It's no surprise to see the New Orleans Saints on top of the Football Outsiders DVOA ratings through Week 12 of the 2020 season. We've had the Saints at No. 1 for a few weeks now, and they just demolished the Denver Broncos 31-3 on Sunday. But now we're stuck addressing the question of what to do about that very strange game.
Some readers have argued that we should simply not count Sunday's game in the same way that our 1987 stats do not count the strikebreaker games in Weeks 4-6 of that season. The problem is that it's a lot easier to eliminate three full weeks of games than it is to eliminate a single game during the season. Do we pretend that the Saints and Broncos had a bye week? That would require reprogramming our systems to deal with dual bye weeks, which is hard enough with the past 1993 data; at least that year, every team had two bye weeks. What about the fact that the Denver defense and New Orleans offense also took the field? Yes, New Orleans was starting a backup quarterback, but we've always counted every game in DVOA even when teams were stuck with second-string and third-string quarterbacks. The New Orleans offense and Denver defense and both special teams should be counted for Sunday's game. With the 1987 games, there were whole teams that were using completely different rosters with no regular players crossing the picket lines. In Sunday's game, we had one team that had to use one unprepared practice squad player.
So we're going with the idea that Sunday's game counts for DVOA, with the knowledge that we need to talk about it with a bit of a mental asterisk. That asterisk also includes the waves that emanate from Sunday's weirdness and hit other teams. The ratings for the Denver offense and New Orleans defense are affected, but so are the opponent adjustments for every team that has played against Denver or New Orleans this year, and the opponent adjustments for every team that has played a team that has played against Denver or New Orleans this year. For the most part, we're just going to have to live with it. However, just like we adjust our ratings for backup quarterbacks when we put them through the playoff odds simulator, so too we are going to adjust the Saints and Broncos for last week's game in the ratings we use to project the rest of the season.
The good news is that one single game has less of an impact on DVOA than you might think. Denver was already dead last in offensive DVOA before this week and they only dropped another two percentage points. Right now the Saints have a lead of almost 10 percentage points in overall DVOA, with 38.3% compared to Pittsburgh at 28.6% (with the Steelers yet to play this week, of course). The Saints are at -21.3% defensive DVOA, which is second in the league behind the Steelers. Take out this week's game, and the Saints go to -18.1% defensive DVOA, which would still be third in the league. And they would still have an overall lead of 6.5% over the Steelers.
It's after New Orleans and Pittsburgh that the weirdness sets in, because the No. 3 team is... the Tampa Bay Buccaneers, who drop just one spot from last week. They stay one spot ahead of the team that just beat them, the Kansas City Chiefs.
I will note here that the weighted DVOA ratings which lower the strength of early-season games have Kansas City ahead of Tampa Bay. Still, this is going to look strange to a lot of people. I keep writing about this. Most advanced metrics around the Internet have Kansas City as the No. 1 team. We don't. Meanwhile, the conventional wisdom is that the Tampa Bay Buccaneers are struggling and could even miss the playoffs. We have them still as one of the top teams in the league.
Some of what's going on here is adjustments for schedule strength. Kansas City's schedule ranks 27th so far this season. Tampa Bay's schedule ranks third, behind only Chicago and Carolina. In our non-adjusted VOA numbers, the Chiefs are still behind New Orleans and Pittsburgh but clearly lead Tampa Bay, 27.4% to 22.8%.
I also want to point out that DVOA is definitely in agreement with other analytic methods that the Kansas City passing game is something special. Despite the adjustment for the easy schedule, and despite heavy adjustments for the fact that the offensive levels around the league are so high this year, Kansas City is on pace to rank as one of the best passing games we have ever measured. However, I do want to take issue with all the people who have been saying that what Kansas City is doing this year is "unprecendented." One of the teams that has a higher passing DVOA than Kansas City is... Kansas City themselves, just two years ago. That Chiefs team two years ago averaged 0.35 yards per play more than this year's team. That's not specifically because of rushing, either. The 2018 Chiefs averaged 0.40 net yards per pass more than the 2020 Chiefs, despite playing in a less impressive offensive environment league-wide.
Here are the best pass offenses over a full season, and where the current Chiefs would stand among them. These are version 7.3 of DVOA, so scrambles are included as pass plays, though that doesn't matter much with most of these teams. I put the two Chiefs teams in red:
|BEST PASS OFFENSE DVOA, 1985-2020|
|* through Week 12|
|** 12 games, no strikebreaker games|
As an aside, it's worth noting that most of the quarterbacks on these teams won the MVP award, making Mahomes the clear frontrunner for this season. The exceptions were Mark Rypien of the 1991 Redskins (lost to Thurman Thomas); Philip Rivers of the 2009 Chargers (lost to Peyton Manning); and Joe Montana of the 1987 49ers (lost to John Elway). Montana did win the MVP in 1989, and that 1989 49ers team is 13th on this list.
Boy, am I looking forward to finishing up 1984 to figure out where Dan Marino and the Dolphins land on this list. But I digress... Clearly the Kansas City pass game is not keeping the Chiefs out of the No. 1 spot in DVOA. Everything else is. It's not that the Chiefs are terrible in the other parts of the game, but they aren't good.
The Chiefs are just 15th in rushing DVOA, though that has improved in recent weeks. Kansas City had -22.9% DVOA on rushing plays through Week 4, but that's gone to 4.3% DVOA since Week 5, which ranks fourth in that period. That's one reason the Chiefs are better in weighted DVOA than in full-season DVOA.
Kansas City ranks 17th in defensive DVOA. Unlike the running game, that hasn't improved in recent weeks. Kansas City's two best defensive games of the year were Weeks 3 and 4.
The Chiefs are also 23rd in special teams DVOA. This one is a surprise because the great Dave Toub is still the coordinator and the Chiefs had one of our best preseason projections for special teams. However, a lot of this special teams rating is one bad game way back in Week 3. The Chiefs' special teams are probably fine the rest of the year.
Anyway, if we put all the numbers together, we end up with the Kansas City Chiefs as the No. 4 team in DVOA for the season. There's definitely something to the argument that Kansas City should be considered stronger than its total DVOA because pass offense is the most predictable and consistent part of the game. I've played around with different combinations of offense, defense, and special teams to try to account for this in DVOA and I don't have anything that I'm comfortable with yet, but it does matter. It matters that the Chiefs are the best at the most predictive thing, while the three teams ahead of them are the top three defenses and defenses are less consistent than offenses.
But there's also something to the fact that Kansas City has had problems putting opponents away. In Week 9, they let a mediocre Carolina team get within two points in the fourth quarter. The Panthers had a shot at a game-winning field goal and just couldn't get the ball close enough. Two weeks ago, they needed a last-minute drive to beat the Raiders. This week, they let the Buccaneers score two touchdowns on them in the fourth quarter. Did Kansas City's win expectancy stay above 75% despite these Tampa Bay scores? Yes, but these scores were not in "garbage time." The Chiefs were trying to stop the Bucs from moving the ball. They were trying to move the ball themselves on offense. They didn't do a good job of either in the final 22 minutes. What Tampa Bay did in the second half does give us data about these teams going forward.
So then, what about Tampa Bay? As noted above, some of what's going on with Tampa Bay is schedule. Take away the schedule adjustments and the Bucs drop from sixth to eighth on offense and from third to fifth on defense. There's also an issue with yardage that gets counted in DVOA but not in traditional stats. Add in that yardage -- stuff like defensive pass interference and yards lost from aborted snaps and intentional grounding -- and Tampa Bay's yards per play on offense goes from 5.81 (12th) to 6.02 (ninth), a lot closer to their unadjusted offensive rating. Tom Brady leads the NFL with 16 DPI calls for 297 yards.
The football commentariat is writing off the Bucs because they've lost three out of four, but two of those losses came by just a field goal. Yes, the second Saints game in Week 9 was a debacle. A close loss to the Bears with a lot of penalties was a little embarassing. But the Bucs are not some kind of disaster right now. Take out the Saints game, and the Bucs have just been good on offense in the last three weeks as they were in the first eight weeks! The defense has gotten worse, but has still played above average over the last three weeks once we adjust for strong opponents.
|Tampa Bay by Week, 2020|
|Weeks||OFF DVOA||DEF DVOA|
The Bucs have a schedule full of games where both teams had positive DVOA after opponent adjustments. In fact, the Saints faceplant in Week 9 is the only game all year where Tampa Bay has a negative DVOA. They aren't the only ones, by the way. The other top four teams also have positive DVOA for almost every game:
- New Orleans as of now has no games with negative DVOA. We wrote earlier this season about the surprising results in their losses, particularly the Week 2 Raiders game in which DVOA rated the Saints better than the Raiders. The only game that comes close to zero is the Week 7 win over Carolina, 27-24, which is currently at 1.0% DVOA. The last four New Orleans wins all have single-game ratings over 50%.
- Undefeated Pittsburgh does have two games under zero, but those ratings are in single digits: -2.0% for the Week 5 win over Philadelphia and -5.3% for the Week 9 win over Dallas.
- Right now, Kansas City's only negative rating is not for the Week 5 loss to Las Vegas but rather, surprisingly, for the Week 1 win over Houston. That's also in single digits at -8.1%.
Tampa Bay's schedule gets much easier after the bye week, ranking 28th the rest of the way. My best guess is that the Bucs will finish the season either 4-0 or 3-1 and get into the playoffs, and we'll hear about how Bruce Arians and Tom Brady solved their disagreements over the offensive scheme and the Bucs are coming into the playoffs hot and in reality nothing will have really changed except that they aren't going to have a game like Week 9 again.
Let's hit the rest of this week's list. Green Bay moves up four spots to fifth with the big win over Chicago. Tennessee's big win over Indianapolis moves the Titans up one to 12th and the Colts down three to eighth. The Seahawks also move down three spots to ninth after a "close" win over a poor Eagles team. Yes, the final Hail Mary touchdown counts in our numbers. The game was close enough that it still counts at full strength. At some point, I do plan on doing more work on adjusting DVOA for current win expectancy which may help with things like this. Seattle's defensive DVOA would go from 8.5% to 7.7% without that play, so they would still be in ninth place. I'm assuming that a COVID-ravaged Baltimore squad is going to drop after Wednesday which would move the Colts and Seahawks back up a spot, but we'll have to see.
Other big moves this week include Houston moving up three spaces to 17th, Detroit moving down five spaces to 24th, and Las Vegas moving down four spaces to 18th.
There's one other team I want to talk about today, which is the 8-3 Cleveland Browns. You've probably heard that the Browns are not very good for an 8-3 team. DVOA agrees and has them all the way down at No. 22. The Browns have actually been outscored by opponents this year despite being 8-3. Their schedule ranks 29th. Cleveland has negative DVOA for this week's win over Jacksonville as well as last week's win over Philadelphia due to opponent adjustments. DVOA also thinks they were outplayed by Indianapolis despite winning back in Week 5. As a result of all of this, the Browns have the worst DVOA ever for an 8-3 team, by a substantial amount:
|WORST 8-3 TEAMS BY DVOA, 1985-2020|
|* Actually 8-3-1|
|** No strikebreaker games|
The good news for Cleveland fans is that all of the teams on this table made the playoffs despite their low DVOA ratings. The 2003 Panthers even made it to the Super Bowl. The latest playoff odds report has Cleveland making the postseason 67.3% of the time. But this low rating is why we have Cleveland (Week 16) as the most likely win that would prevent the New York Jets from a 0-16 season. Cleveland is still favored, of course, but it's the most likely Jets win.
Speaking of the Jets and 0-16, let's finish up the week with updated odds for all the extreme possibilities we've been tracking the last couple weeks. The Steelers odds here include a backup quarterback penalty for Baltimore in the simulation of Wednesday's game.
- Pittsburgh goes 16-0: 18.1% (up from 15.7%)
- New York Jets go 0-16: 32.8% (up from 20.9%)
- Entire NFC East has a losing record: 84.2% (up from 80.4%)
- Entire NFC East is 6-9-1 or worse: 41.5% (up from 36.5%)
- Entire NFC East is 5-10-1 or worse: 4.4% (up from 4.1%)
* * * * *
Football Outsiders playoff odds, snap counts, and the FO+ database are now all updated through Week 12. We'll be updating everything again on Thursday morning once the Pittsburgh-Baltimore game is (hopefully) in the books.
Big news! The entire DVOA database has now been updated to the new version 7.3 of DVOA that counts scrambles as pass plays and fixes adjustments for defenses playing indoors. All the FO+ pages as well as the free pages going all the way back to 1985 now represent updated numbers. The only exception, as of now, is that DVOA "as of a Specific Week" is still the older version for the years 1985-1998, as it takes a lot of time to re-run every single week of past years. We're gradually getting to them, though!
Also, 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, and picks against the spread.
* * * * *
Here is the Football Outsiders Top 16 through 12 weeks of 2020, 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 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>
For the full table, including DAVE, schedule strength, and non-adjusted VOA, visit the Football Outsiders DVOA database.