Patriots Dominate in Weighted DVOA

Mac Jones, New England Patriots
Mac Jones, New England Patriots
Photo: USA Today Sports Images

NFL Week 13 - The Tampa Bay Buccaneers drop in our overall DVOA ratings this week but they hold on to the top spot over New England and Buffalo. The big story, however, may be in our weighted DVOA ratings that have a slightly higher correlation with future performance.

As you may know, weighted DVOA lowers the weight of early-season games to get a better idea of how well teams are playing now. It doesn't lower the weight of those games as much as you might expect. A lot of fans have a very short-term view on which teams are good and bad, and only think about the last month or so. Weighted DVOA is still considering every game this season and won't stop considering Week 1 for another couple of weeks. Right now, Week 1 games still count at 40% strength in weighted DVOA. Every other game counts at least 50% strength or more.

Right now, the New England Patriots are the leaders in weighted DVOA by nearly 10 percentage points over the Tampa Bay Buccaneers. The gap between the Patriots and the Buccaneers is larger than the gap between the Buccaneers and the No. 8 Los Angeles Rams. This advantage in weighted DVOA is then translated to our playoff odds simulation, which uses weighted DVOA (in conjunction, for one more week, with a tiny bit of the preseason projections). As a result, we have the Patriots right now as a favorite way out of step with the market. Over half our simulations now end up with the Patriots in the Super Bowl, and one-third of our simulations end up with the Patriots winning the whole thing.

It certainly does seem to be too high. Perhaps there needs to be a variable that considers the quarterback's resume for each team, giving a boost to the teams with the best veteran quarterbacks who have the most consistent records of high performance. Perhaps there needs to be a "rookie quarterback" variable that docks the Patriots for the fact that no rookie quarterback has ever started a Super Bowl. (Kurt Warner wasn't a rookie in 1999, since he had been bouncing around football for five years beforehand.) However, any metric based on 2021 performance alone is going to have the Patriots very high, and any metric that gives a stronger weight to how the Patriots are playing over the last couple months should have them on top of the league.

It's hard to overstate how good the Patriots have been over the last two months. (It's not impossible -- Patriots fans are very good at overstating things.) Let's go all the way back to Week 4, the week of the big reunion game against Tom Brady and the Buccaneers. Opponent adjustments give the Patriots a positive DVOA in that game, even though they lost to the Bucs in the end. Since that game, the Patriots are 8-2. That's a good-sized sample, 10 games going back to the start of October including two losses. The New England Patriots have put up a positive DVOA in nine of those 10 games; the exception was Week 5's narrow victory over Houston. They also put up DVOA over 70% in three straight games from Week 10 to Week 12. Only the Bills and Rams have more games over 50% over the course of the entire season than the Patriots have in just the last month.

Patriots Week-by-Week Graphic

The Patriots' recent dominance stands out if we look at DVOA ratings from Week 4 onwards. Packers fans want to pretend Week 1 never happened? Instead, let's pretend that the entire month of September never happened. Of course, it did happen and it can still teach us something, which is why we still include it in the weighted DVOA formula. But let's see what the top of the DVOA ratings would look like if we only looked at the last 10 weeks:

2021 DVOA, Weeks 4-13
NE 41.5% 1 8-2 16.2% 3 -20.7% 1 4.7% 4
TB 29.1% 2 7-2 24.3% 1 -6.9% 7 -2.1% 24
BUF 26.0% 3 5-4 6.2% 8 -18.4% 2 1.4% 10
IND 24.6% 4 7-3 14.8% 5 -7.1% 6 2.7% 8
DAL 22.3% 5 6-3 5.8% 10 -12.7% 4 3.8% 5
SF 19.9% 6 4-5 13.9% 6 -8.9% 5 -2.9% 25
ARI 18.7% 7 7-2 5.3% 13 -13.2% 3 0.2% 19
GB 16.1% 8 7-2 15.5% 4 -6.1% 9 -5.6% 31
KC 16.0% 9 7-2 5.9% 9 -3.9% 13 6.2% 1
LAC 13.0% 10 5-4 18.6% 2 0.8% 19 -4.8% 30
LAR 12.4% 11 5-4 5.7% 11 -5.7% 10 1.0% 11
MIN 11.6% 12 4-5 5.4% 12 -2.6% 14 3.5% 6

Wow. That is a lead. Unfortunately, it's not easy for me to go back and look at every 10-game stretch dating all the way back to 1983. But I can look at the first 10 games of each season. Since 1983, only 16 teams have put up DVOA over 40% over their first 10 games of the season. That's what the Patriots have done over their most recent 10 games. You probably already knew how good the Patriots have been on defense and special teams over the past two months. But notice that the Patriots come out third on offense behind the Buccaneers and Chargers. That includes ranking sixth rushing the ball since Week 4 and -- I swear I am not making this up out of some sense of homer obligation -- first passing the ball since Week 4. With a rookie quarterback! The Patriots have 37.9% passing DVOA since Week 4. Tampa Bay is at 37.6% passing DVOA since Week 4. This will stun you, but the two leaders in net yards per pass since Week 4 are the San Francisco 49ers at 7.7 yards and the Patriots at 7.6 yards. Dallas is third at 7.2 yards.

(Aside: obviously, those passing numbers include only three plays from last night's game. For those who are curious: there are not at this time specific weather adjustments in DVOA for high-wind or high-precipitation games. Offensive and defensive DVOA simply have an indoors/outdoors adjustment. Special teams has general weather adjustments but they're based on the week of the season and the usual temperatures in that city, not the specifics of each game.)

I hope this table helps explain why the Patriots are so far ahead of everyone else in our Super Bowl simulations. They really have been that good over the last 10 games.

Below the Patriots, the top 10 mostly stays the same as last week. The Dallas Cowboys jump Arizona into the No. 4 slot and the Green Bay Packers jump Minnesota into the No. 10 spot. (I'll write more on the Vikings before Thursday's game.) The big mover is the Los Angeles Chargers, who went from 16th to 12th after beating up on the Cincinnati Bengals. The Chargers were also the team with the biggest improvement in playoff odds this week, going from 52.4% to 80.6% odds of making the postseason.

Houston and the History of Grotesque Ground Games

Thank you to reader Todd Singer for pointing this out during yesterday's Football Outsiders livestream: the Houston Texans have a historically bad running game this year. In fact, the Texans have the worst run offense DVOA we've ever measured through Week 13. It's not like they haven't done this before. Last year's Texans had the fifth-lowest run offense DVOA ever tracked through Week 13. And the Texans are not alone. Despite some of Cordarrelle Patterson's heroics -- which have mostly been as a receiver, not as a runner -- the Atlanta Falcons also rank among the worst run offenses we've ever tracked through Week 13.

Worst Run Offense DVOA
Through Week 13, 1983-2021
Year Team W-L Run Offense
2021 HOU 2-10 -43.4%
2019 MIA 3-9 -43.1%
2016 JAX 2-10 -37.8%
1991 IND 1-11 -35.4%
2020 HOU 4-8 -35.0%
2013 BAL 6-6 -34.7%
2021 ATL 5-7 -34.0%
1998 NO 5-7 -33.8%
2015 CLE 2-10 -33.4%
2005 ARI 4-8 -33.3%
2015 IND 6-6 -33.0%
2017 ARI 5-7 -32.8%
2010 CAR 1-11 -32.1%
1986 NE 10-3 -32.1%

The Texans are averaging just 2.95 yards per carry when we remove scrambles and only look at planned runs. The Falcons are in a weirder place. They are partly so low in run DVOA because we now count aborted snaps as running plays and the Falcons have five of those. However, the Falcons are also 27th in the league with 3.82 yards per carry and they are 30th with a 35% success rate on running plays.

(By the way, going back in time, how about those 1986 New England Patriots? The Patriots averaged just 2.9 yards per carry as a playoff team, with all three of their main running backs under 3.0 yards per carry.)

However, as I've written about in the past, things are a little more complicated than this. See, the average offensive DVOA is 0.0%. But the average run offense DVOA is not 0.0%. Team DVOA, including run/pass splits, have a baseline of all offensive plays (plus the penalties that are included in DVOA). Since passing is more efficient than rushing, passing will always have a positive average DVOA and rushing will generally have a negative average DVOA. The passing premium has generally increased over time. Here's a look at the average pass offense DVOA each year as of Week 13; 1987 is missing since that year had only 12 weeks.

Average Pass Offense DVOA Week 13

Average run offense DVOA bounced around but tended to average about -3.0% or so from 1983 to about 2007, but in the 2000s it fell dramatically before rebounding a little bit over the last five seasons. This year, the average run defense has a DVOA of -8.4%.

Average Run Offense DVOA Week 13

You might have figured out where I'm going with this. Our list of terrible running games above compared each team to the league average for all plays, which every year is 0%. But a more proper way to compare running games is to compare rushing DVOA to the average DVOA for run plays only in that season. So I did that. And guess what? The worst running game still belongs to the 2021 Houston Texans! The Atlanta Falcons have fallen from the top of our list, but the Texans still have the worst running game ever tracked through Week 13 since 1983.

Worst Normalized Run Offense DVOA
Through Week 13, 1983-2021
Year Team W-L Run Offense
Run Offense
vs. Average
2021 HOU 2-10 -43.4% -35.0%
2019 MIA 3-9 -43.1% -34.2%
2005 ARI 4-8 -33.3% -31.6%
1998 NO 5-7 -33.8% -30.4%
2002 HOU 3-9 -29.7% -29.9%
1991 IND 1-11 -35.4% -29.7%
2000 SD 1-11 -28.8% -28.5%
2004 SF 1-11 -28.9% -27.9%
2013 BAL 6-6 -34.7% -27.5%
1986 NE 10-3 -32.1% -27.3%
2016 JAX 2-10 -37.8% -27.0%
2000 ATL 3-10 -27.2% -26.9%

Can the Texans break the all-time record for running game futility? Tune in over the next few weeks and find out.

* * * * *

Football Outsiders playoff odds, snap counts, and the FO+ database are now all updated through Week 13. A reminder that all our free stats pages, including DVOA and player position stats, require registration to view. This is not a paywall! You only need to register (for free) and then log in to the site to view these pages. While you're at it, you can get a seven-day trial of FO+ and check out the FO+ features like a deeper DVOA database, weekly fantasy projections, fantasy football research tools, and picks against the spread.

* * * * *

Here is the Football Outsiders Top 16 through 13 weeks of 2021, 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 for opponent and performance indoors and consider all fumbles, kept or lost, as equal value. SPECIAL TEAMS DVOA is adjusted for type of stadium (warm, cold, dome, Denver) and week of season. As always, positive numbers represent more points so DEFENSE is better when it is NEGATIVE.

WEIGHTED DVOA gives recent games more strength than older games to get a better idea of how well teams are playing now.

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>

1 TB 28.7% 1 26.8% 2 9-3 25.6% 1 -4.5% 8 -1.5% 25
2 NE 28.0% 2 36.5% 1 9-4 8.1% 11 -17.0% 2 2.8% 4
3 BUF 25.0% 3 22.9% 4 7-5 3.4% 14 -21.0% 1 0.6% 14
4 DAL 24.7% 5 23.0% 3 8-4 10.7% 7 -11.8% 4 2.2% 7
5 ARI 22.3% 4 18.9% 7 10-2 7.8% 12 -13.4% 3 1.1% 12
6 LAR 20.1% 6 17.2% 8 8-4 13.6% 5 -6.9% 6 -0.4% 20
7 SF 16.7% 7 19.2% 5 6-6 13.8% 4 -3.9% 9 -1.0% 22
8 IND 14.2% 8 19.1% 6 7-6 8.8% 9 -3.2% 11 2.2% 6
9 KC 13.6% 9 16.1% 9 8-4 12.0% 6 3.8% 25 5.4% 2
10 GB 10.6% 11 14.6% 10 9-3 15.0% 3 0.0% 14 -4.4% 30
11 MIN 10.2% 10 11.8% 11 5-7 9.3% 8 1.7% 17 2.7% 5
12 LAC 7.6% 16 9.4% 12 7-5 15.5% 2 2.2% 19 -5.6% 31
13 NO 5.7% 12 1.7% 14 5-7 -3.5% 21 -9.3% 5 0.0% 17
14 CLE 4.9% 14 1.4% 15 6-6 5.8% 13 0.6% 16 -0.3% 19
15 PHI 4.2% 15 4.2% 13 6-7 8.2% 10 4.2% 26 0.3% 16
16 BAL 2.1% 13 -2.0% 16 8-4 0.2% 19 3.8% 24 5.7% 1

Click here for the full table.


76 comments, Last at 11 Dec 2021, 1:55pm

#1 by nat // Dec 07, 2021 - 4:24pm

Is DAVE screwy for the Patriots? I don’t see how DAVE could equal weighted DVOA for them. Pre-season projections weren’t all that high for them, certainly not 36.5%.

Points: 0

#66 by MC2 // Dec 08, 2021 - 2:26pm

All the teams that have already played 13 games have a DAVE that is identical to their weighted DVOA, so I would assume that after a team has played 13 games, the projection component of DAVE no longer applies.

Points: 0

#67 by nat // Dec 08, 2021 - 2:28pm

That’s got to be it.

I should have figured it out myself. Thanks.

Points: 0

#3 by Hoodie_Sleeves // Dec 07, 2021 - 5:01pm

This will stun you, but the two leaders in net yards per pass since Week 4 are the San Francisco 49ers at 7.7 yards and the Patriots at 7.6 yards. Dallas is third at 7.2 yards.


It shouldn't stun anyone. They're 5th in the league in receptions > 20 yards despite being 15th in total passing yards, and near the bottom in attempts. And 7th in Y/Attempt. And they've been drastically better since the Covid/Line/Etc issues in the first 4 weeks.

But instead we've got official Walkthrough hot-takes like "Mac Jones is performing on the same level as Mark Sanchez his rookie year"

Points: 0

#5 by Independent George // Dec 07, 2021 - 5:36pm

The problem is that the very best comparison to how the Pats are using Mac Jones is... the 2001 NE Patriots. And I imagine almost every writer, whether they love the Pats or hate them, absolutely does not want to make that comparison for obvious reasons.

Points: 0

#12 by Alternator // Dec 07, 2021 - 7:30pm

2004 Steelers are also an excellent comparison - high efficiency rookie bolstered by a quality running game and an excellent defense.

Points: 0

#60 by b1ff // Dec 08, 2021 - 1:56pm

I think a lot of sports prognosticators are making the connection for the obvious reasons, but they're also saying that they aren't saying that he's the second coming of Brady.

Points: 0

#50 by PatsFan // Dec 08, 2021 - 10:52am

Yeah.  Tanier has decided his anti-NE schtick is more profitable than paying attention to analytics.  Odd for an analytics site, but then one does need to keep the lights on.

Points: 0

#4 by Raiderfan // Dec 07, 2021 - 5:02pm

I am always grateful when “open the curtain” and let us see behind it to how the calculations go.

Points: 0

#6 by Bobman // Dec 07, 2021 - 5:41pm

Colts vs Pats, weighted DVOA 6 vs 1, rankings 9/11/6 versus 11/2/4. Not Manning-Brady, but Reich-Belichick.

5:20 pm Saturday, Seattle time. I'm coaching a wrestling tournament that day and will have to figure out an excuse to make sure I am home early.  If Indy had one more win (say the OT game vs Tenn) I'd be a lot more relaxed about this.  But too much is on the line. At least it's at home. 

Points: 0

#7 by nat // Dec 07, 2021 - 6:04pm

If home filed is worth 15% DVOA and we go by DVOA instead of weighted, it’s anybody’s game.

As always, if DVOA says it’s near a tossup, look to specific matchups, injuries, etc.

I have no idea who is going to win that one. At least it won’t be windy.

Points: 0

#9 by Independent George // Dec 07, 2021 - 6:25pm

If Indy can run it like Tennessee did (except without the turnovers), this could be an amazing battle. 

Points: 0

#44 by Aaron Schatz // Dec 08, 2021 - 10:39am

Historically, home field was worth about 17% DVOA. The last three years, it's been worth nothing.

That being said, as a matchup, for me as a Patriots fan, I think this is the hard one. I'm more scared of the Colts running game and the Colts front seven stopping the Patriots running game than I am scared of losing to Buffalo in the rematch.

Points: 0

#51 by Lost Ti-Cats Fan // Dec 08, 2021 - 11:04am

"and the Colts front seven stopping the Patriots running game"

If there is a running game.  This might be the "NE only runs 3 times" game.  You know, just to keep things balanced overall.


Points: 0

#52 by Pat // Dec 08, 2021 - 11:10am

If I were a Patriots fan I'd be scared of that, too. The Patriots didn't run every play just for lulz: they at least partly did it because they were worried about how their quarterback would respond. Belichick's been pretty freaking amazing at adapting his team to put them in the best position to win, but I can't see how a sane New England fan wouldn't be able to recognize that this team's limited, so it's definitely possible for them to be schemed into a corner.

Points: 0

#57 by Anon Ymous // Dec 08, 2021 - 12:30pm

The Patriots didn't run every play just for lulz: they at least partly did it because they were worried about how their quarterback would respond.

A key question here is, respond to what?  To extreme weather?  To the pressure of being the primary driver of the offense?  Or the pressure of it being a big division game?

The former is absolutely reasonable.  Mac's arm strength is fine, but it's not the proverbial cannon.  And perhaps his best attributes are how precise he is and how catchable his passes are, neither of which translate well to those kinds of conditions.  

The latter two, though, are hard to support.  It was just a week ago that Mac threw 32 times vs. 21 rushing attempts.  21 rushing attempts that were mostly futile until the game was well in hand in the 4th quarter.  

Then there's the game against Tampa, where NE ran for negative yards and had no choice but to turn the reigns over the Mac.  31/40 for 275 with 2/1 TD/int with no running game, an offensive line that was getting him regularly pummeled and weapons that weren't clicking yet.  All on a rainy night in about as high pressure a situation as a rookie can face in their first month.  Tampa's dreadful secondary certainly helped those numbers, but the other factors still count.  

Not saying that Mac couldn't have a stinker if Indy stops the run.  Just that the Buffalo game offers little reason to believe he will or that the team is concerned about putting the game in his hands.  

Points: 0

#59 by Lost Ti-Cats Fan // Dec 08, 2021 - 12:39pm

That's pretty much my take, too.  As long as Belichick thinks he can win on run-heavy and D, he's going to avoid giving Jones much opportunity to mess the game up.  But if the game situation / match-up dictates "pass first", I think at this point in the season (maybe not in the first month or so), Belichick lets McDaniels go pass-heavy and sees what happens.  The game against Indy's as likely a game as any for that to happen.  Then maybe it's a disaster for NE or maybe not.

Points: 0

#62 by Pat // Dec 08, 2021 - 1:58pm

Not saying that Mac couldn't have a stinker if Indy stops the run.  Just that the Buffalo game offers little reason to believe he will or that the team is concerned about putting the game in his hands.  

So, to be clear, all I was really saying is that the team is concerned about somethingYou're absolutely right - we have no idea what went into that reasoning. But Monday was a flamingly obvious "change your stance to guard an Achilles heel" move, so the idea of being forced into exposing what might be a weak spot would definitely worry me.

Obviously, weather's a totally reasonable explanation, and that doesn't apply next week. But again, if I'm a fan and I see my coach historically protect a portion of my team, that'd make me worried almost no matter what. Belichick's a great coach but in my opinion, one mistake I've seen him do is overreact, to the overall detriment of the team (see: SB LII, 4th down conversions, etc). That game versus Buffalo could've easily backfired dramatically.

Points: 0

#72 by nat // Dec 08, 2021 - 6:22pm

I think you’re overreacting here.

The winds were gusting to 45+. The Bills are the #1 pass defense. The Patriots got an early lead on a day where scoring was going to be very hard. The Patriots felt they had a personnel advantage running the ball. Any of those could lead to a run-heavy strategy. All four of them together made running a lot the smart thing to do.

There’s no need to inject the idea of an “Achilles heel”. They’ve gone the other extreme (85% passes) and came close to upsetting the Bucs. It’s more that Belichick and company trust their strategic analysis and game plan. If the plan says come to the line expecting to run and check to a pass only in the most favorable conditions, and if “favorable conditions” has been adjusted to deal with high winds, then that’s the plan.

Points: 0

#19 by turbohappy // Dec 07, 2021 - 10:36pm

I'm actually in Indiana for a bit so I'm going to be at this one. Should be a good battle I think. Indy really does have a lot more on the line...NE could lose this game and still have a pretty easy path to winning the division. I'm nervous Belicheck is gonna be able to convince Wentz to do something stupid a couple times but if not this should be a hell of a game.

Points: 0

#27 by Bobman // Dec 08, 2021 - 12:55am

Enjoy the game. I miss being optimistic for my team and BB has been the bully stealing my lunch money, candy, and happiness for so long. So I'll be a little nervous for two weeks, getting more tense each day.

I suspect it will be less cerebral than some of the old Manning-Brady games, and more even than the Legarrette Blount steamroller games, with kind of an "I punch you, you punch me" feel to the run games. So long as Wentz plays within himself (and runs a few times rather than try to fix every broken pass play with his wondrous arm)....

Points: 0

#69 by horn // Dec 08, 2021 - 2:59pm

Don't worry, even if Wentz is 36 for 31 passing for 599 yds and 6 TDs, Tanier and FO will still find a way to say he's terrible. [no typos in the above]

Points: 0

#71 by Independent George // Dec 08, 2021 - 5:22pm

If Wentz creates a universe-destroying temporal paradox, then he would indeed be the absolute worst

Points: 0

#31 by PapaNarb // Dec 08, 2021 - 6:08am

Colts also get the pleasure of playing 5 rated Arizona, After already having played 1 TB, 3 BUF, 6 LAR, and 7 SF. Talk about a well tested team!

Points: 0

#8 by Jimmytossinbeachble // Dec 07, 2021 - 6:20pm

Weighted DVOA paints a pretty good picture as to why being "hot" is more than just a win streak. Miami's weighted DVOA is barely better than their overall despite being 5-0 L5. 

Points: 0

#10 by AFCNFCBowl // Dec 07, 2021 - 6:35pm

That's what happens when you play the Texans, Ravens, Jets, Panthers and Giants.

The most sickening part is that the Dolphins still get to play the Jets, Saints, and Titans

Points: 0

#17 by RickD // Dec 07, 2021 - 10:18pm

There's more to that team than Derrick Henry.  

I don't see Miami getting past the problems posed by their 7-game losing streak.  If they'd been playing at this level all season, they'd be a playoff team.

Points: 0

#11 by big10freak // Dec 07, 2021 - 7:09pm

I have been fairly consistent in sharing that the Packers defense looks to be better this season.  Figured the numbers would eventually catch up


Separately I am hopeful for a Xmas miracle on the special teams.  Having experienced multiple post seasons I know deep in my bones that special teams will be the team's undoing at the worst possible time.  A bitter pill to say the least

Points: 0

#56 by Arkaein // Dec 08, 2021 - 11:46am

I think the Christmas miracle could be getting Jaire, Z, and Bakh all back and playing in time for a playoff run.

It's not very often that an already good team can add 3 Pro-Bowl caliber players in December, but it's certainly possible in Green Bay.

Points: 0

#13 by Anon Ymous // Dec 07, 2021 - 7:31pm

Speaking of including September, NE's loss to Miami means they cannot hold the tiebreaker over Buffalo if they end up splitting the season series with them.  So, even if Buffalo loses to Tampa this week, losses by NE to Indy and Buffalo - eminently plausible - would give the Bills the division lead heading into the final two weeks.  

As well as they've been playing, the perch is anything but secure.

Points: 0

#14 by Lost Ti-Cats Fan // Dec 07, 2021 - 7:55pm

"Speaking of including September, NE's loss to Miami means they cannot hold the tiebreaker over Buffalo if they end up splitting the season series with them."

Are you assigning a 0% probability that BUF loses their rematch against NYJ? 

Points: 0

#15 by AFCNFCBowl // Dec 07, 2021 - 7:57pm

I mean, realistically, there is a 0% chance of BUF losing to NYJ.

Points: 0

#21 by Independent George // Dec 07, 2021 - 10:44pm

#1 in variance. They might beat Tampa, and then get blown out by the Jets.

Points: 0

#23 by Q // Dec 07, 2021 - 11:03pm

An argument can be made that Buff is 1 of the more predictable teams this year. In general, they smash the terrible teams and play close games (and generally lose against good teams).

A few exceptions exist (vs Jax and IND. Possibly KC depending on your thoughts of them at that time of the year).

In general, though, Buff can be understood pretty well with the simple: Destroy Bad, Close (usual loss) vs Good Heuristic 

Points: 0

#33 by lauers // Dec 08, 2021 - 7:30am

Add the Steelers to the exceptions list too. The existence of exceptions probably means that the odds of a Jets victory are non-zero. But if you're willing to offer infinite odds, I'll gladly bet a penny on the Jets!

Points: 0

#28 by Bobman // Dec 08, 2021 - 12:57am

we're talking a "negative zero" chance?

Points: 0

#35 by Mike B. In Va // Dec 08, 2021 - 8:24am

I dunno. I don't put anything past the Bills at this point, since McDermott is basically saying "it's not my fault". His messaging after the NE loss is a bad look.

Points: 0

#22 by Q // Dec 07, 2021 - 10:56pm

I wonder if this Pats team has the record for Highest DVOA this late in the season despite 0 wins against what the general public would consider top level competition. Their 3 Best Wins to most people seem to be:

-A squeaker vs a pass heavy, yet disappointing team in a windstorm that is currently the #7 Seed in the AFC.

-Tenn after almost all of their Top Offensive Players were injured

-A road win against tbe wildly up and down Chargers that most have given up on.

The cloest parallel I see to them is Brady's last year that that through 3/4 of the Season DVOA thought was 85 Bears caliber after like 2 straight months of High School caliber QB's they faced.

Points: 0

#29 by Alexdoe42 // Dec 08, 2021 - 2:13am

I don't trust the Patriots. Like you point out, they have lot of wins against teams like browns, titans when they had major injuries in whole position groups.

Similarly chiefs are big suspect. All the best teams are in nfc. 

Points: 0

#32 by Lost Ti-Cats Fan // Dec 08, 2021 - 7:16am

In reply to by Alexdoe42

"All the best teams are in nfc. "

I suspect that's a mirage based on there being fewer competitive teams in the NFC and more weak sisters.  The larger pool of AFC competitors are beating each other up and suppressing their records.  More than just the extra win or two, though, the NFC teams look better more often because they're playing weak opposition more frequently.

There's almost perfect symmetry in the out-of-conference records of the 4 division leaders in each conference.  The difference lies in their inter-conference games, where the top NFC teams are benefitting from a softer schedule.  Flip the 4 division leaders to the other conference and I'd guess it would still look like the best teams are in the NFC, even though it would now be different teams.

Out-of-conference record:

Ari 4-0, KC 5-0

TB 3-0, BAL 3-0

GB 2-1, TEN 3-1

DAL 2-3, NE 2-3

Points: 0

#37 by Q // Dec 08, 2021 - 8:46am

I don't think people would feel the same if the Conferences were flipped.

-Ari is considered a Top team and doesn't need any explanation.

-GB has only 2 losses with its MVP QB and a road win at ARI. Sweeping the NFC West is a pretty nice feat.

-TB is the Defending SB Champs and has The Most Accomplished QB ever. They also have a win @NE, @healthy DAL, etc

-DAL is an unknown. When healthy their Offense is loaded. A win @NE is a pretty nice accomplishment.

Let's look at the AFC:

-KC's best win on their whole schedule is probably @Philly? @Wash? Beating Jordan Love is far from impressive. Beating DAL without Lamb or Cooper is like beating the 90s Vikings with both Moss and Carter injured.

-BAL's best win is @Chiefs. Their best win vs the NFC is a home win vs MIN.

-TENN has loads of quality wins. They are not considered top tier by the public due to crippling injuries to their best players. This can change if they return.

-NE has lost to every quality NFC team they have played, despite both being home games. 

The NFC teams just have far more impressive wins as a group. TENN has more good wins than NE + KC + BAL combined.


Points: 0

#38 by Anon Ymous // Dec 08, 2021 - 9:14am

If we're going to dig into the shape of each roster at gametime (reasonable, I think), it could be noted that:

* The OL was in shambles and was at least as bad relative to their current performance as TN's and Dallas' receiver units were for their games against NE and KC, respectively.

* Bourne and Henry hadn't emerged yet.

* Rhamondre Stevenson was still in the doghouse, with 5 rushing attempts combined in both games.  

* The defense was still finding its way.  The secondary is playing a much different style now, the front seven had a lot of new guys that didn't really gel until the SD game and LBs like Van Noy (nursing an injury) and Hightower (slowly returning from taking last year off) weren't playing with at their current levels.

I could add that the coaches were also still finding their way.  That part of the season was plagued with silly errors, like 12 men penalties, punt protection errors, time outs that had to be taken at inopportune moments, etc.  For the most part, these self-inflicted errors have been corrected (having Harry returning a punt this past week notwithstanding).  

I applaud your effort to have a nuanced discussion, and as such, shouldn't your inclusion of NE's losses to TB and Dallas take the above into consideration?

Points: 0

#41 by RBroPF // Dec 08, 2021 - 10:01am

Those are all fair enough points, but the larger issue is that there is nothing "nuanced" about just looking at Ws and Ls. 

In the Tampa game, NE doinked a go-ahead FG in the final two minutes.  The Cowboys game went to OT.  Both of those games could have easily gone either way if one play had gone differently.  If someone is basing their "analysis" of a team's strength on the results of one play, then they should probably re-think that.

The entire driving insight of this website is that Ws and Ls are not at all the best way to assess teams' performances.

Points: 0

#34 by Anon Ymous // Dec 08, 2021 - 8:19am

Well, Buffalo would have been the #2 seed had they scored a TD on either 4th quarter drive and SD was 4-2 at the time, considered by most to be at least a near-elite team.  As much as you can discount stopping TN and Cleveland's injured offenses, there really isn't much that can be done to minimize how well NE moved the ball against both teams.  

All told, this isn't comparable to the 2019's run of opponents.  The opponents themselves are a perfectly ordinary bunch that every team faces in a 5-6 game stretch every year.  The injuries have definitely been a break, but that's more a streak of good fortune that happens several times each year (the Chiefs have gotten at least as much luck during their recent run, for instance), only most teams don't take advantage of it to the extent that NE has.  


Points: 0

#54 by Alternator // Dec 08, 2021 - 11:11am

Minor correction, Buffalo would have been the #2 seed, not the #1 - the Titans would have been the #1 if Buffalo won on Monday.

Points: 0

#24 by jheidelberg // Dec 07, 2021 - 11:49pm

Aaron Schatz:

It certainly does seem to be too high. Perhaps there needs to be a variable that considers the quarterback's resume for each team, giving a boost to the teams with the best veteran quarterbacks who have the most consistent records of high performance. Perhaps there needs to be a "rookie quarterback" variable that docks the Patriots for the fact that no rookie quarterback has ever started a Super Bowl.

I remember someone doubting his own data and picking KC over TB last year, questioning the DVOA model as so many other respected colleagues in the analytic field had selected KC.  Trust your data!!!  So many of your readers have the faith.  If there needs to be a variable that considers the QB's resume for each team or a rookie QB variable, that can be another study done in the off season.

I must assume that NE is a sample size of 1 in which weighted DVOA is number 1 this late in the season let alone it being by a large margin with a rookie QB starting all or most of their games.  Even 2004, 15-1 Pittsburgh  (Roethlisberger) was number 2 in weighted DVOA at the end of the season.

Trust the process!!  Trust the data!!

Your modesty is admirable, I appreciate your lack of boasting about your model and constant studies to improve upon said model.



Points: 0

#26 by Alternator // Dec 08, 2021 - 12:37am

Plus, who did 2004 Pittsburgh lose to?  The eventual Super Bowl winners, in the AFCCG.  Getting that far justified the DVOA rating.

Points: 0

#65 by b1ff // Dec 08, 2021 - 2:17pm

I would argue that the Pat's DVOA number probably is too high. I don't buy the rookie QB thing though, he's been a rookie QB all season long in the data.

What (I think) DVOA doesn't have visibility to is the luck the Pats have had w.r.t to key players out in some of the games they've played, or in the case of the Bills game, extreme weather conditions that happen to play perfectly to the relative strengths of the Pats.


Points: 0

#25 by mrh // Dec 07, 2021 - 11:57pm

Chiefs and the rest of the way:  if offense plays like September and defense plays like November/December, they win the Super Bowl, NE's weighted DVOA notwithstanding.  But if the offense plays like November/December and the defense plays like September, they won't even make the playoffs.  If the current trend continues, they'll make the playoffs but a deep run once there looks unlikely.

My personal theory is that Mahomes has a bum shoulder, which is why he's dropping down to 3/4 motion so much.  The pain isn't too bad when the games start, so the 1st drive or two goes well.  But once the pain kicks in, he can't throw as well (and is more reluctant to tuck and run than he has been in the past) and the offense tails off. 

Points: 0

#39 by barf // Dec 08, 2021 - 9:35am

You're assuming a lot about Mahomes with no proof - he hasn't been playing up to his normal standards, but still has 25 TDs. He would show up on an injury list if there was a problem. This is more about the offense regressing from historic levels to being merely decent. He's not hurt, they're just not as good as they were the last 3 years. 

Points: 0

#40 by Anon Ymous // Dec 08, 2021 - 9:41am

He would show up on an injury list if there was a problem.

Is this true?  Isn't the rule that someone only has to be listed if they are limited in some way?  It also seems to me that we hear all the time about guys who end up having surgery for some malady that hadn't been disclosed before then.

Not saying that I agree with the speculation about Mahomes, just that I'm not convinced a shoulder issue would definitely be listed.

Points: 0

#47 by Independent George // Dec 08, 2021 - 10:47am

Famously, Peyton Manning needed offseason neck surgery and apparently couldn't feel his fingertips all year, none of which ever showed up on the injury report. 

Points: 0

#53 by Lost Ti-Cats Fan // Dec 08, 2021 - 11:11am

NHL injury "reports" famously say only "upper body injury".  While NFL players are, no doubt, more civilized and respectful than their NHL counterparts, if Mahomes had a reported issue with a shoulder, I'd be concerned that opponents might accidentally contact that shoulder somewhat more often than otherwise might occur.

Perhaps my fears are unwarranted, but I wouldn't take a chance with my franchise QB.  Err on the side of saying nothing.

Points: 0

#55 by Independent George // Dec 08, 2021 - 11:32am

Or you can go the opposite direction, and report Tom Brady as 'probable (shoulder)' every week for over a decade.

Points: 0

#30 by RobotBoy // Dec 08, 2021 - 3:19am

The Pats running game was excellent in '85 but dreadful in '86, even though it lost only one starter from the previous year. Of course, that player was John Hannah, who retired after the superbowl. Tony Eason was mediocre at QB but got the starts over Steve Grogan, who was still better at 34 and semi-broken. The team finished 11-5 because the defense was still good.

Points: 0

#36 by dryheat // Dec 08, 2021 - 8:44am

It's interesting, at least to me, to recall that when Steve Grogan retired, he physically looked like a cripple, and hadn't  been able to walk normally in several years.  Tom Brady is now seven years older than Grogan when he retired.

It's a cautionary tale for running quarterbacks.  And Grogan had good offensive line play for most of his career.

Points: 0

#42 by Anon Ymous // Dec 08, 2021 - 10:04am

KC's defensive DVOA only being 13th since October surprises me.  They've been very productive lately and I would have expected a large boost from shutting down GB and Dallas.  What is DVOA not as impressed by?  Or were the Buffalo and TN games bad enough that they weigh things down?

Points: 0

#45 by Aaron Schatz // Dec 08, 2021 - 10:42am

All a matter of which weeks you look at. The Kansas City defense was bad in Week 5 and absolutely horrendous in Week 6. Since Week 7, the Chiefs are second in the NFL behind the Patriots!

Points: 0

#48 by AFCNFCBowl // Dec 08, 2021 - 10:48am

All a matter of which weeks you look at. The Kansas City defense was bad in Week 4 and absolutely horrendous in Week 5. Since Week 6, the Chiefs are second in the NFL behind the Patriots


    Points: 0

    #43 by samueljames530 // Dec 08, 2021 - 10:24am

    The NFC West race might be more interesting down the stretch than it looked two weeks ago. The playoff odds are pretty much giving it to Arizona right now, but interestingly weighted DVOA seems to think the Cards and Rams are more similar than different over the past few weeks. If the Rams get the upset on the road Monday night (+2.5, which seems favorable to the Rams!), then you've got the 9-4 Rams and 10-3 Cards down the stretch like so:

    Wk 15:

    Cards at Lions

    Seahawks at Rams

    Wk 16: 

    Cards vs Colts

    Rams at Vikings

    Wk 17:

    Cards at Cowboys

    Rams at Ravens

    Wk 18:

    Rams vs 49ers

    Cards vs Seahawks


    That's a pretty even slate of opponents for both teams. 


    Points: 0

    #64 by MikeWhiteGOAT // Dec 08, 2021 - 2:11pm

    In reply to by samueljames530

    Well the Cardinals were starting Colt McCoy at QB for 3 weeks so of course they haven't looked as good.

    And last week Andy Dalton was playing a fairly aggressive style of quarterbacking which leads to high yards per play. Also some of the offense was in garbage time. 

    There's nothing that implies the rams are as good as the Cardinals so it's a bit strange the Cards are only getting 2.5 at home

    Points: 0

    #68 by samueljames530 // Dec 08, 2021 - 2:39pm

    There's nothing that implies the rams are as good as the Cardinals

    DVOA sure seems to imply that...22.3% for AZ, vs 20.1% for the Rams. They are currently 5th and 6th respectively. That's total DVOA, not weighted, by the way. DAVE tells the same story: 17.5 for the Cards, 16.7 for the Rams....7th and 8th respectively. 

    Of course, none of this means the Cards can't dominate the Rams on Monday night. They absolutely could, and that would pretty much end the division race right then and there. But DVOA strongly suggests that both teams are pretty evenly matched. if the Rams split the series on Monday, the stretch run in the West is well worth watching. 




    Points: 0

    #70 by Lost Ti-Cats Fan // Dec 08, 2021 - 4:06pm

    In reply to by samueljames530

    Three big re-matches I'm looking at:

    ARI v LAR

    KC v LAC

    NE v BUF

    Any of those series that end up as a split instead of a sweep change the complexion of their division - and the perception of the teams involved - significantly.

    On the other hand, completing the sweep would solidify the SB-contender status of the winner significantly.

    Points: 0

    #46 by Aaron Schatz // Dec 08, 2021 - 10:43am

    I realized that I originally meant to put the Patriots week-to-week graph in this article and forgot that yesterday afternoon, so you'll now find it if you scroll up to the article. It's impressive!

    Points: 0

    #61 by nat // Dec 08, 2021 - 1:57pm

    Nice chart.

    BTW, did you have anything to say about NE’s surprisingly high DAVE. I would have thought the pre-season projection would have made their DAVE significantly lower than weighted DVOA, even this late in the season.

    Am I misunderstanding how weighted DVOA and pre-season projection are combined to make DAVE?

    Edit: probably explained by some teams having 13 games, and some 12. Teams with 13 games don’t really use pre-season projections at all, I think. For them DAVE is weighted DVOA.

    Points: 0

    #49 by Independent George // Dec 08, 2021 - 10:51am

    With a historically bad Houston rush offense this season, what does the median look like? Or is median a bad metric to use because of DVOA using per-play averages?

    Points: 0

    #73 by KaosTheory // Dec 10, 2021 - 1:09pm

    Has any team had a better run DVOA than pass? I would assume so but maybe not in decades. Who was the last team to do so?

    I guess I could look it up but does anyone happen to know the answer?

    Points: 0

    #74 by jheidelberg // Dec 10, 2021 - 7:20pm

    Every year there are teams that have better run DVOA than pass DVOA.  Basically all of these teams are bad offensive teams.  Basically a team has a better run DVOA when their pass DVOA is so pathetic.

    Here is this year's list:

    CAR (overall offense DVOA 31)

    DET  (overall offense DVOA 30)

    JAX   (overall offense DVOA 29)

    CHI   (overall offense DVOA 28)

    Yes even the Colts with Taylor and Titans with a lot of Henry have better pass DVOA than run DVOA.




    Points: 0

    #75 by mattyj94 // Dec 11, 2021 - 5:05am

    Has a team ranked bottom in Total DVOA ever made the playoffs (the Atlanta Falcons still have a fair chance this year)?

    Points: 0

    #76 by DarthGrizz // Dec 11, 2021 - 1:55pm

    What was going on in 2002? It appears that average Run Offense DVOA was positive at this point in the season, thus violating Aaron's statement that it's always negative and Passing is always positive. I'm guessing that reversed itself by the end of the season.

    Points: 0

    Save 10%
    & Support Aaron
    Support Football Outsiders' independent media and Aaron Schatz. Use promo code SCHATZ to save 10% on any FO+ membership and give half the cost of your membership to tip Aaron.