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

Week 19 DVOA Ratings

Chad Henne
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 1-4, and Weeks 5-12 are somewhat discounted.
  • The ratings listed do not include the adjustments used in the ratings for our playoff odds report. For example, Kansas City moves up to No. 5 in those ratings because we remove the Week 17 game due to sitting starters.
  • Only weighted DVOA is listed for offense, defense, and special teams. Total DVOA is also listed, but adding 10 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.

Buffalo remains No. 1 in weighted DVOA this week despite having the lowest single-game rating of any of the divisional round winners. Nevertheless, the Bills now have a single-game rating over 30% in eight straight games. Only two teams have more games over 30% this entire year, including the playoffs.

Those two teams happen to be sitting right behind the Bills in weighted DVOA: Green Bay at No. 2 and Tampa Bay at No. 3. (The Bucs also pass the Saints for No. 1 in full-season DVOA with the playoffs included.)

Then we have Kansas City down at No. 6. Obviously, our ratings differ substantially from conventional wisdom and even some of the other advanced metrics out there on the Internet. We've been writing about this with the Kansas City Chiefs for the whole second half of the year. It's based on the idea that the Chiefs were essentially not trying hard over the last two months, waiting to "flip the switch" in the playoffs.

This week certainly presented some evidence to support the theory. The Chiefs had 71.5% offensive DVOA in the first half of the game against Cleveland, and that's with a small penalty for playing a below-average Browns defense. This was the best offensive DVOA put up by the Chiefs in any one half of football this season. They only had five other halves with offensive DVOA over 50%, most recently the second half of their Week 12 win over Tampa Bay.

But how do we take that and apply it to Kansas City's performance over the course of this season to get a "real" rating for their offense? It's also possible that Kansas City's excellent first half against the Browns was just the natural variation of a very good offense. Do we assume that Kansas City's offense is as good as their offense from 2018, which would raise their offensive DVOA from 26.1% (their weighted DVOA without Week 17 or the Chad Henne snaps this week) to 35.4%? That would put them neck-and-neck with the Green Bay Packers. Or do we guess that the "true quality" of Kansas City's offense is even better than that?

This is where I would usually say that we don't do that kind of guesswork around here, but the fact is that we do have to do guesswork around here to try to estimate each remaining team's odds of winning the Super Bowl. Patrick Mahomes' concussion means that we have to guess at the chances that Mahomes will be able to start in the AFC Championship Game and factor that into our playoff odds. We also have to estimate what the Kansas City offensive level would be with Chad Henne as the starter. Henne played pretty well in his Week 17 start, and he didn't have Travis Kelce or Tyreek Hill for that game. However, the running game was terrible, I assume because the Chargers could load the box against Henne. So overall, Kansas City had 5.9% offensive DVOA in Week 17.

The playoff odds report right now assumes a 75% chance that Mahomes will play in the AFC Championship. In all simulations, he's healthy for the Super Bowl. We can show how the odds change for each team if we assume Mahomes or Henne is the starter this week. We can also show how the odds change if we assume that Kansas City's offense is actually as good as it was two years ago:

Odds to Win Super Bowl LV
Scenario KC BUF GB TB
Mahomes 75% to play AFC Championship 20.2% 28.0% 31.5% 20.3%
Mahomes healthy 22.5% 25.2% 31.7% 20.5%
Henne starts AFC Championship 13.3% 36.2% 30.7% 19.8%
Kansas City offense as good as 2018 29.4% 21.6% 29.8% 19.3%

You may have noticed that Green Bay and Tampa Bay's odds to win the Super Bowl actually go up if Mahomes is healthy enough to play in the AFC Championship. That's because a healthy Mahomes would give Kansas City a better chance to beat Buffalo, but weighted DVOA would still make Kansas City an underdog to the NFC champion in Super Bowl LV.

The "Kansas City offense as good as 2018" odds are closer to what they are listing in Las Vegas, although even there we still have Green Bay slightly ahead of Kansas City. Clearly, conventional wisdom believes that the Kansas City offense is truly better than what it has shown over the second half of 2020, and probably better than what it showed in 2018.

Note that all of these simulations include home-field advantage. If we truly believe that home-field advantage no longer exists, or does not exist in 2020 -- although both Kansas City and Green Bay will have some fans in attendance -- then Buffalo and Tampa Bay's odds would be higher. Here's another set of odds with home-field advantage removed:

Odds to Win Super Bowl LV (No Home-Field Advantage)
Scenario KC BUF GB TB
Mahomes 75% to play AFC Championship 16.7% 32.1% 27.0% 24.2%
Mahomes healthy 18.9% 29.5% 27.2% 24.4%
Henne starts AFC Championship 10.4% 39.6% 26.3% 23.6%
Kansas City offense as good as 2018 25.4% 25.9% 25.7% 23.1%

Which of the above numbers are the "real odds" depends on your own assumptions about the Kansas City offense and home-field advantage in the 2020 playoffs.

* * * * *

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 championship games on FO+, and snap counts are now updated with playoff data.

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
RK WEI DEF
DVOA
RK WEI S.T.
DVOA
RK TOTAL
DVOA
RK
1 BUF 38.0% 1 15-3 19.9% 4 -10.7% 8 7.4% 3 25.8% 4
2 GB 36.2% 2 14-3 35.3% 1 -4.1% 14 -3.1% 27 27.7% 3
3 TB 35.5% 3 13-5 27.2% 2 -11.8% 6 -3.5% 28 32.8% 1
4 NO 28.2% 4 13-5 5.0% 10 -20.0% 2 3.2% 8 28.8% 2
5 SEA 23.0% 5 12-5 5.7% 9 -7.5% 11 9.7% 2 19.3% 7
6 KC 19.4% 7 15-2 22.6% 3 6.3% 22 3.0% 10 21.2% 5
7 BAL 18.0% 6 12-6 4.9% 11 -10.1% 9 3.0% 11 19.9% 6
8 IND 14.6% 9 11-6 9.6% 7 -4.5% 13 0.5% 17 15.1% 8
9 LAR 12.9% 8 11-7 -5.7% 18 -20.7% 1 -2.0% 25 14.1% 9
10 PIT 7.1% 10 12-5 -8.0% 21 -13.7% 5 1.4% 14 14.1% 10
11 WAS 6.0% 11 7-10 -17.1% 29 -18.8% 3 4.3% 6 -1.1% 16
12 CHI 4.2% 13 8-9 -4.5% 17 -4.0% 15 4.7% 5 0.9% 15
13 SF 3.0% 16 6-10 -7.5% 20 -14.2% 4 -3.7% 29 5.4% 11
14 CLE 2.3% 12 12-6 10.4% 6 8.3% 26 0.1% 18 -3.9% 17
15 ARI 1.9% 17 8-8 -9.6% 24 -11.0% 7 0.6% 16 3.6% 13
16 ATL 1.8% 18 4-12 -2.6% 15 -4.6% 12 -0.2% 20 -4.2% 18
17 MIA 1.4% 14 10-6 -6.2% 19 -7.9% 10 -0.3% 21 4.6% 12
18 TEN 0.0% 15 11-6 14.6% 5 13.0% 30 -1.6% 24 1.2% 14
19 NE -4.5% 19 7-9 -10.5% 25 6.5% 23 12.5% 1 -7.1% 22
20 DAL -7.8% 20 6-10 -11.4% 26 2.7% 18 6.4% 4 -11.1% 23
21 LV -9.9% 21 8-8 -4.3% 16 9.2% 27 3.6% 7 -6.3% 19
22 NYG -10.0% 23 6-10 -9.0% 23 1.0% 17 0.1% 19 -13.8% 25
23 CAR -13.0% 22 5-11 -8.5% 22 6.8% 24 2.3% 13 -6.7% 21
24 MIN -15.4% 24 7-9 6.1% 8 9.6% 28 -11.9% 31 -6.4% 20
25 HOU -17.1% 25 4-12 2.2% 12 18.2% 31 -1.1% 22 -12.5% 24
26 LAC -17.5% 27 7-9 -0.7% 13 4.5% 21 -12.3% 32 -14.7% 26
27 PHI -18.3% 26 4-11-1 -17.0% 28 -1.3% 16 -2.6% 26 -18.8% 28
28 DEN -21.0% 28 5-11 -16.2% 27 3.4% 19 -1.3% 23 -22.1% 29
29 CIN -22.5% 29 4-11-1 -17.7% 30 8.0% 25 3.1% 9 -24.7% 30
30 DET -23.3% 30 5-11 -2.2% 14 24.2% 32 3.0% 12 -15.7% 27
31 NYJ -27.6% 31 2-14 -17.8% 31 4.1% 20 -5.6% 30 -30.5% 32
32 JAX -31.4% 32 1-15 -20.0% 32 12.8% 29 1.4% 15 -27.9% 31

Here are the single-game ratings for the divisional round. Tampa Bay with the highest DVOA of the week may be a surprise, but remember that opponent adjustments are based on full-season DVOA, which had New Orleans first overall. Also, while New Orleans had more yards per play than Tampa Bay, the Bucs had a higher play success rate in both the first and second halves -- a lot of the Saints' yardage came from that one Jameis Winston deep pass to Tre'Quan Smith -- plus the four turnovers.

DVOA (with opponent adjustments)
TEAM TOT OFF DEF ST
LAR 5% 13% 18% 10%
GB 57% 61% 7% 3%
BAL -19% -48% -43% -15%
BUF 33% -7% -51% -10%
CLE -25% -18% 12% 6%
KC 48% 29% -20% -2%
TB 72% 28% -50% -7%
NO -7% -29% -12% 9%
VOA (no opponent adjustments)
TEAM TOT OFF DEF ST
LAR -18% 10% 38% 10%
GB 42% 48% 10% 3%
BAL -45% -52% -21% -15%
BUF 16% -16% -43% -10%
CLE -35% -9% 32% 6%
KC 49% 36% -14% -2%
TB 41% 6% -42% -7%
NO -39% -46% 2% 9%

Comments

34 comments, Last at 21 Jan 2021, 9:48am

2 Cleveland VOA?

"Eye Test" wise this felt like the most competitive game and Cleveland's special teams were also better. Surprised they have the largest gap of the day. 

10 Agreed with eye test

In reply to by zmillard

But on closer inspection, the VOA makes sense. Browns were outgained 7.0 yards/play to 5.1, had an extra giveaway, had a much lower success rate (52% to 42% from Baldwin's box score). The final score was close due to late down variance (although the Chiefs benefitted from that too) and the Chiefs were just 2/5 in the red zone. 

18 eye tests

In reply to by zmillard

KC was crushing the Browns at will.  The Browns had great difficulty stopping massive gains on nearly every play.  

 

The numbers aren't lying here.

20 I’m pretty sure you only…

In reply to by RickD

I’m pretty sure you only watched the first 2 KC drives. They averaged 8.8 yards per play for those. From drive 3 until Mahomes got hurt, KC had 4.8 yards per play. 
 

Conversely, the Browns had 3.2 yards per play for their first 2 drives and 8 yards per play for the rest of the game. 
 

I am very surprised by the divergent DVOA. All plays matter, but the Browns were better for most of the game. I guess that massive first quarter plus the turnover in the end zone were really impactful. I’m guessing there’s a divergent success rate as well since KC generally moved the ball, just slowly after the first couple drives. 

25 Not quite the right numbers

After accounting for penalties and spiking the ball to stop the clock, KC was averaging 7.3 yards per play prior to Mahomes leaving and that includes one drive where they went 3 plays for 4 yards and missed a field goal.

They gained:

70 in 10

85 in 7

68 in 11

70 in 7

4 in 3

60 in 11

That is absolutely moving the ball at will. A full yard better than the best in the NFL this year. This was a complete mismatch until Mahomes went down.

26 Not sure how you are…

Not sure how you are calculating Browns yds/play.  From p-f-r drive stats:

3rd drive: 6 plays/74 net yards (ended in a fumble)

4th drive: no plays; KO return ended half

5th drive:  3/0 (ended in INT)

6th drive: 8/77 (TD)

7th drive: 18/75 (TD)

8th drive 7/12 (punt)

total: 42 plays/238 net yards or 5.7 yds/play.  2 successful drives, 2 turnovers, 1 punt. 

Conversely, the Chiefs:

3rd drive: 13/53 (FG)

4th drive: 9/70 (FG)

5th drive: 4/4 (missed FG)

6th drive (Mahomes): 5/23 (injury)

6th drive (Henne): 5/37 (FG)

7th drive: 5/37 (INT)

8th drive: 8/24 (end of game)

Total:  Mahomes 31 plays/150 yards 4.8 yards per play.  2 successful drives, 1 missed FG, 1 ended by injury

           Henne: 18/98 5.4 yards per play (17/99 5.8 before the kneeldown).  2 successful drives (killing clock is successful), 1 INT

The Mahomes Chiefs from their 3rd drive on had fewer yards per play but a higher drive success rate than the Browns in that span and that's counting the missed FG as a failed drive.  So if you want to take two artificial endpoints, the Browns were "better".  But we don't know how the Mahomes Chiefs would have done the rest of his injury-shortened drive; the Henne Chiefs averaged 7.4 yards per play on the rest of that drive.  Anyone want to argue that the Mahomes Chiefs would have done worse?

Basically the Chiefs with Henne had the same yards per play (taking out the kneeldown) as the Browns from their 3rd drive on, plus a higher success rate on drives (67% vs. 40%).  The backup QB Chiefs were as good or better on a per-play or drive basis than the last 5 drive Browns. 

And if I chose MY desired endpoints, the Browns averaged 3.5 yds/play after Mahomes left the game while the Chiefs averaged 5.8 non-kneeldown yards/play.  Clearly, the Chiefs were "better." 

 

27 Chiefs yards/play

I was just calculating the Chiefs yards per play for the first 6 drives. I adjusted for penalties and spiking the ball to stop clock to look at how much they were actually gaining per snap that counted. KC in drives 3-6 (so excluding the first 2 drives) was averaging 6.3 yards per play which is right around their (very good) season average. But again, that is cherry picking because you are taking out 2 drives in which they were lethal and scored 2 TDs and including a drive where they had incredibly short field and gained 4 yards on 3 plays.

Either way, its hard to argue that KCs offense when Mahomes was in there wasn't elite. They were doing what they wanted.

3 Note that all of these…

Note that all of these simulations include home-field advantage. If we truly believe that home-field advantage no longer exists, or does not exist in 2020 -- although both Kansas City and Green Bay will have some fans in attendance -- then Buffalo and Tampa Bay's odds would be higher. Here's another set of odds with home-field advantage removed:

Home-field probably persists even without fans in Lambeau, especially in late January against a team from Florida.

I am curious to see players' takes on which fields offer a local advantage independent of fans. GB, Chicago, and NE come to mind, with substantial location-specific weather effects. Candlestick used to. Probably Oakland. Pittsburgh, maybe.

6 While I doubt the Bucs will have any issue playing

in the conditions next Sunday all of the Packer beat writers said Saturday's fan crowd did a fantastic job creating noise.  Apparently this is where Lambeau's old school metal bleachers come in handy as fans beat on the bleachers to generate a racket.

 

Regarding the weather my perception is that highly conditioned, mentally focused athletes wearing appropriate gear can manage most weather.  The exceptions being extreme cold (for reference say zero degrees Fahrenheit or less), wind and rain.  The former because that is legitimately uncomfortable and the latter because of how it alters the playing context.  Wind causes tracking any ball in the air tricky.  Extreme rain makes handling the ball and ball exchanges be it handoff or pass more difficult.

 

The Green Bay forecast right now is mid-20's and maybe light snow.  This could affect the footing for Tampa Bay who may have to tinker with what footwear to use.  Otherwise, likely a non-issue for the visitors

15 Buffalo is another candidate…

Buffalo is another candidate, IMO.

The question for me boils down to "which stadiums have weather that genuinely impacts play, but that one team maybe can't prepare for in their own market?" I often see this sort of thing touted for early-season Florida games: the heat and humidity are outside the range of what teams based further north can prepare for at home.

4 Why did VOA like TB's…

Why did VOA like TB's offense so much? They did pretty much squat on drives that didn't start in the red zone.

9 God, I have PTSD just…

God, I have PTSD just thinking about the last 3 NFCC games.  

14: epic choke job that will haunt me forever 

16: completely dominated by Falcons 

19: completely dominated by 49ers

At least with those last two, I knew it was over after one quarter. 

11 2014

Was 7 years ago today, I'm pretty sure. I dread the thought of another Brady SB. 

12 Henne played pretty well in…

Henne played pretty well in his Week 17 start, and he didn't have Travis Kelce or Tyreek Hill for that game. However, the running game was terrible, I assume because the Chargers could load the box against Henne. So overall, Kansas City had 5.9% offensive DVOA in Week 17.

Sure, it probably hurt the Chiefs run offense a bit that Henne was at QB.  Maybe 10-20%?  Here's why the run game was so bad (offense Week 17 snaps/rest of year snaps):

G Wisniewski – 53/98

T Rankin – 53/0

T Durant – 53/38

G Omameh – 53/17

C Kilgore – 53/183

RB Thompson – 49/31

The OL was backups or 3rd or 4th stringers.  Even Wisniewski, who started at G in the 2019 SB run, was a street FA when the Chiefs added him and while he knows the system, the snaps he got after joining the team were with the starting line.  Same with Kilgore; he filled in credibly a couple of games when Reiter was out.  Thompson has been in the doghouse and the 4th string back since an early season fumble.  And Henne's passing DVOA is even more impressive behind that makeshift line.impressive because

 

14 No

No, we did not give Tampa home-field advantage for the Super Bowl. We'll have to consider what to do about that if they win this week.

28 interesting...

In reply to by Aaron Schatz

Under normal circumstances, if it did happen, I would think that homefield advantage would be somewhat diminished given the corporate nature of the event and pricing and so forth (e.g., if Minnesota had made it a few years ago). 

But now, with travel greatly reduced, but at the same time with capacity (probably?) also reduced, would it offset?   Aside from that, normally in florida sporting events you get a lot of support for the away team as Florida has a lot of transplants and is also a popular destination in winter....

Its probably a lot of speculation for something that has never happened and probably won't happen this time.

Even if it does, this will be the sample size of superbowl HFA.

16 Chiefs Again

Does FO staff get as tired of explaining the Chiefs' rating in these weekly columns and we get of reading it? Was there any thought to centering the DVOA column around one of the other 4 team's nuance and details? 

24 When the facts change, my…

In reply to by Aaron Schatz

When the facts change, my opinion changes ... or at least my punchlines do.

 

31 Probably not

No other quarterbacks with college records similar to Allen have turned in the kind of season he turned in this year. We're always reviewing the formula, but you don't make subjective changes to it based on a single player.

21 Postseason parity

The very last line of those SB odds is as close to "all four teams at 25%" as dammit.

I still think the AFC teams are being slightly undersold by DVOA because of an unusual number of blowouts in interconference games during the regular season. One of the sample-size issues for DVOA is that teams play three times more intra-conference games than extra. So it's much more reliable in ranking NFC teams vs other NFC teams than it is in ranking NFC vs AFC. Not to say that it's wrong about the Chiefs – all those close wins are going to depress their numbers. More, I think, that the Packers' odds might be slightly elevated, in part because they dominated the Bears twice, and the Bears played their best games against AFC opponents Houston and Jacksonville.

22 FWIW the betting markets…

FWIW the betting markets have now corrected, with Green Bay as very marginal favorite, and the difference between KC and Buffalo being little more than the equivalent of home field advantage this weekend. Tampa are rated the worst, but not by much. It's clearly very close. 

34 Baltimore - Buffalo

Why would Baltimore's number be so much worse than Buffalo's? Baltimore had more first downs, were better on third down, had more total yards, and 12 more minutes of possession.  Baltimore was inefficient inside the opponent's 35 but was there 5 times and missed two field goals and gave up a pick 6.  Buffalo, if you take away the pick 6 only had one touchdown on offense.