Panthers No. 1 in DVOA for First Time Ever
NFL Week 2 - Boosted by two impressive defensive performances against the Jets and Saints, the surprising Carolina Panthers take the top spot in the Football Outsiders DVOA ratings this week. This is first time in history that the Panthers have been on top of our ratings after any week, going back to their founding in 1995. The 1996 Panthers peaked at No. 2 in DVOA, while the 2015 Panthers peaked at No. 3.
Carolina leads a big pack of 2-0 teams at the top of the DVOA ratings. The Los Angeles Rams and Tampa Bay Buccaneers are second and third, both over 40% as they go into a game against each other this Sunday. Denver is just below that, followed by the three other NFC West teams: Arizona fifth, San Francisco sixth, and Seattle as the top 1-1 team ranked seventh.
Carolina's appearance on top of our table means there are now only two or three franchises that have never topped the DVOA ratings, depending on how you want to count the Cleveland Browns. The new Browns have never ranked any higher than fifth, going back to 1999, but the original Browns were on top of DVOA for a long stretch in 1989. The Jacksonville Jaguars and Detroit Lions have also never reached No. 1 in DVOA, both peaking at No. 2. The Jaguars were second for almost all of 1999. The Lions were second after Week 1 of 2014; they have never ranked in the top four any later than Week 4.
After two games, of course we have to ask how "for real" the Panthers are. We're not adjusting for opponents yet in DVOA, not until after Week 4, and while it's hard to say just how good or bad the New Orleans Saints are after two very different games, I think we all agree that the New York Jets are a bad team making that a less impressive win. How much can we learn from a high DVOA with no opponent adjustments and a two-game sample?
I went looking for teams that were in a similar situation to the Panthers since the expansion to eight divisions in 2002. These were teams that put up a DVOA over 40% through Week 2 (without opponent adjustments) coming off a season in which they ranked in the bottom half of the league. (Carolina ranked 21st last season.) From 2002 to 2020, there were 22 such teams. Fourteen of these 22 teams eventually made it to the playoffs, plus the 2015 Jets and 2017 Lions who each had winning records and would have been in the playoffs under the current seven-team format. Recent teams that started 2-0 after bad seasons but didn't eventually make the playoffs included the 2018 Dolphins (7-9) and the snakebit 2019 Cowboys (8-8). Two other Panthers teams were part of this list: the 2002 Panthers (who started 3-0 and then lost eight straight, eventually finishing 7-9) and the 2017 Panthers (who ended up with a wild card at 11-5).
If we want to look at a slightly bigger group of teams, we can lower the baseline from 40% DVOA after Week 2 to 35% DVOA. Now we've got teams we can compare to both the Panthers and the Broncos, who have 38.9% DVOA so far. This group has 17 of 28 teams making it to the playoffs.
While we're asking about the Panthers and similar teams, let's also focus on the Panthers and similar defenses. That's the unit that's powering the Panthers so far this season. Carolina has -52.0% defensive DVOA through two games. The Panthers rank No. 1 in both adjusted line yards and adjusted sack rate. This is quite the unexpected improvement, as the Panthers ranked 24th in defensive DVOA a year ago.
This time, there are fewer teams for comparison purposes. Only 12 teams since 1983 have put up defensive DVOA of -40% or better in Weeks 1-2, a year after they ranked in the bottom half of the league. Here, though, we have good news for the Panthers. Eight of these 12 teams ended up ranking sixth or higher in defensive DVOA at the end of the year. Only the 2003 49ers and 2018 Dolphins saw their defenses collapse after that early success. Here's a full table:
| Early Turnaround Defenses, 1984-2021 | |||||||
| Year | Team | DVOA Y-1 |
Rk | DVOA Week 2 |
Rk | Final DVOA |
Rk |
| 1988 | LARM | 4.5% | 19 | -52.6% | 1 | -7.7% | 6 |
| 1990 | CHI | 3.3% | 18 | -49.5% | 1 | -15.3% | 3 |
| 1996 | DEN | 11.3% | 27 | -54.2% | 1 | -15.2% | 3 |
| 1996 | GB | 6.5% | 22 | -50.9% | 2 | -19.6% | 1 |
| 2000 | NO | 5.4% | 21 | -43.1% | 2 | -7.5% | 12 |
| 2001 | CLE | 11.0% | 25 | -41.9% | 3 | -12.4% | 4 |
| 2002 | CAR | 2.9% | 23 | -42.4% | 2 | -13.2% | 3 |
| 2003 | SF | 2.6% | 21 | -40.4% | 1 | 3.0% | 20 |
| 2005 | PHI | -1.6% | 17 | -44.2% | 2 | -5.6% | 10 |
| 2018 | MIA | 10.7% | 28 | -40.0% | 1 | 7.1% | 26 |
| 2019 | NE | 1.4% | 19 | -61.1% | 1 | -25.2% | 1 |
| 2019 | SF | 6.5% | 24 | -41.7% | 2 | -20.3% | 2 |
| 2021 | CAR | 7.3% | 24 | -52.0% | 1 | -- | -- |
Why Aren't The Las Vegas Raiders Higher?
When I ran through the top teams in DVOA a few paragraphs above, you may have counted six different 2-0 teams. You also may know that there are currently seven 2-0 teams in the NFL. Where's the seventh team, the Las Vegas Raiders? You won't even find them on the partial table run below. You need to go to the full DVOA table, where you will find the Las Vegas Raiders ranked 19th at 2.3% DVOA, right above the winless Minnesota Vikings.
(The Raiders are below average in rank but above average in DVOA. Because the worst teams this year have been very, very bad, we actually have 21 teams right now that come out above the average of 0% and only 11 teams that are below 0%.)
What's going on? It's not that the Raiders have had a bad game. Their single-game rating is 1.4% for the Week 1 overtime win over Baltimore, and 3.0% for this week's win over Pittsburgh. This isn't a case of DVOA thinking the "wrong team won," either. The Raiders have the higher rating for Week 1, and the two teams are essentially tied for Week 2. It's just a case of DVOA feeling these games, especially against Pittsburgh, were closer than the final scores indicate.
For Week 1, I probably don't have to explain why DVOA is close in an overtime game. In case you are curious, the Ravens and Raiders both had 6.1 net yards per play. The DVOA ratings for Week 2 are more curious. The Raiders controlled most of this game. They outgained the Steelers, 6.6 to 5.9 yards per play, and they won the turnover battle by intercepting Ben Roethlisberger once. (Neither team lost a fumble, although the Raiders fumbled twice to Pittsburgh's once.) The Raiders also had the higher success rate on offensive plays, 44% to 37%. Yet the Steelers were slightly higher in offensive DVOA, 11.8% to 6.9%.
Getting down into the weeds on this one, the Raiders were the better team on first downs and the Steelers on second downs. The difference is on third downs, and the reason is that the Raiders had more baseline value on their third- and fourth-down plays than the Steelers did. Both teams converted these plays at about the same rate, but the Raiders were doing it from easier positions. For example, they had seven third/fourth downs with 4 or less to go. The Steelers had only three.
So what happens to 2-0 teams that DVOA isn't high on? Let's do the same thing we did with the Panthers, and go looking for past teams in a similar situation. Once again going back to 2002, there have been 23 different teams that were 2-0 despite a DVOA between -10% and 10% after Week 2. On average, those teams finished the season with 8.3 wins. They ranked 16th on average with 0.4% DVOA. Ten of these 23 teams made it to the playoffs. On one hand, the 2013 Saints and 2006 Colts proved early DVOA wrong by finishing with double-digit wins and top-six DVOA ratings. On the other hand, teams such as the 2007 Lions and 49ers sank to the bottom of the league. The 49ers that year started 2-0 but ranked 20th in DVOA. They ended the year 5-11 and dead last in DVOA.
The Raiders situation isn't very strange. In fact, last year we had it with four different teams including the Raiders themselves! Tennessee, Arizona, Chicago, and Las Vegas all ranked between 13th and 20th in DVOA last season despite 2-0 records. None of these teams saw their DVOA particularly improve over the rest of the year, but Tennessee and Chicago both made the postseason -- although for the Bears, they made it as the new seventh seed with an 8-8 record.
Will Defense Keep Kansas City Down?
One other team we want to talk about today, and that's the Kansas City Chiefs. I got a question over Twitter yesterday from Twitter user @PATRICKMAHOMES, who I'm going to guess is not actually Patrick Mahomes. He asked:
What are the worst defenses by DVOA that made the playoffs and how does the 2021 Weeks 1-2 Chiefs defense compare to them?
Let's go back into the database and find out, shall we? Kansas City's defense is at 33.7% DVOA after two weeks. The bad news is that no team has ever made the playoffs finishing the season with a defense this bad. The good news is that Kansas City isn't going to finish the season with a defense this bad because no team ever has finished a season with a defense this bad! The worst defense in DVOA history is the 1986 Tampa Bay Buccaneers at 26.7%.
What if we compare Kansas City to teams that started out the season with really bad defenses? I went and looked at teams with defensive DVOA of 30% or worse after Week 2, once again going back to the start of the eight-division setup in 2002. There are 45 different teams that qualify. Only 10 of these teams eventually made the playoffs, which are not good odds for the Chiefs, even though the playoff teams here include the Chiefs themselves in 2018, when they ranked dead last in defensive DVOA after two games.
But of course, this is not really a good list of teams comparable to the Chiefs, because what matters with the Chiefs is the offense. So instead, let's look for teams that have a defensive DVOA of 30% or worse but also have an offensive DVOA of 30% or better after Week 2. Now we've got a group with only seven other teams in it. Five of those teams made the playoffs, with the exceptions being the 2002 Chiefs who finished 8-8 and the 2010 Texans who crashed out to 6-10. Even if we include teams from before 2002, we only add in two other teams, one of which made the playoffs. This really isn't a good list of comparable squads, especially given that a few of them eventually lost their starting quarterbacks to injury (2000 Rams, 2013 Packers). But you do have the 2006 Colts, who won the Super Bowl.
| Great Offense with Terrible Defense Early, 1983-2021 | ||||||||||
| Year | Team | W-L | Off. DVOA |
Rk | Def. DVOA |
Rk | Final W-L |
Final DVOA |
Final Rk |
Playoffs |
| 1983 | GB | 1-1 | 52.5% | 1 | 31.3% | 27 | 8-8 | -11.5% | 24 | None |
| 2000 | STL | 2-0 | 35.3% | 2 | 35.9% | 31 | 10-6 | 15.3% | 10 | Wild Card |
| 2002 | KC | 1-1 | 40.0% | 2 | 35.1% | 31 | 8-8 | 23.4% | 4 | None |
| 2004 | MIN | 1-1 | 33.5% | 3 | 33.4% | 32 | 8-8 | -4.1% | 17 | Wild Card |
| 2006 | IND | 2-0 | 40.4% | 2 | 36.0% | 31 | 12-4 | 17.9% | 7 | Won Super Bowl |
| 2010 | HOU | 2-0 | 48.6% | 1 | 32.6% | 31 | 6-10 | 3.6% | 12 | None |
| 2013 | GB | 1-1 | 64.1% | 1 | 35.2% | 31 | 8-7-1 | -7.7% | 23 | Won Division |
| 2016 | OAK | 1-1 | 37.5% | 1 | 36.5% | 32 | 12-4 | 7.6% | 10 | Wild Card |
| 2018 | KC | 2-0 | 63.3% | 1 | 34.5% | 32 | 12-4 | 33.2% | 1 | Won Division |
| 2021 | KC | 1-1 | 36.1% | 2 | 33.7% | 31 | -- | -- | -- | -- |
In the long run, the Chiefs are still one of this year's top two or three contenders, despite some early defensive problems. Probably.
* * * * *
Football Outsiders playoff odds, snap counts, and the FO+ database are now all updated through Week 2.
A reminder that all our free stats pages, including DVOA and player position stats, now 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.
* * * * *
This is the Football Outsiders Top 18 through two 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.) Please note that there are no opponent adjustments in DVOA until after Week 4, which is why it is listed here as VOA.
OFFENSE and DEFENSE VOA are adjusted for 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.
DAVE is a formula which combines our preseason forecast with current DVOA to get a more accurate projection of how a team will play the rest of the season. DAVE is currently 85% preseason forecast and 15% actual performance.
To save people some time, please use the following format for all complaints:
<team> is clearly ranked <too high/too low> because <reason unrelated to DVOA>. <subjective ranking system> is way better than this. <unrelated team-supporting or -denigrating comment, preferably with poor spelling and/or chat-acceptable spelling>
| RK | TEAM | TOTAL DVOA |
LAST WEEK |
TOTAL DAVE |
RANK | W-L | OFF. VOA |
OFF. RANK |
DEF. VOA |
DEF. RANK |
S.T. DVOA |
S.T. RANK |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | CAR | 47.5% | 7 | -2.2% | 18 | 2-0 | 4.1% | 15 | -52.0% | 1 | -8.7% | 31 |
| 2 | LAR | 43.0% | 2 | 13.2% | 3 | 2-0 | 32.0% | 5 | -19.9% | 4 | -8.9% | 32 |
| 3 | TB | 41.1% | 13 | 23.8% | 1 | 2-0 | 25.9% | 6 | -8.8% | 12 | 6.4% | 4 |
| 4 | DEN | 38.9% | 9 | 8.4% | 9 | 2-0 | 32.9% | 4 | -12.9% | 10 | -6.9% | 29 |
| 5 | ARI | 32.3% | 4 | 4.0% | 16 | 2-0 | 11.4% | 10 | -16.2% | 6 | 4.7% | 8 |
| 6 | SF | 28.8% | 5 | 10.6% | 8 | 2-0 | 18.6% | 8 | -8.6% | 13 | 1.7% | 11 |
| 7 | SEA | 25.4% | 6 | 12.5% | 5 | 1-1 | 33.0% | 3 | 5.2% | 19 | -2.4% | 23 |
| 8 | PHI | 21.9% | 3 | -5.7% | 22 | 1-1 | 8.6% | 13 | -14.0% | 8 | -0.7% | 16 |
| 9 | CLE | 21.1% | 11 | 5.3% | 13 | 1-1 | 40.1% | 1 | 16.0% | 25 | -3.1% | 24 |
| 10 | WAS | 19.7% | 18 | 1.6% | 17 | 1-1 | 13.1% | 9 | 1.6% | 18 | 8.1% | 2 |
| 11 | NE | 18.5% | 20 | 6.7% | 10 | 1-1 | -4.8% | 21 | -24.0% | 3 | -0.7% | 17 |
| 12 | NO | 14.3% | 1 | 6.0% | 11 | 1-1 | -14.0% | 24 | -19.7% | 5 | 8.6% | 1 |
| 13 | BUF | 14.3% | 24 | 11.2% | 6 | 1-1 | -19.5% | 26 | -37.1% | 2 | -3.4% | 25 |
| 14 | DAL | 8.4% | 23 | 6.0% | 12 | 1-1 | 24.4% | 7 | 10.5% | 22 | -5.5% | 28 |
| 15 | PIT | 5.4% | 16 | 4.7% | 14 | 1-1 | 0.2% | 17 | -7.5% | 14 | -2.4% | 22 |
| 16 | KC | 5.0% | 22 | 18.4% | 2 | 1-1 | 36.1% | 2 | 33.7% | 31 | 2.6% | 10 |
| 17 | BAL | 5.0% | 21 | 13.1% | 4 | 1-1 | 11.0% | 11 | 13.1% | 23 | 7.1% | 3 |
| 18 | HOU | 5.0% | 8 | -18.3% | 29 | 1-1 | 2.4% | 16 | -6.7% | 15 | -4.2% | 26 |
Click here for the full table.
Comments
77 comments, Last at 24 Sep 2021, 1:05pm
#3 by Boots Day // Sep 21, 2021 - 3:38pm
The same would apply to the Chiefs defense. They've played two strong offensive teams in the Ravens and the Browns (who are No. 1 in offensive DVOA!), which means opponent adjustments will make their defense look better down the road.
#5 by Aaron Schatz // Sep 21, 2021 - 3:47pm
Are the Browns No. 1 in offense because they played the Chiefs, or are the Chiefs No. 31 in defense because they played the Browns? That's part of the problem with the early season.
I'm guessing the Raiders' ratings will go up for these games once we incorporate the opponent adjustments, but not enough to get them into the Top 10 or close to the other 2-0 teams through two weeks.
#39 by Pat // Sep 22, 2021 - 9:22am
Well, except Keyshawn Johnson. But, y'know. 3-time Pro Bowl WR, what the heck could he know. Oh, and Rex Ryan. Y'know. Dumb people like that.
If Carr isn't elite, he's sure as hell borderline elite. I could see putting him outside of the top 5, but definitely not out of the top 10.
#49 by Pat // Sep 22, 2021 - 11:04am
Nope. Just preemptively answering the possible response to Keyshawn and Ryan being "no one" to justify your previous statement of "absolutely no one."
You're welcome to your opinion of Derek Carr, but your statement that it's universal (and/or obvious, depending on how you take the statement) is easily demonstrated to be false.
#52 by Pat // Sep 22, 2021 - 11:55am
It's hyperbole that's not even close to reasonable. Of course someone reasonable could consider Carr elite: he's had a top-10 passer rating the past two years, top 10 ANY/A, top 10 DYAR, top 10 DVOA. The only QBs who have a combined DYAR higher than Carr for 2019-2020 (1989 DYAR) and still (effectively) in the league are Mahomes, Brady, Rodgers, and Wilson.
I'm not saying everyone should say he is. I'm saying that saying "absolutely no one" would is totally unreasonable.
You then gave your opinion italicizing definitely.
Yeah. Because it's far more unreasonable that someone would put Carr outside of the top 10 than someone calling him elite.
#58 by Pat // Sep 22, 2021 - 12:28pm
Flashbacks of people hyping up Fitz since 18 and disregarding the years prior.
Literally the only years you need to disregard are his rookie year and 2018, the year Gruden tanked the Raiders. This isn't Fitz where you have to disregard the guy's career or something. Outside of those 2 years, every year of his career's been Pro Bowl (*) worthy.
(*: by relaxed modern standards)
#75 by ImNewAroundThe… // Sep 23, 2021 - 7:04pm
But Carr is 30. He's not Josh Allen young. The hand waving of rookie years and a supposed tank year sure is a great pass! Carr is generally underrated but raiderfan seems quite serious in his notion that Carr is elite. Come on now...he's not and we know it. Try to twist it however you want like Pat but we all know what we see in the t12th ranked QB by the EPA+CPOE composite since 2015 (excluding his rookie year because...rookie years never count and you can't get any good info from, and that ranking excludes the retirements of Brees and Rivers and whatever you think of Watson even though he will return).
#48 by BeatThe9ers // Sep 22, 2021 - 10:31am
Is being in the top 10 out of 32 elite? Is being in the top 5? Is the elite stamp a point in time label, so whoever the top QBs are at any given time, they are the elite ones? Or is this in historical context, so that it is possible that in some seasons there are no 'elite' QB's? I favor the second.
In any case, Carr as elite seems a stretch. Like a huge stretch.
#53 by Pat // Sep 22, 2021 - 12:06pm
But if he's elite then that opens the floodgates of Tannehill, Baker, Kirk, Stafford, etc.
Except that Carr's been better than all of those guys over the past two years by DYAR. If you really want to set the bar to "elite" where there's no question, the only guys you have in there are Mahomes and Brady. Even Rodgers wouldn't hit there: there's no reason 2020 can't be an outlier rather than 2017-2018 being the outliers.
Brees is gone, Rivers is gone, Watson's gone, Prescott's coming off an injury, Allen's only got the one year, Lamar was way down as a passer last year. Wilson's better than Carr, sure, but he collapsed so hard last year that I absolutely wouldn't put him on the same shelf as Mahomes and Brady.
Honestly I have no problem with "elite" currently having a population of 'two', but I also totally recognize that that's a super strict definition, and I don't begrudge anyone else who wouldn't be that strict.
#59 by Pat // Sep 22, 2021 - 12:36pm
Or is this in historical context, so that it is possible that in some seasons there are no 'elite' QB's? I favor the second.
I would too (e.g. only Brady and Mahomes are elite) but the Hall of Fame pretty much proves that idea's in the minority, so I tend to follow the common definition, which is more "hanging at or near top 5 with at least one stellar year"-ish.
You could then say "but Carr hasn't had a stellar year" - to which the response is to point to this year. He's certainly playing at a top-end level right now, so saying he's "elite" doesn't bother me at all.
But hey, if we want to go with "far and away no questions asked," I'm fine with that too as long as people accept the whole "then Jim Kelly's a laughable Hall of Famer" corollary.
#74 by Pat // Sep 23, 2021 - 12:30pm
Pretty much the only metric he's not consistently top 10 is ESPN's QBR. EPA/play, EPA+CPOE composite, DVOA, DYAR, 538's ELO, etc. Literally all of them he was top 10 in 2020 (and except for the tank year pretty consistently there).
And obviously we've lost Brees, Rivers, and Watson from those metrics. I really don't understand the low perception of Carr by fans. His performance so far this year is completely in line with what we've seen from him before. Little better, sure, but not a lot at all.
#7 by Aaron Schatz // Sep 21, 2021 - 3:50pm
Good memory! At -129.1%, the 2019 Dolphins were the worst team ever through Week 2. The 2013 Jaguars and 2018 Bills were also worse through two games (2013 Jaguars were not much worse, at -78.2%). Also really bad through two games in recent years: 2011 Chiefs, who finished 7-9, and 2005 Vikings, who finished 9-7.
#11 by Pat // Sep 21, 2021 - 4:22pm
Here, though, we have good news for the Panthers. Eight of these 12 teams ended up ranking sixth or higher in defensive DVOA at the end of the year.
Well, "good" here is iffy. Obviously the '88 Rams, '96 Broncos and '96 Packers are bad comparisons, since they're much stronger offenses (all of them were top-5 DVOA/points scored), which leaves you with 3/9 teams making the playoffs.
So... yeah. Bodes well for Carolina's defense staying good, not so well for the Panthers making the playoffs.
#34 by David // Sep 22, 2021 - 4:16am
Bodes well for Carolina's defense staying good, not so well for the Panthers making the playoffs.
Sure, but making the playoffs is somewhat outside a team's control. If a team plays well, it must still rely on some number of other teams playing not quite as well in order to make the playoffs. In Carolina's case, early returns indicate that the division title might be a bit of a stretch, so they need to do better than 9 of the other 11 non-division winners. Given the strength in the West (through two games), it's doable, but difficult. I don't think the West is going to fill all three wild card slots, but I guess it's possible. Best hope for the Panthers is probably that the teams in the West all trade wins to reduce the win totals all around
Without looking, I'm guessing that schedule strength may help out the Panthers - they were bad last year, so should have some easier games on the slate.
#12 by DisplacedPackerFan // Sep 21, 2021 - 4:23pm
I can cook up a plausible causation between a bad defense and the QB getting injured but I have no idea if there is a correlation to attach it to. If you have a bad defense the QB will feel pressured to do more and place themselves in situations where they could get hurt more often. If you feel you have to make a play you are less likely to throw the ball away and hence take a sack you are more likely to try and scramble into a bad a situation, etc. So if teams with bad defense do have a higher QB injury rate I can come up with the BS as to why.
Mostly it came to mind because I wonder if 2013 Green Bay is actually a bad comparison to the 2021 Chiefs. Maybe having a great offense and a bad defense puts the QB at higher risk and that is exactly what KC could be facing. They are both teams that had/have young but established MVP caliber QB's who were/are great extending plays with their legs. Rodgers was in his 6th year as a starter Mahomes is in his 4th and 2009 - 2012 Rodgers rushing is very similar to 2018 - 2020 Mahomes rushing. Both teams had the two previous years where they had great offensives and mediocre at best defenses. Maybe 2013 GB is a cautionary tale about what could happen to a young elite QB when you keep pairing him with a poor to atrocious defense.
Then again there very likely isn't any correlation between crappy defense and higher QB injury rates. But still having watched the Packers since the early 80's and the Chiefs since the early 2010's (because of where I currently live). The McCarthy Packers and the Reid Chiefs have a lot of similarities. A LOT. The LeFleur Packers play a very different brand of highly efficient offense so they may have similar looking numbers but get there in different ways. The early 2010's Packers and the late 2010's early 2020's Chiefs actually got their similar numbers in similar fashion. That 2013 GB team which had nearly identical DVOA after week 2 to the 2018 KC team very likely would have been a 11-5 or 12-4 team without the Rodgers injury too. Basically if there is no higher chance of a Mahomes injury then the 2021 Chiefs are probably fine and will probably have a nicely positive DVOA by the end of the year too.
#63 by Dan_L // Sep 22, 2021 - 2:12pm
I think you're more likely to see the QB injury connection with Quarterbacks on the hot seat. Deep down do Andy Dalton and Tyrod Taylor think that if they make a bunch of "good throwaways" and "smart check downs", they will no longer be starting by week 9? (I think they are correct if they do suspect this). And so they would extend plays and do the type of risky things with their body necessary to keep a job. I think this also could describe most of Ryan Fitzpatrick and Josh McCown's careers, never having a safe starting job.
Just thought about this because in my impression, QB injuries in the past 20 years have been much more common from the mediocre-to-bad Quarterbacks than the good ones.
#13 by Eddo // Sep 21, 2021 - 4:37pm
"Getting down into the weeds on this one, the Raiders were the better team on first downs and the Steelers on second downs. The difference is on third downs, and the reason is that the Raiders had more baseline value on their third- and fourth-down plays than the Steelers did. Both teams converted these plays at about the same rate, but the Raiders were doing it from easier positions. For example, they had seven third/fourth downs with 4 or less to go. The Steelers had only three."
This is really interesting to me, as it implies that a team that goes 4 yards on three straight plays earns less value than one that goes 0, 0, 12, right?
#16 by theTDC // Sep 21, 2021 - 5:19pm
I remember pointing out last year that any offense that can legitimately run for 4 yards per carry every single play would be perfect. But the DVOA for said offense would be around 30%, despite them being absolutely perfect, and scoring on every drive minus time issues. The argument being that no real team can actually run for four yards per carry every single time, which is true.
DVOA tries to predict how things will be in the future, so sometimes you get those weird results that don’t make sense. I’d guess that what happened is the Raiders had lots of 2nd and 3 runs for 1 yard type plays, before converting the 3rd and 2. Whereas the Steelers might have some sacked for 7 yard loss on second down, before throwing for 10 yards twice in a row. It’s plausible that the latter works out to a better team in the long run.
#35 by David // Sep 22, 2021 - 4:23am
DVOA tries to predict how things will be in the future
Not exactly. DVOA is based around a few different ideas, but the key one that I think is germane to your comment is that DVOA is looking for sample size - because larger samples are more inherently demonstrative than small samples. The simplest measure of a team is the win-loss record, but it's a small sample size, so susceptible to the noise overwhelming the signal. It's what matters, though - it's what teams are actually trying to achieve. The largest sample size we can find is the play, but teams are NOT trying to win plays, they are trying to win games - so we need to find a way to link the success (or otherwise) of a play to the success of a team (winz).
Yes, IF a team could run for four yards every play - but it can't. There is noise in the signal. If there wasn't, then we wouldn't need DVOA
#38 by Eddo // Sep 22, 2021 - 8:48am
My guess now is that, for the first and second down plays, DVOA looks at two 4-yard gains and says "OK, that's not altogether atypical, those both get a little positive value and are fairly sustainable (though not perfectly so)". Then it looks at the two 0-yard gains by the other team and says "those are bad, but it's very rare for teams to chain together 0-yard plays, so we discount their negative impact."
#17 by RickD // Sep 21, 2021 - 5:26pm
It's much harder to convert 3rd and 12, so the points there are higher. DVOA is measuring difficulty of individual plays, not the soundness of a sequence of plays.
A team that tried to get 4 yards per play is far more likely to succeed than a team that tries to get 12 yards every third play.
#37 by Eddo // Sep 22, 2021 - 8:46am
Right, and I have no issue or confusion with the specific 12-yard third down conversion being worth more than the specific 4-yard one. But over the course of all three plays, it is surprising that 4-yard gains on first and second down don't count for more positive value than two 0-yard gains count as negative, because, as you say, the 4-4-4 team is more likely to succeed.
#65 by DisplacedPackerFan // Sep 22, 2021 - 3:05pm
It's actually 45% of the yardage on 1st down not 40% (https://www.footballoutsiders.com/info/glossary and scroll way down to success rate). A 4 yard gain on a first down is a binary failure. A 4-4-4 team is a fail-success-success team.
#44 by takeleavebelieve // Sep 22, 2021 - 9:47am
It’s likely because converting 3rd and 12 is a lot less common, and much higher stakes, than converting 3rd and 2. (Even if you fail on 3rd and 2, 4th and 2 is still manageable.)
Robo-offenses don’t exist in reality, so examples like that aren’t always the best way to understand a complex model that’s based on what actually did happen. I’d guess that the model just sees that one thing happens a lot less often than the other thing, so it favors the team that can do the harder thing.
#14 by RickD // Sep 21, 2021 - 5:15pm
This thing whereby the NFL pretends the Browns never skipped town to Baltimore is weird. No other franchise in any sports gets this kind of treatment. What happens if the Browns move again?
Since the Colts skipped town, should the Ravens take over their franchise records? After all, Unitas, Berry, and company have more of a connection to Baltimore than to Indy.
Imagine if MLB had done this with the Washington Senators. Would Walter Johnson's franchise records would have gone to Texas instead of Minnesota? Would the Nats have taken them over? But the Nats appear content to have the Expos' franchise records.
#18 by Eddo // Sep 21, 2021 - 5:27pm
The NBA did the same thing with the Charlotte franchise (note the gap from 2001-02 to 2004-05) when the Hornets moved to New Orleans. Charlotte got an expansion franchise a few years later that eventually got renamed back to Hornets.
#21 by Mike B. In Va // Sep 21, 2021 - 5:43pm
The Nats are *way* more complicated, since all the original, Johnson-era records would have gone to the Twins, since the original Senators/Nationals/Nats franchise moved there in 1960.
The expansion franchise that moved to Texas overlapped with the Expos. It's a mess.
Back to football - I thought the retention of the Browns' name and records by Cleveland was part of the deal for the move to Baltimore? Also, the Ravens didn't retain their records, but DID retain the Colts' marching band.
What were talking about again?
#77 by JimZipCode // Sep 24, 2021 - 1:05pm
Also, the Ravens didn't retain their records, but DID retain the Colts' marching band.
"Fight on, you Baltimore Colts!
Go in and strike like lightning bolts!"
https://www.youtube.com/watch?v=HOjuV9rFQJU
They changed the lyrics to match up with the current team name – shockingly late, I think that didn't happen til 2010 – but retained the tune, after a fan vote.
https://en.wikipedia.org/wiki/Baltimore%27s_Marching_Ravens#Background
#31 by Scott P. // Sep 21, 2021 - 10:17pm
I don't think it's weird since I conceive of a team name/identity to be something that is connected to the city. I find it bizarre that Oklahoma City pretends to have some claim on the Supersonics. For me, the actual personnel come and go in any case, the main constant is the city.
In fact, there were two versions of the Senators, one from 1901-1960, which became the Twins, and another from 1961 to 1971, which became the Rangers, and I find it natural to think of them as the same team.
In a parallel universe, sporting corporations (owner, GM, players, equipment) are hired as a package, and act essentially as mercenaries on contract to a city. Imagine if the Kraft-Belichick-Brady combo were only hired short-term by "New England," and could sign in free agency to any city that would outbid them. Could that be a 'stable' equilibrium, or would it inevitably turn to such corporations signing such long-term contracts that they became essentially fixed?
#36 by David // Sep 22, 2021 - 4:26am
No other franchise in any sports gets this kind of treatment
Lots of them do - see the examples in the thread.
Even more so, see all the examples that Bryan ran through on this site when going through the 'worst dynasties' series over the summer. This sort of thing was endemic in the early days of the NFL (through to the fifties, really)
I know this may come as a shock to you, but history didn't start in the early 80s :-)
#56 by Bryan Knowles // Sep 22, 2021 - 12:11pm
Sorting out franchise identities from the 1940s was a mess and a half.
My opinion on the Browns situation is that I'm OK with the OldBrowns and the NuBrowns being considered the same franchise, because that decision was made at the time the Browns moved to Baltimore. That was part of the agreement to get out of their obligations in Cleveland. The Browns never technically stopped existing; they just stopped playing football for three years and then came back with an entirely new roster, coaching staff and ownership. Sometimes it makes sense for treating the two eras differently -- it makes sense to consider the 1999 Browns an expansion team when looking at how new teams performed, for example -- but I'm OK with the history being stapled together, because there has never been a break in that history.
That's different then the history of Baltimore football. There have been three NFL teams in Baltimore -- the old Colts, the modern Colts and the Ravens -- and there was never any sort of consideration of the teams being the same thing at any point. The Irsays didn't say "don't worry, Baltimore; we'll have a new team here in 13 years, just be patient!", nor did the collapse of the 1950 Colts come with any sort of agreement that a new franchise would come back in 1953, even one that nicked the name. They're different situations because they were treated differently at the time.
That being said! I think it is fair and interesting to look at city records. So, Peyton Manning is the Colts' all-time leading passer, Joe Flacco is the Ravens' all-time leading passer, and Johnny Unitas is Baltimore's all-time leading passer. It's interesting to see how the Texans match up to the Oilers' old records, and to see which Los Angeles players come on top out of the motley collection of Raiders, Rams, Chargers and Buccaneers (no, not those ones), and maybe the Dons if you're counting AAFC stuff.
#67 by DisplacedPackerFan // Sep 22, 2021 - 3:45pm
Like LA New York and Chicago get an extra layer of fun with some of that too.
I do think it's a fun thing to look at, but even it has to have some definitions around the edges.
Do you have to take away the games the Packers played in Milwaukee to get the Green Bay numbers and give those numbers to Milwaukee? They stopped doing it in 95, but from 1933 to 1994 they played 1 to 4 games a year in Milwaukee. The Milwaukee Badgers didn't exist long and the stats from 1922 - 1926 would easily get swamped, but would Milwaukee get to lay claim to the stats when the Packers played there? It's still the majority of the franchise history where they played a significant amount of home games in a stadium that was over 100 miles away (like 118 by the most direct road route). That's in the range of Cleveland to Pittsburgh's, Washington to Philly, Indy to Cleveland. It's noticeably farther than Philly to Baltimore, Philly to New York, or the basically in the same city of Washington to Baltimore or San Fran to Oakland were. So no arguing that it's just a different stadium same city they are not close to each other.
It is clearly the same franchise so no weirdness there and they claimed a home city of Green Bay, but especially in the years they played 4 home games in Milwaukee could Milwaukee try and claim not only the Badgers but a share of the Packers? Had Green Bay only ever played 1-3 in Milwaukee I might not even thing about it, but when half your home games are there and the city has franchise history, if we're talking city records...
You could have fun with London too they get stats of all the "home teams" that played there over the years. But that is also a clearly defined situation and you would probably have to have a "permanent" London team at some point.
Again it's just a bit of silliness and pretty easy to go, no they are Green Bay city records because it was the Green Bay team and they always played at least half their home games there and stop trying to muddy the waters of a mostly silly argument. But if a city with multiple teams gets to sort through all those for the records Milwaukee might want in on that action.
#23 by Pat // Sep 21, 2021 - 6:36pm
I... have a lot of trouble believing the Raiders 3rd down argument. It's... pretty weird. I mean, not that it's there - sure, I can see that. But that it has any significant effect. That just seems super-odd.
First, there just aren't that many actual 3rd down conversions in the game. The teams both have 5. And sure, of those, a few of the Raiders were short conversions. 3rd and 1 for 2, 3rd and 4 for 4, 4th and 1 for 2. Sure.
And yeah, the Steelers had a 3rd and 8 for 17 and a 3rd and 8 for 41, and a 3rd and 10 for 25 and a TD. Cool.
But the Raiders had 3rd and 10 for 61 and a TD, plus a 3rd and 9 for 23. And the Steelers had a pick on 3rd and 10! On balance... I just can't see how those would make any significant difference. It feels like I could just line up all the 3rd down plays and cancel guys out. Like, the 17-yard play cancels the 23 yard play, the 3rd and 10 for 61 cancels the 3rd and 8 for 41, etc. And then you're left with a bunch of near-identical plays plus a Steelers TD and a pick. And I can't see how that would work out to "significant positive" for Steelers. I could buy a minor difference there, sure. But really not enough to even mention.
#25 by BiscuitsNGravy // Sep 21, 2021 - 6:47pm
The top six teams by total DVOA have faced a strength of schedule ranking all below average so far this year, and 4/6 have had strength of schedule rankings in the bottom 19% so far.
Are you concerned that this suggests that you are not weighting the results by opponent strength sufficiently?
#29 by poplar cove // Sep 21, 2021 - 9:23pm
I'm a huge Lions fan (no laughing) and probably every year for the last decade early in their 1st game of the season I'm watching thinking if they could just play their best game today maybe they will be number 1 ranked in DVOA.
Usually by the middle of the 2nd quarter that dream is long gone and just hoping they don't embarrass themselves.
That said couple years ago they absolutely dominated the Cardinals in Kingsbury and Murray's debut game in Arizona. Late into 3rd quarter I thought this is it they really have a chance to be ranked number 1 overall after week 1 only to watch them completely collapse in the 4th quarter and end up in a tie after overtime. This is the type of stuff you think about when the team you follow has just one playoff win in your lifetime
#71 by DIVISION // Sep 22, 2021 - 6:20pm
The Lions came in to Arizona and pulled out a victory they had no business winning. It was early season of what would be the first of several letdowns by ST (Zane Gonzalez) and head-scratching play-calls by Kingsbury.
If I had known the kicker would be that bad, I'd rather he'd just go for it on 4th and short every time. The Cards likely win against the Lions, Patriots at least.
You guys had Stafford and Megatron for a handful of years. You had your opportunity to win.
#30 by Lost Ti-Cats Fan // Sep 21, 2021 - 9:39pm
The #1 and #3 Defenses have each played NYJ, who somehow still rank #29 on Offence, not dead last. The #2 and #3 Defenses each played MIA, who rank #30 on Offence.
a) Opponent adjustments are likely to see some big swings in some of these numbers.
b) The AFC East is not the AFC West.
#33 by jheidelberg // Sep 22, 2021 - 12:17am
I find KC raw numbers interesting. There are no opponent adjustments. It is good to see the raw numbers. They will not get to adjust there defensive DVOA in the playoffs if they play the Ravens or Cleveland. They had better improve on defense, or just do their thing and have the offense and special teams be better on the positive side than the defense is on the negative side.
#42 by panthersnbraves // Sep 22, 2021 - 9:43am
We were expecting great things out of the Defense (although the Secondary taking such a big leap forward has to be on the coaching staff?). The Offense is alright - CMC is CMC. Darnold is pretty good, if he can stay within himself, and trust the receivers. TE group is improving, but the O Line is still a work in progress (and lost the LG for a few weeks...)
... and then there's Kicking...
ever since they let Butker get poached, that has been an issue.
#70 by ImNewAroundThe… // Sep 22, 2021 - 6:01pm
Unless you're committed to them. And most teams usually are (to a fault sometimes).
But yeah the Panthers defensive investments from last year better start paying off. Beating a rookie HC and QB isn't too surprising. And the Saints W was a toss up going into the year but of course the GB W they got changed things but probably shouldn't have that much.
#72 by DIVISION // Sep 22, 2021 - 6:31pm
I was impressed with them at all, however they forced Murray in to one of his worst games as a pro, last year.
They seem fundamentally solid, if unspectacular. They will go as far as the offense can take them. Their defense is legit.
It remains to be seen if Hassan Reddick is a one-year wonder or if he has staying power.