12th in DVOA history once the playoffs are included. 1987 San Francisco 49ers listing does not include strike games.

BEST TOTAL DVOA
INCLUDING PLAYOFFS, 1985-2019
Year Team W-L DVOA Result
1991 WAS 17-2 61.7% Won Super Bowl
1985 CHI 18-1 59.6% Won Super Bowl
2007 NE 18-1 50.3% Lost Super Bowl
1989 SF 17-2 48.1% Won Super Bowl
1996 GB 16-3 45.8% Won Super Bowl
1987 SF* 10-3 44.1% Lost Divisional
2013 SEA 16-3 43.2% Won Super Bowl
1992 DAL 16-3 42.4% Won Super Bowl
1998 DEN 17-2 41.9% Won Super Bowl
2002 TB 15-4 41.0% Won Super Bowl
2010 NE 14-3 40.6% Lost Divisional
2019 BAL 14-3 40.5% Lost Divisional

Comments

78 comments, Last at 16 Jan 2020, 10:45am

1 I didn't look closely last…

I didn't look closely last week, but the Patriots weighted defensive DVOA really fell to -10.4%, 7th in the league? Dang.

6 This Titans team reminds me…

This Titans team reminds me of the 2007 Giants. A nice team that you didn't expect to get out of the first round and if they did, we merely serve as division round slaughter. 

Remarkably that 2007 Giants team Slade 2 very formidable opponents, 1 thanos level opponent, and the Tampa Bay buccaneers en route to the super bowl. If the Titans happen to somehow follow suit, basically beating four formidable opponents, I will have to sit and really think about which run seemed more improbable.

 

58 It's been stated by other…

It's been stated by other analysts that while the Packers have been average in many other places, they are above-average in:

-red-zone offense

-red-zone defense (bend *very* much but don't break)

-Turnover margin.

 

Does DVOA account for the red-zone success?

 

 

61 A very simple answer is "yes…

A very simple answer is "yes".  DVOA accounts for all plays, regardless of where they occur on the field.  Are you implying that DVOA might throw away all the red zone plays?

A more complicated answer is that I actually vaguel recall reading that DVOA puts slightly more weight on red zone plays.  I'll see if I can dig up a source for that.

And turnovers are one thing that DVOA is huge on, so the Packers' turnover differential is something that DVOA would love.  Look at the 2010 Patriots (518 points scored, 10 turnovers, 37.4% offensive VOA) vs. the 1999 Rams (526 points scored, 31 turnovers, 23.2% offensive VOA).  Similar points scored, but the team with far fewer turnovers has a much higher VOA.  (Note that I used VOA to remove the strength of schedule component, in order to compare unadjusted numbers to unadjusted numbers.)

7 I don't know if it makes…

I don't know if it makes sense to compare them to the 2010 Jets. That team was 6th in DVOA and had a record (11-5) about commensurate with that rating. Green Bay has an excellent record but a much worse DVOA then you would expect for going 13-3. Maybe I'm missing the way in which they are similar.

4 Maybe it would helpful to…

Maybe it would helpful to include variance in the tables? Based on the regular season results both BAL and TEN had pretty high variance; that makes the outcome of this weekend a little bit less surprising and would also give us some insight into the upcoming match ups.

5 GB this year

I wonder if any team in DVOA history is even comparable to GB this year. By my count they had had Double Digit 2nd Half Leads in 9 of their 14 wins this year, yet almost all their games end up quite close. They regularly take the lead early in games and then end up barely holding on. They might be the worst team ever at extending 2nd half leads without actually blowing the game. Truly a fascinating team.

10 The 'why?' has been bugging Packer fans all season

In reply to by Q

Because this fan base has seen some REALLY good teams, right?  And not just the good teams like 1996 and 2010/2011 but those teams that improved so at season's end they were legitimately good teams like 2014 and 2003. 

 

Sensible Packer fans grasp that this 2019 Packers squad is not of that quality. But what makes GB competitive is that there are really strong elements of the team.  The offensive line is very good all around. The team has a legit number 1 running back and number 1 wide receiver. The front seven on defense applies good pass pressure.  The field goal kicker is rock solid. The qb does not turn the ball over. The head coach does not do things that get in the way of winning (poor clock management, foolish challenges, etc)

But if GB had not been super healthy this season the cracks would have formed quickly.  The drop off from the starters especially on the offensive and defensive line is severe. The pressures on D are highly concentrated.  The pressures allowed on offense spike when a starter misses even a minimal amount of time. The second running back is not close to the quality of the starter.  The gap between the number one receiver and the next guy is HUGE. 

 

When more than one of the above is not performing WELL in a game the Packers struggle. A LOT. And if the opposition creates matchup issues then the score can rupture in a truly ugly direction very quickly. Against the Chargers the Packers offensive line did not get off the bus and the d-line could not apply any real consistent pressure to Rivers who had his last great day. Against SF Bulaga went out early, Light was a turnstile and the cascade effect across the line was dramatic. But even before that the qb who doesn't turn it over turned it over.  The team's inability to cover tight ends was exposed. 

 

Matt has been leveraging the team's positives while spackling the negatives. There just is not much margin of error

 

13 Like you mention the Pack…

Like you mention the Pack have enjoyed excellent health, and that's been a big, big factor for them winning as many games as they have. But one thing that I saw someone else point out - might have been in audibles, might have been you! - Gutekunst didn't stop working at the margins of the roster during the season, and Veldheer and Ervin have turned out to be very important contributors. Very solid and timely work to get that additional depth and special teams help. In the back half of Ted Thompson's tenure he seemed too content to let guys who are barely playable like Alex Light or Jason Spriggs remain as the first line of depth. Those additions have gone a long way this past month and a half.

17 That is a great explanation…

That is a great explanation that I have not been able to fully put together. 

I'm not sure Matt is a better coach than Mike M (not going back to the other really good Mike H coach) overall, but as I mentioned further down, and you put so well, he does seem to be a better game day coach than Mike was. McCarthy really was good at prep work though. I'll take the game day improvements as it's less frustrating, also I think he's proved his prep phase is good enough and could end up being better. It's his first year, and he's delivered great results from a fan perspective (because the fan in me cares more about the W than how we got there, the analyst in me is still evaluating). I feel tackling has improved as well and that is worth a lot to me.

It's the flashes that you see from every unit that raise your expectations and then drive you nuts when it's a different performance next week. It's more noticeable, like you said, because we've been spoiled with seeing what that talent, backed with solid depth, can do. We don't have the depth so consistency is lacking. That is fortunately something that can be remedied fairly quickly with a good draft, and the flashes are coming from players with several years left.

I've enjoyed the team a lot because of the line play on both sides. Seeing good pressure without blitzing has been rare this last decade. The offensive line has often been above average, but it's downright good this year. I think Rodgers has slipped a little physically and he only really has Adams, which doesn't help, but as I mentioned in the preview thread his game could be greatly effected by small changes in physical ability. For the WR Lazard could be a good #3, but he really isn't a #2, and neither is MVS, Allison, or Kumerow. They don't have a tight end weapon, though the corpse of Jimmy Graham will re-animate from time to time, and that can be huge, but it's not consistent. 

I still don't know what to think of King at corner, but I'm close to settling on talented but injured. That doesn't lead to consistent play. Alexander is good CB who is playing like it's his 2nd season and his size can be exploited. 

I like being a spoiled fan though. The other team I get to see a fair bit is KC, since I've lived here since 2013. That means it's been the Andy Reid Chiefs. That has always been fun because they've always had at least one super fun player to watch, if not more. I also got to get passed by Jamaal Charles on one of the running trails once too, which is fun as well. LOL

18 I too have really enjoyed…

I too have really enjoyed watching this GB team even though it is flawed. Completely agree that the depth is the issue here, especially on offense. The top 3 skill players (Rodgers, Jones, and Adams) are one of the best in the league but the depth beyond that is highly suspect, especially in the roles they’ve been asked to contribute in.

The end of the Ted Thompson years left a bare cupboard on both sides of the ball they’re still recovering from, but I have optimism that it will continue to be filled up this offseason. I honestly think the D is a good linebacker away from being a very good unit. They have a really hard time when teams are in 11 or 12 personnel because they don’t have enough dual-role players in the back 7 yet. Martinez would be fine if he was the only issue, but they don’t have a linebacker that is good in coverage at all. I think that is a top priority for the offseason, along with finding a #2 WR.

I really hope the Packers take a run at AJ Green or Robby Anderson in the offseason. Both have been exemplary deep threats. Otherwise, I think a top pick should go to a top-tier WR prospect.

11 It has been a weird year for…

In reply to by Q

It has been a weird year for sure. I've enjoyed seeing the offense hold onto the ball for 4+ minutes at the end of games several times to seal the win. That was so rare with McCarthy calling plays. I'd rather not have games get that close at the end of course, but I've seen the offense and the defense both make plays to seal wins this year. 

 

I wonder if they manage to make me stupidly happy and win the SB in the same fashion they have been winning nearly all their games this year if that would make them the worst DVOA 3 loss super bowl champion. They already have the best 3 loss team based on the post chart. Could be kind of fun. I really don't expect them to even get the SB, let alone win it, but I am a fan so I have to hope.

33 As I recall the "take an…

In reply to by Q

As I recall the "take an early lead and hang on to make the game look close" was what the pre-2007 Patriots seemed to do. It kept them under the radar (relatively) with ordinary commentators thinking their success was more down to luck and wouldn't be repeatable longterm.

But that was a long time ago and I had less access to follow their games then. Would be interested to hear what someone like RickD thinks to that.

77 hmmm

Well, there were the 2003-2005 Pats and the Pats of the other seasons.  

Before Moss and Welker joined the team, the Pats' strength was their defense.  (Well, except for 2002, when they had serious problems in their run defense.)

The model you mention certainly applies to Super Bowl XXXVI.  And it really shouldn't have worked.  The Rams were clearly the better team, but they got caught napping, the Pats got a few turnovers, and then they were able to do just enough to win at the end.

The team that went back to the Super Bowl 2 years later was much, much better.  

8 Regarding Pittsburgh

Is this the lowest overall DVOA for a team with the league's best defense?

9 1991 Red Clouds and 1985…

1991 Red Clouds and 1985 Bears are so close I wonder if a future update to the DVOA formula could flip them. Of course, it could just as easily extend the Clouds' lead.

19 wonder that too.  with '85…

wonder that too.  with '85 in the books, we'll have dvoa for all of Jerry Rice's career;  '84 will bring another 'all time greatest' contender in the 18-1 49ers (one three-point loss to Pittsburgh), plus Marino's legendary season, and the '83 will give us the full careers for that epic quarterback class (as well as another all-time great offense in the '83 Red Clouds)

68 I’m also interested exactly…

I’m also interested exactly how much the 2007 Patriots Super Bowl defeat cost them in the DVOA race (perhaps this has been covered previously?). IIRC they were right alongside the Clouds and Bears throughout the regular season, but of course those two teams slaughtered their Super Bowl opposition. 

76 ...and throughout the…

...and throughout the playoffs;  closest game either the Bears or the Clouds had was the Clouds' Super Bowl (13 points), and it wasn't really that close;  Pats were pushed in both their playoff games before the SB

15 Again, I am curious to see…

Again, I am curious to see DVOA on the '84 Niners, a somewhat overlooked great team, especially on defense.

20 Will Allen's said it before…

Will Allen's said it before better than I could, but 'tell me who's healthy in January and you'll have a good idea who'll win the tourney'. Still, that graphic during the broacast that teams resting their starters in week 17 are now only 9-6 since 1990 seems really bad for teams good enough to be able to rest starters. Still like the Ravens' medium-term prospects, maybe it'll work out for them like it did this season for the Chiefs, where the year they get the breaks in the postseason doesn't coincide with the year they get the breaks in the regular season (seemed to happen with quite a few teams over the last 20 years that way). I hope, but also trust, that Harbaugh won't be dumb enough to change his aggressiveness just because every break went against him in one high-profile game. This season seems to settle the question of whether Vrabel's a good coach and this Titans team is going to be a tough out

69 The flip side of that stat…

The flip side of that stat is that all playoff teams that didn't rest their starters in Week 17 are 3 games below .500 on aggregate.

Intuitively we assume teams that get to rest their starters are very good teams, but we also have scenarios where teams know they're locked into the 3 or 4 seed, or into the 5 or 6 seed.  So really, the set of teams able to rest their starters on week 17 includes teams entering the playoffs in every seeding.  To really evaluate rest vs not rest, you'd need to compare the success of those teams who rested against the success of other teams in an equivalent situation, such as breaking out home teams versus away teams, and bye week teams versus teams playing in the Wildcard round

But by the time you segregate the data to that point, you're dealing with very small sample sizes, at which point the data is mostly noise.

 

 

25 I'm surprised Minnesota's…

I'm surprised Minnesota's defensive VOA clocks in at 1%. For a while there, they were allowing 4-6 yard runs at will, sometimes many of those in a row. Tevin Coleman had the weekend's top DYAR and never had a run longer than 12 yards. I'd have thought VOA would ding the Vikings more for that.

26 It's Chuckling Time

If the Packers have done anything this year-- other than win 14 games, which is not an achievement to be scoffed at by anyone-- it is the incredible gymnastics and histrionics they have forced both the chattering commentariat and the analytical establishment (yes, FO-- you have become the enemy-- it happens to the best of them-- deal with it) into to proclaim that they really aren't that good. Winning a playoff game most of both of those groups did not expect them to win has only heightened the absurdity, and i damn well expected that DVOA would conclude they really didn't win that game at Lambeau anyway. Because that more than anything else is the new world we live in-- 8-8 Dallas is really good; 14-3 GB is not (and that 31-3 lead in Jerry World was a mirage anyway because Dak proved how much better he was than Rodgers in leading them back to a respectable defeat).

Listen. I'll be simple. Winning matters. It mattered for Seattle that they managed to do it in 11 of 13 close games before yesterday-- but not the one at home that would conceivably have changed everything in the NFC playoffs. it mattered for GB that they have now done the same in 9 of 10 such games. DVOA or no DVOA, eye test or no eye test-- in each of those games somebody on the Packers did what was necessary to prevail. A Jones run. A Rodgers pass. A Smith sack. A Martinez tackle. An Amos INT. A Crosby FG. A rookie coach has had more success than anyone-- ANYONE-- in his fan base, in the league, and, i suspect, even among his own team-- expected. So Rodgers is right-- who cares if it's ugly?

They may get croaked by SF Sunday. They may not. They may pull off an upset and then get croaked by the Chiefs in the SB. In any case, the true takeaway from this season will be one of accomplishment-- and Aaron Schatz and everyone else be damned. I suspect with the draft and free agency their offense will improve next year-- and you know what? They might only win 10 or 11 games despite the improvement. But right now... 28 other teams are watching on TV-- 29 by the start of the game at Levi's-- and the crappy Packers are still playing.

The reason i get on my soapbox about all this is not because i think DVOA is horribly flawed. it is, as Churchill once said about democracy, probably better than all the other failed attempts to measure this particular sport. The reason is because so many people have left their judgement, their romance, their spirit at the front desk when they journey into Analytics Land. It's still a game. Sh*t happens-- it can't be confined to a bunch of measurements. The Packers were better than the Seahawks-- and you know why I know that? Because the scoreboard told me so. They dominated the first half. And they made crucial plays in the second half when they needed to in order to preserve a deserved victory. And oh by the way, they got hosed much worse with the "non-fumble" than the Fail Marys ever did at the end with the Graham spot.

7 or 8 times out of 10, SF wins Sunday, But we'll all be watching to see if the two or three times is what actually happens. And DVOA won't have a damn thing to do with it. Thank Goodness.

28 I notice you never show up…

I notice you never show up when the Packers lose.

 

Edit let me say what others might be too polite to. The fact that you post these DVOA be shamed posts only after the packers win makes you look even worse. At best its a roll your eyes post. At worst, it turns everyone who has no rooting stake into a marginal packer hater.

 

Also note...regular posters who are packer fans also roll their eyes at you. Please consider revisiting your biases if you want to be taken seriously. 

 

And I say all of this as a colts fan...someone who has 0 interest in rooting for or against the packers. 

39 At what age do we outgrow…

At what age do we outgrow irrational fandom? I have plenty of friends who are Packers fans and all, to a person, can't believe they continue winning in this fashion, are enjoying this run for what it is, and also realize it could end disastrously at any point. (One comment: "How does this *crappy* team keep winning?!" followed by incredulous laughter.) They have, without realizing it, echoed Aaron and DVOA's conclusions about the team's quality. This is a decent-to-good team, but nowhere near 13-3 quality. Guess what? It doesn't detract from their ability to cheer for the team. So just enjoy the damn games, sir/madam. There's nothing wrong with being a cheerleader, but own it and stop chiding those who are seeking a comprehensive understanding of the whys behind the results. 

41 Bingo.Remember, Football…

Bingo.

It's good to remember that Football Outsiders was started by a Patriots fan who was looking to show that the 2001 Patriots weren't actually the best team in the league just because they won the Super Bowl!

40 "yes, FO-- you have become…

"yes, FO-- you have become the enemy-- it happens to the best of them-- deal with it"

Dang, if Aaron and Vince and Bryan and company are the ones suiting up on Sunday wearing 49ers uniforms, I really like the Packers' chances of reaching the Super Bowl.

 

49 I'm No Different

I expect the Niners to win-- as i did in November-- because they are a better team. Everybody can tell-- Packer fans or not-- that this year's edition is not a great team.  But the value of 13-3 and now a playoff win is still being undersold in the fealty to metrics that are not be-all and end-all. That's all I'm saying..

And I have a constructive suggestion.

Does the value of plays in the second half (I am excepting the KC 2nd quarter amazement) when team A is leading by a lot change in any way? Either because of a garbage time situation? or because Team A essentially has to perform fewer good plays to seal the ONLY outcome that truly matters-- a WIN than Team B has to perform to try to make it a game or, perhaps, to come back and win the game.

There were roughly 6-7 key plays by GB that sealed that victory. As opposed to at least double that amount by Wilson and Co which brought the Seahawks back to within striking distance. Now in the latter case, several of those plays had redundancy built in-- a key Wilson scramble or pass that had taken place on 1st or 2nd down-- or even 3rd down in GB territory-- could, if hadn't succeeded, still be equalled by a subsequent play. Whereas nearly every one of the key Packer plays was crucial in contributing to the ultimate outcome and was either not able to be replicated (3rd down completion) or had a much weaker chance of success in a subsequent play.

The drive that led to a punt-- which backed up Seattle and then ultimately led to their final punt-- featured a 3rd down completion to Allison and a Rodgers scramble-- both for first downs--which prolonged the drive and got better field position. The 3rd down sack by Preston Smith stopped Seattle from driving for a potentially winning score. (The Alexander sack of Wilson denied them two points) The 3rd down completion to Graham in the 3rd quarter was followed a play later by a TD pass to Adams-- the latter might still have been replaced by capable plays and a Packer TD, but the former if unsuccessful would have led to a punt and the Packers would have been stuck at 21.  And then of course the two final 3rd down passes to Adams and Graham to end the game.

My point is-- and don't tell me it's too complicated to compute-- nothing is--  that these handful or so of plays had a wildly disproportionate impact on the final outcome. They ensured that GB-- not Seattle-- would win. They were more significant than even the first half success which created the 21-3 lead. Because of CLOCK and SCORE. My guess is that DVOA undervalues them... maybe not by a lot, but by something. BIG PLAYS MATTER....A Preston Smith sack of Wilson 3rd and 5 with 4 minutes left in a 5 point game is infinitely more important than the same play in the first quarter...Does DVOA weight these appropriately?

Anyway-- hope for a good game Sunday because sometimes the lesser team wins... or becomes the greater team in a given situation and dynamic (TENN)

 

 

50 Constructive criticism

In reply to by oaktoon

If you had posted this alone without the galactically laughable and frankly absurd original post the collective here might actually engage

 

As it is you have by your own words, not just recently but over the course of years, completely destroyed any semblance of credibility.   As such, and speaking I think for the group...........................whatever dude

51 I am so confused by this…

In reply to by oaktoon

I am so confused by this post given what you wrote in the above post. You concede that this team is worse than the 49ers and also seemingly concede that they aren't as good as their 13-3 record. Yet above, the 13 wins plus the playoff win were used as bullet points A, B, and C for why DVOA is wrong. Sometimes its all about the wins, metrics be damned right? Do I take this first paragraph as a concession that maybe wins aren't the metric we should be gauging team quality? Or are you planning to walk all of this back and revert to the wins mantra if the packers win the game?

 

"and don't tell me it's too complicated to compute-- nothing is--" Lots of things are super complicated that cannot be measured or solved via the tools or technology we have today. So this a pointless remark to throw in.

 

I understand your general point. Big plays swing the game therefore they should count for DVOA more so than the thousand little paper cuts. That point is true for one game, but again as I and others have said over and over again, that's not what DVOA's main goal is. They aren't throwing a bunch of variables to predict the win loss probability of a single game. They are trying to forecast future wins. And the little paper cuts are more predictive than the big swing plays that happened. You may object to that fact; either because you feel DVOA should not be about forecasting but storytelling or you just think that big plays are better forecasting variables than little one's. Either way, that's a problem with your expectations, not an inherent flaw in DVOA. 

 

 

53 You are so enamored with the…

In reply to by oaktoon

You are so enamored with the beautiful big play tree that you completely ignore the continent sized forest of randomness that dominates the outcomes of close games. If meathead football fans are going to use a stupidly miniscule 16 data points to judge the quality of something as massively complex as a football team's performance over 4 months, they at least ought to have enough brains to understand that any victory of less than 6 or 7 points is essentially a tie, that could of had it's binary w/l outcome flipped by a large number of random events that have absolutely nothing to do with how the teams in the contest performed. The ball bounces right instead of left.  A penalty flag  is thrown one time, and kept stowed away another, despite the player behavior being largely the same. A camera angle provides one play a decent review, but not another. Human A, instead of Human B, performs the review. A field goal placement takes place on a decent patch of grass, instead of muck. A gust of wind hits one field goal attempt, or one pass, but not another. A player puts his foot on a particularly bad piece of turf, but the player next to him does not. It goes on and on, and that before we consider how the timing of injuries, and scheduling sequences, can produce 2, 3, 4, or even more, swings in win totals. 

Now, if football players were robots who played a game every day, after a couple hundred games, we would only then begin to have some (and I emphasize some, since even with baseball's 162 game schedule, run differential illuminates more than w-l record) confidence that this stuff was evening out, and that a close loss or victory was telling us something of import. We only get 16 data points either labeled "W" or "L"  however, so those data points which are produced by close scores really don't illuminate much of anything with a high degree of confidence.

It's human nature to reject or discount Randomness, The Great and Terrible God, when examining the outcomes that humans are subject to, in both matters of huge import, or even trivial matters like football team performance. It's kind of unsettling to admit how limited is the control we follically challenged chimps, with superior opposable thumbs, actually have over outcomes. Unsettling or not, however, nope, sorry, but 24-20, or something similar, on the scoreboard, on a half dozen days from September through December, doesn't tell us much. To actually know something, or even strongly suspect something, requires hugely more data.

54 You are so enamored with the…

You are so enamored with the beautiful big play tree that you completely ignore the continent sized forest of randomness that dominates the outcomes of close games. If meathead football fans are going to use a stupidly miniscule 16 data points to judge the quality of something as massively complex as a football team's performance over 4 months, they at least ought to have enough brains to understand that any victory of less than 6 or 7 points is essentially a tie, that could of had it's binary w/l outcome flipped by a large number of random events that have absolutely nothing to do with how the teams in the contest performed.

Now I need to figure out why the quantum superposition always collapses to a loss when the Lions open the box.

65 "Winning a playoff game most…

"Winning a playoff game most of both of those groups did not expect them to win has only heightened the absurdity,"

 

"That said, [a Seahawks win] would still be an upset. This is also clearly a flawed Seattle team that hasn't played nearly as well as its record would suggest this season. With practically no ability to run the ball and a defense with no particularly dominant traits, they have to go on the road to face a team that is healthier and more rested. There's a reason teams fight so hard in the regular season to win those first-round byes: they often make the difference in games like this."

 

https://www.footballoutsiders.com/game-previews/2020/nfc-divisional-round-preview-2020

27 And The Coda Is...

Someone at FO made an editorial judgement in the Tipping Point analysis of the Seahawk-Packer game to spend about 80% of the article extolling the virtues of Russell Wilson and the near-comeback. I couldn't have made my point any better. Damn that 21-3 halftime deficit from that lousy Packer team!!

64 Can't tell if this is a…

Can't tell if this is a serious complaint or not but just to be clear, Tipping Points is a column about fourth quarters so we don't analyze how Green Bay built a lead, we analyze how Seattle tried to come back from a deficit, because that was the story of the fourth quarter.

32 Oaktoon, I will have to…

Oaktoon, I will have to agree with you about the atrocity that is DVOA. Mr. Schatz, don't think I'm not on to you, I've noticed the New York Jets creeping up these rankings even though they haven't played in two weeks. Obviously, you are fudging the books here. Since your favorite dynasty has gone down, you've been reduced to playing favorites with AFC East rivals.  YOU ARE RIGGING THE SYSTEM SO THAT ADAM GASE LOOKS COMPETENT AND THE JETS KEEP HIM FOR TEN YEARS, TEN YEARS OF FUTILITY.  I'M ON TO YOU, BUT IT'S NOT OVER.. NO, IT'S ONLY OVER WHEN I SAY IT'S OVER.  WAS IT OVER WHEN THE GERMANS BOMBED PEARL HARBOR?  NO.  WHEN THE GOING GETS TOUGH, THE TOUGH GET..............

42 positives from this round for the Vikings

* 2nd best Special Teams DVOA

That's about it. Dan Bailey made every kick he attempted. Amir Abdullah had some decent kickoff returns. That managed to offset the fumbled punt returns by Sherels somehow.

44 copypasta

The niners are clearly ranked too low because weighted DVOA is removing all the games in which Dee Ford and Kwon Alexander were healthy, and is reducing the impact of games where Tartt was healthy. Full-season DVOA is way better than this. Selectively choosing which stat to believe based on my region-of-origin is the only way to find real truth!!!

But in all seriousness, I do wonder if weighted DVOA is under-rating the defense in particular. When this defense was healthy at the beginning of the year, it was just bonkers. Now they've got three major contributors back* (Alexander was in as the Sam last game, and was replaced in nickel situations, whereas he was the primary LB before his pec), and played an incredible game Saturday. There are other key rotational guys on the DL who were there early who aren't there now (DJ Jones and Ronald Blair in particular), but getting those three back is so huge.

45 Full season or weighted DVOA…

In reply to by jimbohead

Full season or weighted DVOA are never going to capture those kind of injury/return from injury nuances.  As a Chiefs fan, I'm convinced that where their offense is right now is at the 2018-level, about 10 DVOA points higher than the 2019 level, standard or weighted.  The current weighted offense DVOA has the Titans ahead (it's close) of the Chiefs.  As well as TEN is playing, does anyone buy that? 

This offense is a lot more dangerous when Mahomes can scramble and Williams is taking touches instead of McCoy. But weighted DVOA includes games where Mahomes was still in a knee brace and several that Damien Williams missed.  So over the full-range of games in the weighted DVOA, probably the two offenses are very close.

Context matters of course.  If I were down 24-0, I'd want the Chiefs' offense.  But if I were ahead by the same score, I would prefer the Titans'.

46 Agree fully that these sorts…

Agree fully that these sorts of stats aren't going to catch these nuances. And if we're going to engage in this sort of analysis, we have to be equally charitable or harsh to each team. Since we're being charitable here, for TEN, we absolutely should believe what weighted DVOA is telling us. This is a different team now that Tannehill is playing (sentences I didn't expect to be writing 6 months ago). KC is healthy and playing great ball. GB hasn't lost a game since their post-bye hangover loss in SF. 

There are good reasons to believe in every team remaining, and I'm so here for it. 

52 Did Alexander have an…

In reply to by jimbohead

Did Alexander have an incredible game?  He had no tackles, no sacks, no QB pressures, no passes defensed, and a PFF grade of 58 of 25 snaps.  

I'm genuinely asking.  Everyone on the team seems so happy to have him back there, but the reason isn't showing up on PFR.  

56 I don't know, to be honest…

I don't know, to be honest. I need to re-watch. As I said, he played in base only, not nickel/dime like he did early in the season, and the Vikings had 10 rushes for 21 yards. My biggest concern with Kwon off a pec was his capacity to stack/shed in the run game, and to wrap up tackles. No tackles, but no missed tackles. They weren't breaking big runs to his side or victimizing him in coverage (hi Al'Shaair). So yeah, idk.

To me, it's huge that they got him in the game but didn't need to rely on him for 60 snaps (like JJ Watt in the wild card round). So, he's got his feet wet, isn't going to be super worn out, and maybe there's a better chance that he can do some meaningful work in the next game.

62 Ah I see that part of the…

Ah I see that part of the question is that my original post has some ambiguous grammar. "Had an incredible game" refers to "they", where "they" refers to the 49ers. So, the 49ers got three defensive contributors back, and the 49ers had an incredible game.

71 Is weighted DVOA wrong?

from the article

"The surprising team in DVOA is Green Bay. For the full season (with or without the playoffs), Green Bay is down at No. 10 in our ratings. In the weighted ratings, they're down to No. 15 this week, actually dropping three spots despite the victory over Seattle. The reason for the drop has to do with the nuances of the weighted DVOA formula. Essentially, their big Week 7 win over Oakland dropped significantly in weight this week and they weren't good enough in their 28-23 win over Seattle to make up for that drop."

A week 7 game has a more "significant" effect on a teams rating then a just finished playoff game??? The common sense test says that can't be right. If that what the math says then the formula is wrong. would like to discuss more

72 I think you have this…

I think you have this backward. Weighted DVOA has de-emphasized the impact of that week 7 game, which was highly positive for Green Bay (it looks like their second strongest win on that graph), and has emphasized the impact of last week's win, which was only slightly positive. So, GB's overall weighted DVOA drops, despite last games' rating being slightly higher than their weighted DVOA. 

73 Is weighted DVOA wrong?

"Weighted DVOA has de-emphasized the impact of that week 7 game, which was highly positive for Green Bay"

That's how i understand it. I find it hard to believe a game in week 7 should have such a "significant" effect on a teams outlook today especially when that game looks like an outlier (2nd best of year)

75 Is weighted DVOA wrong?

Can outliers be better weighted? 

picture the largest snake in the world eating a good sized mammal. There's going to be a bulge, that's the Packers v Radiers game, but that bulge should start normalizing rapidly. with in weeks, not months