Blowout Moves Cowboys to DVOA Top Spot

NFL Week 16 – We’ve got a new team on top of the Football Outsiders DVOA ratings this week. With their huge 56-14 blowout of Washington, the Dallas Cowboys climbed into the top spot for both full-season and weighted DVOA, passing the Tampa Bay Buccaneers. The Cowboys are now our No. 1 defense but are strong all around: eighth on offense and fourth on special teams.

The rest of the top 10 remarkably stays the same despite some big one-sided results on the schedule this week. However, there’s more movement for the weighted DVOA ratings that lower the importance of early games. This is the first week where a game has completely disappeared from the weighted DVOA formula. Week 1 games are now gone.

Buffalo hops up from fifth to second after beating New England, and the Patriots drop from first to fourth. Tampa Bay falls one spot from second to third. Green Bay hops to eighth, past San Francisco, and Philadelphia moves up to 10th, passing the Chargers and the Cardinals. Arizona is now down to 14th in weighted DVOA after losing this week to Indianapolis.

But the strangest movement of the week belongs to Kansas City. Last week, the Chiefs ranked sixth in total DVOA and fourth in weighted DVOA. This week, after beating Pittsburgh by a very comfortable 36-10 margin, the Chiefs are… sixth in total DVOA and fifth in weighted DVOA. Kansas City’s total DVOA actually dropped this week, from 18.5% to 17.6%. This is the team that is pretty much regarded as the best in the NFL by acclamation. The Chiefs are everybody’s Super Bowl favorite — including ours. They won the game easily and coasted through the second half. How on earth did they drop in DVOA this week? Why did the other blowouts in Week 16 end up so much better in DVOA? Even with opponent adjustments, the Bengals, Cowboys, Eagles, Buccaneers, and Dolphins all ended up with single-game ratings above 50%. Kansas City’s single-game rating was 15.9%.

This is not the entire explanation, but it looks like once again we’re going to run into a problem with the Kansas City Chiefs taking their foot off the gas. You may remember that this was a very big deal last year when DVOA did not have Kansas City rated as highly as conventional wisdom. The argument was that while most teams are just trying to kill clock with big late leads, the Chiefs take it to another level and really don’t try very hard on offense or defense with a late lead. As long as they win, people said, the Chiefs don’t care how many yards they gain or give up late.

So what happened in this week’s game is that the Chiefs were horrible in the fourth quarter. They brought in backup quarterback Chad Henne, who proceeded to botch two snaps and then just handed off for a bunch of runs that went nowhere. Meanwhile, the defense had a strip-sack of Ben Roethlisberger early in the fourth quarter but then allowed a meaningless 85-yard touchdown drive as well as a 54-yard drive to end the game that ended without a score.

Some of these plays are already downweighted in DVOA. Offensive DVOA only gives half strength to any play in the fourth quarter where the scoring margin is over 21 points, unless the offense is in the red zone where prevent defense doesn’t mean as much. Defensive DVOA doesn’t have an adjustment for late-game leads; perhaps I need to test again to see if it makes sense to add one. But even with the lower weight, the Chiefs get dragged down by what they did in fourth down, because they were so much worse than the other teams who were finishing up blowout victories this week. Look at how the Chiefs’ DVOA compares to those other blowout teams in Week 16. I think it will stand out to you:

Apologies for being too lazy to split out special teams by quarter. Kansas City’s offense was excellent for the first three quarters. I’m a little surprised that their defense doesn’t get a better rating for the first three quarters but Pittsburgh had a reasonable (i.e. close to league average) 43% success rate through the third quarter. Kansas City still gets a negative defensive DVOA because the Steelers only gained 4.2 yards per play and had an interception.

Then we get to the fourth quarter. If you squint, you will probably notice that Kansas City is the only one of these six teams that allowed a positive DVOA in the fourth quarter of their blowout. The other five teams continued to play good defense, although in some cases not as good as during the first three quarters. But you don’t have to squint to see the number that I assume really jumped out at you: Kansas City’s offensive DVOA in the fourth quarter. Holy cow, that is bad. In fact, that is the third-worst fourth-quarter offensive DVOA put up by any team this year. Here are the five worst offensive fourth quarters:

The worst fourth quarter of the year was the one where the Atlanta Falcons threw four interceptions to the Patriots. As an aside, the last two of those interceptions were thrown by backup quarterbacks, and probably play a small role in DVOA seemingly overrating the Patriots a bit this season.

The overall point here is that not everybody shuts down in the fourth quarter quite like Kansas City did this week. Their performance really stands out, even when compared to other teams running out the clock in blowouts. Even without the two blown snaps because Chad Henne was not used to taking the ball from Creed Humphrey, the Chiefs still would have come close to making this list, with -141.4% offensive DVOA if the two fumbles are removed. Does that tell us something about the Chiefs going forward? Our past research suggests that it does, which is why we downweight these plays to half strength instead of removing them entirely. We did some research on DVOA compared to win probability this past offseason and it suggested leaving the formula as we had it instead of further downweighting garbage time. We’ll probably test this again in this coming offseason, because it’s always one of the big questions about DVOA: is there a better way for us to deal with the way some teams shut it down when they have a late lead while other teams do not?

At the same time, this doesn’t have quite as much of an effect on Kansas City’s DVOA as you might expect. Kansas City’s full-season DVOA right now is 13.8% on offense and 0.3% on defense. If we remove the fourth quarter of this game, that improves to 15.4% on offense and -0.5% on defense. This would be enough to move the Chiefs past the Patriots, but they would still rank just fifth in DVOA for the season (fourth in weighted DVOA). And a number of other metrics also agree that the Chiefs are not the best team this year. Tucker Boynton’s adjusted net EPA has Kansas City sixth. Pro Football Reference’s SRS has them fifth. ESPN’s FPI has them third, but still behind Buffalo in the AFC. FiveThirtyEight’s ELO is the exception, as it currently has the Chiefs at No. 1. (It’s also an exception for the Green Bay Packers, putting them at No. 2.)

I went back into the numbers because I felt confident that I was going to show today that the Chiefs really do treat a late lead differently from other teams. But when it comes to big late leads like the one they had on Sunday, it turns out they really don’t. Over the last three years, the Chiefs’ offensive DVOA with a lead of 21 or more points in the fourth quarter is about average. Last year, for example, the Chiefs had 23.5% offensive DVOA in a limited sample with a lead of 21+ in the fourth quarter. The place where the Chiefs really shut down on offense last year was with a smaller lead, two scores instead of three. Overall, the Chiefs play worse defense when they are defending a big late lead, but even that isn’t too significant. This year, for example, Kansas City’s defensive DVOA with a 21+ lead in the fourth quarter is 6.5%, which isn’t really that far off from their total defensive DVOA for the entire season.

For now, that fourth quarter still counts (albeit at half strength) and Kansas City is sixth in DVOA for the year, fifth in weighted DVOA right now. All I can do is present the numbers the way the system currently spits them out. Then I try to analyze them and I let you, the readers, do the same. We’re stuck here with the same problem we had a year ago. Last year, the football commentariat was pretty much united in its opinion that the Chiefs were not just Super Bowl favorites but were basically inevitable. It was bad enough that I got convinced to go against my numbers in the playoffs even though DVOA suggested that Tampa Bay was the better team.  Obviously, there were extenuating circumstances, the offensive line injuries, but nonetheless, the Kansas City Chiefs were not inevitable. They aren’t this year either. They are properly the favorites right now because they are still very good and they have the inside track to the No. 1 seed. But DVOA suggests that the Buffalo Bills still have an excellent chance of beating Kansas City in a hypothetical AFC Championship, and our numbers would also favor Dallas or Tampa Bay over Kansas City if they meet in the Super Bowl.

As of now, that is. The numbers will change between now and then. And of course, this is 2021, so any decision about which team is favored in any game has to weigh which players are out on the COVID reserve list.

* * * * *

And now, let’s switch from the best teams in the league to the worst. At least, to the team we think is the worst. The Atlanta Falcons are still dead last in DVOA even though they keep pulling off close wins against other bad teams. This week, they managed to come within nine yards of losing to Detroit’s backup quarterback before intercepting Tim Boyle in the final minute. The Falcons are 7-8 and somehow still in the playoff race. It will be completely hilarious if the worst team in DVOA ends up making the playoffs. Right now, the worst DVOA to ever make the playoffs belonged to the 2010 Seattle Seahawks, who won the NFC West at 7-9 with -24.3% DVOA. (They won the Beastquake game over New Orleans in the first round.) The Falcons would have to play really, really well in their final two games to beat that.

It’s very likely that the Falcons will end the season as the worst team to ever have how ever many wins they actually end up with. So I thought it would be fun to run a table of the worst teams to ever finish with every win total, not including ties. Of course, the 17-game season now messes this up, but here goes:

A lot of these teams showed that DVOA is more consistent than win-loss records by having fewer wins both the season before and the season after the year listed above. Not all of them — hello, 2020 Miami Dolphins — but many of them. For example, the 1992 Colts went 9-7 sandwiched between a 1-15 season and a 4-12 season. The 1991 Patriots went 6-10 sandwiched between a 1-15 season and a 2-14 season. The 1986 Chiefs went 10-6 sandwiched between a 6-10 season and a 4-11 season. And the 1996 Cardinals went 7-9 sandwiched between two 4-12 seasons.

* * * * *

Football Outsiders playoff odds, snap counts, and the FO+ database are now all updated through Week 16. 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 16 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>

Click here for the full table.