Arizona Cardinals Take DVOA Lead

The Arizona Cardinals knock the Buffalo Bills out of the top spot in our DVOA ratings this week. Most of the work on that was done by the Bills, playing a lousy game and losing to a bad Jacksonville team. But Arizona’s rating goes up as well after Colt McCoy’s surprisingly strong performance against San Francisco this week.

The biggest surprise from this week of surprises may be how few big moves there are up and down the DVOA rankings after all the Week 9 upsets. There are essentially four major moves this week:

• Buffalo moves down three spots to No. 4, although the difference between the Bills and the No. 3 Rams is a rounding error. Arizona is now on top with Tampa Bay moving up to No. 2.
• Tennessee moves up six spots to No. 14 after their fifth straight win, Sunday night against the Rams. We’ll get to the Titans in a little bit.
• Pittsburgh drops four spots to No. 19 despite a win over Chicago. It was the Bears who ended up with the better DVOA for that game, not the Steelers.
• Jacksonville moved up three spots to No. 27 after beating Buffalo.

The lack of movement in the top 10 means that San Francisco is still No. 9 at 3-5 despite losing to Colt McCoy and the Cardinals. That one certainly seems like a head-scratcher. Even though the Cardinals won 31-17, both teams ended up with very high DVOA ratings for the game: 37.5% for the Cardinals and 31.8% for the 49ers. The Cardinals won the turnover battle and outgained the 49ers 6.4 yards per play to 6.0 yards per play, but the DVOA system likes that the 49ers had a very good success rate on offense, 53% compared to 46% for the Cardinals. This is not related to garbage time. After they kicked a field goal to make it 31-17, the 49ers had two more drives and they didn’t have a crazy good success rate on those drives, just 50%.

A couple of things are going on with the 49ers’ rating. The first is one of the admitted weaknesses of the DVOA system. Opponent adjustments are based on an entire season’s worth of numbers. They are not tweaked when teams play backup quarterbacks. The system doesn’t know that the 49ers lost to Colt McCoy, so it’s giving the 49ers defense credit for playing against a good Arizona offense. This happens around the league, and we have to just understand that it means some of the numbers might be a little off. The Kansas City defense gets credit for facing the Packers based on their season-long stats, not Jordan Love’s numbers. Defenses that played against Russell Wilson now look a little worse than they should, and defenses that played Geno Smith look a little better. And so on. Over a 17-game season, these small differences tend to wash out. But when it comes to single-game ratings, they’re noticeable.

The other issue for the 49ers is the difference between their non-adjusted VOA and their adjusted DVOA. That’s caused by a couple of different things. First, San Francisco has played the fourth-toughest schedule in the league by average DVOA of opponent. Yes, as I just noted, they didn’t play the “real” No. 1 Arizona Cardinals, but the rest of their schedule has been difficult too. Second, the 49ers have had terrible fumble recovery luck this season. They’ve recovered only 4 of 11 fumbles on offense and only 3 of 15 fumbles on defense. They even lost a fumble on a kick return back in Week 4 against Seattle. Right now, only Denver has a worse fumble recovery rate. Those two things, schedule and fumble recovery luck, combine to make the 49ers look a lot worse than they really are. If we don’t account for those two things, the 49ers drop from ninth to 15th.

Incidentally, there’s another team that’s also suffered from a difficult schedule and bad fumble recovery luck. But this team is not as high in DVOA as you would probably expect. Yes, it’s time to talk about the Tennessee Titans again. This is the team that has most confounded DVOA this season, so they keep coming up. As noted above, the Titans did move up to 14th in DVOA this week. That is pretty low for a team that just won five straight games including four against teams that made last year’s postseason. The Titans have played the hardest schedule in the NFL so far this season. So the Titans would be even lower in our ratings if we did not do opponent adjustments! The Titans have also recovered only 31% of fumbles. So if we took out the opponent adjustments and only penalized offenses for fumbles lost, the Titans would drop from 14th to 20th.

As I noted earlier this week on Twitter, Tennessee’s low rating makes a lot more sense if you look at some of their other numbers instead of just admiring their admittedly impressive winning streak against good opponents. DVOA is measuring per-play efficiency, and the best conventional stat for measuring per-play efficiency is yards per play. The Titans have been outgained by their opponents this year, both in total and on a per-play basis. They are being outgained per play both on the ground and through the air. This is not the norm for a 7-2 team. I went looking through the history of the worst 7-2 teams by DVOA to find other teams that had lower yards per play than their opponents at this point in the season. I found these six teams, including the Titans:

It’s not the worst company in the world. Three of these teams significantly improved their DVOA in the second half of the season. The 2003 Panthers got hot in the playoffs and almost won the Super Bowl. The two Vikings teams listed here still had a negative yards per play differential in the second half of the season; the other three teams did not.

Tennessee is the ninth-worst 7-2 team in DVOA history. The 1999 Titans were the fourth-worst 7-2 team and they eventually went to the Super Bowl, so I don’t think Titans fans are that concerned about this fact. The 2015 Vikings team from the table above was the worst 7-2 team. The current Green Bay Packers, by the way, are the eighth-worst 7-2 team in DVOA history, with the Jordan Love asterisk.

A lot of what’s going on with the Titans is recency bias. The Titans were not very good at the start of the season, despite starting 2-2. On their current five-game winning streak, however, they have been outstanding. With current opponent adjustments, the Titans rank 27th in the league for Weeks 1-4. They rank fourth in the league for Weeks 5-9. It’s a huge change with improvement in all three phases.

Here’s the dilemma: I know we all want to look at Tennessee’s winning streak and say, “See, this is one of the top teams in the league, look at how well they’re playing over the last month.” But all our research shows that it is better to have a longer-term viewpoint on how good teams are, looking at a larger sample of games. Even our weighted DVOA formula that’s designed to look at how well teams will play going forward by lowering the weight of early games still considers 12 weeks of games before any games even drop below 50% weight. Those first four games also tell us something about how good the Titans are. And so, the Titans are not in the DVOA top 10 despite being 7-2 and the clear favorite for the No. 1 seed in the AFC. They currently get the top seed in 41% of our simulations, in part because they now go from the hardest schedule in the league so far to the easiest schedule remaining.

As long as I’ve got these numbers run, let’s look at some more teams for Weeks 1-4 compared to Weeks 5-9. Again, this is meant to be interesting more than predictive. After Tennessee, the team with the biggest jump over the past month is… their division rivals, the Indianapolis Colts. Their improvement has primarily been on the offensive side of the ball. The Colts come out with a better total DVOA over the last five weeks than even the Titans have had. Only Tampa Bay and Arizona have been better since Week 5 according to DVOA.

The biggest change in the other direction belongs to the Carolina Panthers. Their offense has completely crumbled since September, while their defense has been good instead of great like it was when they were winning games early in the year. Since Week 5, the Panthers have been the worst team in the league. Remember when the Panthers were No. 1 in DVOA after Week 2? This is why we use DAVE ratings that incorporate projections early in the season.

Another team that’s collapsed over the last five weeks is the Cincinnati Bengals. They had the solid win over the Baltimore Ravens in Week 7 but otherwise they have three losses, one to the Jets, and a win over Detroit. This week’s loss to Cleveland had an awful -59.2% DVOA. Only Carolina was worse in Week 9. DVOA never really thought the Cincinnati offense was that good to begin with, but the Bengals have struggled on defense and special teams over the past month.

Now, the change that you surely know about. The Kansas City offense has suddenly been completely neutered over the past five weeks. You may not realize that their defense has been much better at the same time (although, as noted above, it’s a bit artificially lowered because the Green Bay opponent adjustment is mostly based on Aaron Rodgers rather than Jordan Love). The Chiefs also moved into the top spot in special teams this week after the Ravens gave up a kick return touchdown to Minnesota.

Here are my New England Patriots, just because I find it interesting. The Patriots are a top-10 team over the last five games! Again, this is not meant to be predictive, but Bill Belichick’s teams do have a very long history of regularly struggling in September before they play better the rest of the season. I’ve written about it in the pas and if someone wants to go through all the old DVOA articles, you’ll find it. (This article from early last season just addresses Weeks 1-3 vs. the rest of the year. There should also be another article further in the past that looks at the overall trend for the entire season.)

One more team, because Mike Tomlin’s Pittsburgh Steelers have that same history of playing better at the end of the year than at the beginning of the year. And look at this, despite having the lower rating in last night’s win. The Steelers are riding a four-game winning streak since Week 5.

* * * * *

Football Outsiders playoff odds, snap counts, and the FO+ database are now all updated through Week 9. 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. FO+ is now on sale for just 99 cents per week if you buy an annual subscription, so make sure to check it out!

* * * * *

Here is the Football Outsiders Top 16 through nine 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. Opponent adjustments are currently at 90% strength and will hit full strength next week. 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 36% preseason forecast and 64% actual performance for teams with nine games played, and 45% preseason forecast and 55% actual performance for teams with eight games played. It is not currently adjusted for any backup quarterbacks.

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.