Super Bowl LI Preview

Super Bowl LI Preview
Super Bowl LI Preview
Photo: USA Today Sports Images

by Aaron Schatz

Super Bowl LI is a matchup of two similar offensive juggernauts, but the question that will probably decide the game is this: "When do we know a defense has truly improved?"

Nobody can possibly question that Atlanta and New England are two powerful offenses. Atlanta was No. 1 in points scored, with New England No. 2. Atlanta was No. 1 in offensive DVOA, with New England No. 2. Only seven teams in NFL history scored more points in a season than the 2016 Atlanta Falcons, and as good as the Falcons looked, they were actually better than that. Only Washington played a more difficult schedule of opposing defenses than Atlanta did this season. However, the New England offense is also better than the raw stats would indicate. Raw stats, of course, include four games early in the season where Tom Brady was suspended and the Patriots started backup quarterbacks Jimmy Garoppolo and Jacoby Brissett.

(By the way, New England also played a harder-than-average schedule of opposing defenses, although that is slanted a bit towards those first four games. From Week 5 to Week 17, New England's 12 opponents had an average defensive DVOA of exactly 0.0%.)

The context for these defenses is just the opposite. You'll hear a lot on the Super Bowl LI broadcast about how the Patriots had the league's No. 1 defense based on points allowed. Unless you are brand new to Football Outsiders with this preview, you know that this is a misleading statistic. The Patriots did not allow the fewest points in the league because they had the best defense. That rank comes from a number of other factors. The Patriots faced the easiest schedule of opposing offenses in the NFL this year. In fact, their schedule was even easier than it looks at first glance because they got to face backup quarterbacks such as Landry Jones and Matt Moore. The Patriots also allowed fewer points because their defense constantly got to defend long fields. The average opponent drive began at the 24.9-yard line, the best figure in the NFL. That's a product of the offense's ability to avoid turnovers and strong special teams on both kickoffs and punts. When we look at play-by-play efficiency using Football Outsiders' DVOA ratings, the Patriots come out with an average defense, ranked 16th in the NFL. Football Outsiders now has drive stats for 20 seasons, and the Patriots are the first team to lead the league in points per drive but rank outside the top 10 in DVOA.

However, the problem with the narrative of the overrated Patriots defense is that the season is not static. Players change, schemes change, injuries happen, and teams improve or decline over the course of the year. The Patriots defense has improved very steadily over the course of the season. In Weeks 1-9, the Patriots defense ranked 18th in DVOA. In Weeks 10-17, it ranked eighth. In the postseason, they've been even better, even after the significant downward adjustment that comes from facing The Brockening.

[ad placeholder 3]

Just as with the offenses, there's a lot of similarity between these defenses as well. The Falcons also played an easier-than-average schedule of opposing offenses, 24th in the NFL. (The schedule strength is similar for both teams in part because four of their five common opponents had much better defense than offense: Arizona, Denver, Los Angeles, and Seattle.) The Falcons also benefited from strong starting field position on defense (26.1-yard line average, third in the NFL) because of a lack of turnovers and strong punting (although, unlike the Patriots, not strong kickoffs). And the Falcons have also improved significantly on defense over the last few weeks. In their first 12 games, the Falcons limited opponents to 21 or fewer points only twice. In their last six games, including the playoffs, the Falcons have limited every single opponent to 21 points or less except New Orleans in Week 17.

Ah, but that's the difference here: gradual improvement vs. sudden improvement. The Patriots defense has improved gradually throughout the year, against both good and bad opponents. The Falcons defense improved suddenly over the last few weeks, first against weak opponents in December and then against stronger opponents in Atlanta's two playoff games. Here's a table I ran first a week and a half ago. If we break down the season into five-week segments, it is clear that the Patriots defense has been better than the Falcons defense over both the short term and the long term.

Atlanta/New England Defensive DVOA by Week, 2016
Weeks Atlanta
Defensive DVOA
New England
Defensive DVOA
Weeks 1-5 11.4% 7.7%
Weeks 6-10 10.6% 0.8%
Weeks 11-15 16.0% -5.9%
Weeks 16-20 -11.6% -20.4%

The defensive improvement here brings up two questions. First, where have these defensive improvements come from? What is each team actually doing better on defense? We'll explore the answers for both defenses in the appropriate sections below. Second, how "real" is this improvement? We've talked about this in past postseasons, but counterintuitive though it seems, weighting our ratings for recent games doesn't necessarily predict results any better than looking at a full season. Do we judge these defenses based on just these past few weeks, or the whole season? The problem for Atlanta is that either way, the Patriots have the better defense. But if the longer-term view is more accurate -- even if we're only looking at a couple months, not the entire season back to September -- the Patriots' defensive advantage is much greater. For the Falcons to win Super Bowl LI, they need to show that their defense truly is improved while their offense shows that the Patriots' defensive improvements are entirely based on taking advantage of bad opponents.

For those who may be unfamiliar with the Football Outsiders stats, they are explained at the bottom of the page. Scroll down or click this link. Game charting data appears courtesy Sports Info Solutions, unless noted. Please remember that all stats represent regular season only, except for weighted DVOA and anything else specifically noted.

Patriots on Offense
DVOA 21.1% (2) 8.1% (27)
WEI DVOA 24.3% (3) 4.2% (22)
PASS 50.5% (2) 11.6% (19)
RUSH -3.4% (17) 2.5% (29)
RED ZONE 23.0% (8) 23.1% (29)

Falcons on Offense
DVOA 25.3% (1) -1.5% (16)
WEI DVOA 28.1% (1) -11.0% (6)
PASS 53.0% (1) 13.9% (23)
RUSH 1.7% (7) -23.7% (4)
RED ZONE 3.5% (14) -6.5% (12)

Special Teams
DVOA 2.7% (7) 2.4% (8)
NE kickoff +11.8 (2) -1.5 (29)
ATL kickoff -2.8 (22) -3.0 (26)
NE punts +12.2 (3) +3.5 (9)
ATL punts -7.7 (29) +2.2 (12)
FG/XP +0.1 (17) +10.9 (2)

All readers can click here for in-game discussion on our message boards. If you have FO Premium, you can click here to see all the matchup of DVOA splits for this game.

Check out other Football Outsiders content previewing Super Bowl LI:


Instead of talking about where the Falcons have improved over the past few weeks, let's start by talking about where they haven't improved: stopping the run. It's even worse than it was earlier in the season. The Falcons ranked dead last in run defense DVOA from Weeks 10-17. Even if we included their strong performance against the Seattle and Green Bay running games the last two weeks, they would rank 30th. We're not the first media outlet to suggest that the Patriots hammer away at Atlanta's undersized defense with the running game, but in case you missed it: the Patriots should hammer away at Atlanta's undersized defense with the running game, especially on first down.

The Falcons get killed by the run on first down and the pass the rest of the time. On first down, the Falcons ranked 31st in DVOA against the run. They allowed an average of 4.8 yards per carry compared to the NFL average of 4.3 yards (and the New England offense's average of just 4.0 yards). Against the pass on first down, the Falcons ranked fifth in DVOA. Then on both second and third downs, the Falcons defense ranked 28th against the pass. And while the Patriots were just 18th in DVOA passing on second downs, they led the league in passing DVOA on third downs even if we include the four games without Brady. Overall on third downs, including both runs and passes, the Patriots ranked just 12th on third-and-short but first on third-and-medium (3-6 yards) and second on third-and-long (7-plus yards).

Adjusted line yards show an advantage for the Patriots running in pretty much any direction, but especially up the middle (seventh vs. Atlanta at 27th) and behind right tackle (12th vs. Atlanta 32nd). The Patriots also have a big advantage on outside runs to the right (third vs. Atlanta 24th), but they are less likely to take advantage of this. Only 3.7 percent of their runs this season were marked "right end," which ranked them 31st in frequency.

Of course, the Patriots are still the Patriots, and the NFL is still the NFL of 2017, not the NFL of 1977. There's going to be a lot of passing and Tom Brady distributing the ball to all his receivers. It's the Falcons pass defense which has improved, going from 20.8% DVOA in Weeks 1-13 (tied for 24th in the NFL) to -11.1% DVOA in Weeks 14-20 (tied for ninth). So what has gotten better?

The easy answer would be "more pass pressure," except that doesn't actually seem to be the answer. Based on Sports Info Solutions charting, the Falcons' pressure rate has barely changed between the first and second halves of the regular season. However, pressure does seem to be part of the formula behind the playoff wins over Seattle and Green Bay. During the whole season, the Falcons only blitzed on 14 percent of pass plays, which ranked 28th in the NFL in frequency. In the past two games, that's gone up by either almost double or more than double.

The either/or here is because the ESPN Stats & Info charters and the Sports Info Solutions charters do not agree about just how much the Falcons' use of blitzes has gone up in the postseason. As Bill Barnwell pointed out in his Super Bowl preview over on ESPN, their charters list the Falcons blitzing 36 percent of the time in the last two games. SIS charters have the Falcons blitzing just 24 percent of the time. However, both sets of charting data agree that a) the Falcons blitzed less than almost any other team during the regular season; b) the Falcons are blitzing more in the playoffs, even if they don't agree on how much more; and c) the Falcons have brought more pressure in the playoffs, in large part because those blitzes are getting home. The Falcons ranked 22nd in defensive pressure rate during the regular season, but would rank in the top ten based on their pressure rate in the playoffs.

But is the strategy the Falcons used in the last two games really translateable against the Patriots? The Seahawks offensive line is terrible, and the Packers offensive line was torn apart by injuries to the point where they had to bring defensive lineman LeTroy Guion in to play guard by the end of the game. In addition, Brady is consistently stellar against the blitz. The classic "blueprint" to slow down the Patriots' offense is to get a pass rush with just four pass-rushers, and the Falcons did rank 13th in pressure rate when rushing four during the regular season. However, most of their pass rush comes off the edges, not up the middle against the weaker members of the Patriots offensive line, left guard Joe Thuney and center David Andrews. Interior linemen had only five sacks this season (three for Grady Jarrett, two for Ra'Shede Hageman) and Jarrett was the only interior lineman with more than 10 hurries.

Passing by No. Pass-Rushers, 2016
2-3 Pass-Rushers 4 Pass-Rushers 5+ Pass-Rushers
Freq Yd/Pass Rank Suc Rate Rank Freq Yd/Pass Rank Suc Rate Rank Freq Yd/Pass Rank Suc Rate Rank
Tom Brady 16% 8.9 5 56% 12 63% 7.8 3 57% 6 19% 9.6 4 55% 11
Atlanta D 9% 8.1 29 42% 32 77% 6.2 9 52% 12 14% 6.7 14 43% 32

How about Atlanta's pass coverage. Has that improved since midseason? Well, it depends who we are talking about. There's been a huge improvement from starting cornerback Robert Alford. Through Week 13, Alford allowed 7.5 yards per pass with a 51 percent success rate. Since Week 14, Alford has allowed 5.2 yards per pass with a 62 percent success rate. The other outside cornerback position is a push statistically, but theoretically worse since midseason. Desmond Trufant had surprisingly mediocre charting stats before his season-ending injury nine weeks into the season, allowing 6.3 yards per pass but with a success rate of just 48 percent. His replacement, Jalen Collins, has allowed more yards per pass (7.1) but with a higher success rate (55 percent).

However, the real problem here is in the slot. Nickelback Brian Poole is a great story, getting a regular job in the NFL as an undrafted rookie. But a good story does not necessarily equal a better-than-average nickelback. Poole had only 6.6 yards allowed per pass this season, because he was covering shorter routes, but also had a 48 percent success rate. He's a big reason why the Falcons ranked 29th in DVOA against "other wide receivers" compared to ranking in the top ten against No. 1 and No. 2 receivers. And Poole has gotten worse, not better, in the late part of the season. If we limit our charting to just Weeks 14-20, Poole comes out with 9.3 yards allowed per pass with a dismal 38 percent success rate. Tom Brady finds the open receiver, and a lot of the time the open receiver is going to be whoever is being covered by Brian Poole.

The open receiver probably won't be Martellus Bennett very often, a disappointment for those of us hoping for the most entertaining Super Bowl MVP possible. The Falcons ranked 11th in DVOA against tight ends this season, and this is one of the main improvements in recent weeks, with tight end catch rates against Atlanta dropping from 68 percent over the first half of the season to 52 percent since Week 12.

That leaves running backs in the passing game, and as Scott Kacsmar discussed in a piece on Thursday, this is going to be a major problem for the Falcons. Atlanta ranked 26th in DVOA against running backs in the passing game, and in opponent-adjusted numbers they allowed the highest number of passes (9.1 per game) and receiving yards (53.5 per game) to running backs. The Patriots love to use their running backs as receivers. They'll send them on routes out of the backfield or sometimes line them up wide. Sometimes they'll put James White and Dion Lewis on the field at the same time, with one in the backfield and one on the line of scrimmage. White ranked third in receiving DYAR among running backs this season. Lewis actually had disappointing receiving numbers after his return from last year's ACL injury, but he was a very dangerous receiver in 2015 and had a receiving touchdown against Houston three weeks ago.

However, White and Lewis will probably do their damage by beating the Falcons' young linebackers up the field, not with yards after the catch on screen passes and dumpoffs. Yards after the catch is another place the Falcons pass defense has clearly improved since midseason. Atlanta's broken tackle rate has improved only from 24th in Weeks 1-13 to 20th in Weeks 14-20, but the Falcons have gone from allowing an average of 5.0 yards after the catch through Week 13 to 4.2 average yards after the catch in Weeks 14-20, which ranks third in that time period. This improvement is important because the Patriots led the NFL with 6.3 average yards after the catch. They were just 19th with 7.5 average YAC from running backs, but that's generally a product of the fact that so many of Tom Brady's passes to running backs are planned routes rather than dumpoffs on third-and-long. (The Patriots also don't use a lot of running back screens these days.) However, the Patriots led the league in YAC by tight ends -- and would have done so even without Rob Gronkowski playing half the season -- and ranked fourth in YAC by wide receivers. The Patriots also ranked fifth in broken tackles and ninth in percentage of plays with broken tackles. (As you'll see in a moment, this is all another similarity between the two offenses.)

One last note on the Patriots offense: though they did not use play-action as often or with as much success as Atlanta -- we'll get to that below -- the Patriots did use play-action an above-average 19 percent of the time and gained 9.5 yards per play, fourth in the NFL. However, the Falcons were actually somewhat reasonable against play-action, allowing less than a yard more against play fakes on average (6.3 vs. 7.1).


Back in Week 12, the Patriots finished out November with a close call against the New York Jets. They allowed Ryan Fitzpatrick to throw for 269 yards with only one sack and no interceptions. That game is a good separating point for when the Patriots' defense really turned things around. Their four best performances by single-game DVOA have come in the seven games since.

The Patriots have been strong against the run all year, so that hasn't really changed since Week 13. Their run defense before and after that point is basically the same. But their pass defense has improved significantly, which means their overall defense has improved in every way possible. Split the field into 20-yard segments, and in each one, the Patriots' defensive DVOA after Week 13 is better than its DVOA in Weeks 1-12. But a look at the Patriots' defense by down does the best job of conveying just how extensive the improvement has been:

New England Defensive DVOA by Down, 2016
First Down Second Down Third/Fourth Down
Weeks Yd/Play DVOA Rank Yd/Play DVOA Rank Yd/Play DVOA Rank
Weeks 1-12 5.33 2.5% 16 5.89 6.4% 21 5.34 9.4% 19
Weeks 13-20 4.97 -17.5% 5 3.92 -14.9% 6 5.57 -27.4% 8

Another fun trick is to separate the season one week later, the same as we did for Atlanta. Remember above where we wrote that the Falcons went from being tied for 24th in pass defense for Weeks 1-13 to being tied for ninth in pass defense for Weeks 14-20? The team they were tied with in both time periods was the New England Patriots, which is a crazy coincidence.

So, what has improved for the Patriots pass defense since midseason? It's clearly coverage rather than pressure. The Patriots ranked 21st in defensive pressure rate through Week 12, and that actually has dropped to 30th in Weeks 13-20. However, coverage has improved with Eric Rowe replacing Logan Ryan as the outside cornerback opposite Malcolm Butler and Ryan thriving with his move to the slot to replace Justin Coleman. And Butler himself has been much better in recent weeks. Obviously, the opposing quarterbacks play a role here, since these charting stats are not yet adjusted for opponent. But Butler has gone from 8.2 yards per pass with a 52 percent success rate in Weeks 1-12 to 5.1 yards per pas with a 69 percent success rate in Weeks 13-20.

[ad placeholder 4]

There's been a lot of discussion of Malcolm Butler against Julio Jones because the Patriots depended on Butler to shut down Antonio Brown one-on-one in the AFC Championship Game. But there's a good chance Butler is not going to be covering Jones. Historically, the Patriots have a habit of using their top cornerback on the other team's No. 2 receiver while doubling the No. 1 receiver with the second corner and a safety. There's a reason they would do this with Julio Jones but not with Antonio Brown: size. Malcolm Butler and Antonio Brown are both less than six feet tall and less than 200 pounds. Julio Jones is 6-foot-4 and 220 pounds. That's a better matchup for the 6-foot-1 Rowe. The Patriots didn't play a lot of Julio-like big No. 1 receivers this season, but when they did, Butler did not follow them around in man coverage. They did not use Butler on A.J. Green in Week 6, Terrelle Pryor in Week 5, or Demaryius Thomas in Week 15. In addition, the idea of Butler trying to shut down Mohamed Sanu while Rowe and Devin McCourty try to control Jones matches up with what the Patriots will see on film, that Atlanta's worst offensive performances this year tended to come when opponents shut down Sanu and Taylor Gabriel because nobody can fully shut down Julio Jones.

In the past, the Patriots were known for shutting down opposing No. 1 receivers and allowing a lot of yardage to No. 2 receivers. That didn't happen this year. The Patriots actually had a better DVOA against No. 2s than No. 1s. Other than ranking eighth against No. 2s, they were around average in the categories we track as "defense vs. types of receivers." The 'pass defense by direction" table shows that the Patriots were particularly susceptible to deep passes on the offensive right side of the field, but a look at coverage charting from Sports Info Solutions shows most of these completions were against zone coverage, not particularly poor man coverage by any specific Patriots cornerback.

The Falcons, as you probably know, were spectacular at spreading the ball around this year. Matt Ryan set an NFL record throwing touchdown passes to 13 different receivers. There wasn't a single Falcons player with at least 10 pass targets who had a below-average receiving DVOA. The Falcons had the top two receivers in DVOA with at least 50 targets (Gabriel and Jones), the top running back in receiving DVOA with at least 25 targets (Tevin Coleman), and two of the top three tight ends in receiving DVOA with at least 10 targets (Austin Hooper and Levine Toilolo). It's really important to cover everyone on third downs, because that's when Ryan does the most to spread the ball around. Mohamed Sanu, not Julio Jones, was actually Ryan's top target on third downs this season. Jones was targeted on 30 percent of Falcons passes on first down, 25 percent on second down, and just 17 percent on third or fourth down.

The Patriots love to go zone and force passes short of the sticks on third down, and the other thing that's improved for the Patriots defense over the course of the season is their tackling. Since we first began tracking broken tackles, the Patriots almost always rank near the top of the league for fewest missed tackles on defense. That didn't happen this year, but the problem is concentrated in the first part of the season. In Weeks 1-12, the Patriots had 83 broken tackles, with broken tackles on 10.6 percent of plays (28th). Since Week 13, the Patriots have had 43 broken tackles, coming on 7.8 percent of plays (10th). Good tackling is important against a Falcons offense that ranked 11th in broken tackles and seventh in the percentage of plays with broken tackles.

Oddly, the Patriots' overall tackling has improved even though the player who led the team in broken tackles, Kyle Van Noy, didn't show up until a midseason trade. Furthermore, even the tackling problems early in the year didn't stop the Patriots from preventing big gains after the catch. The Patriots ranked No. 1 in the NFL allowing just 4.1 average yards per catch, and that didn't go down at midseason. It's been that way all year. It's also not limited to any particular type of pass: the Pats ranked third in YAC against running backs (6.3), third against tight ends (3.3), and second against wide receivers (3.2).

This is one of the elements that makes the matchup of the Patriots defense and the Falcons offense so great, as the Falcons ranked second in the NFL with 6.2 average yards after the catch -- trailing only the Patriots. The Falcons rank in the top four at all three positions.

The Falcons pride themselves on the quality of their running backs, but they are going to beat the Patriots by throwing the ball. It makes sense given the strength of the New England run defense and it makes sense given the strength of the Atlanta passing game. The Falcons were fifth in DVOA running on first downs, but slightly below average on second and third downs. And even despite this, it makes sense to throw on first down. There's an absolutely insane stat about their passing game which you may or may not have seen previously. Including the postseason, Atlanta has averaged 10.04 net yards per pass on first-down pass plays. That essentially means that when they throw on first down, the average Falcons play will move the chains. The Patriots have improved on first down over the course of this year, but not enough to fully counter that kind of power. Even the improved Patriots defense ranks just 11th against the pass on first down, allowing 6.05 net yards per pass. (They rank sixth against the run on first down both in Weeks 1-11 and in Weeks 12-20.)

When the Falcons do run, their best bet may be to go right up the middle; that's the one place they have an advantage according to this year's adjusted line yards stats. Atlanta ranked fourth in ALY on runs up the middle, while the Patriots defense is 13th. Of course, that advantage goes away if Alex Mack's leg injury prevents him from playing or has him playing at less than 100 percent, and it's a bit worrying that Mack has spent this week's practices coaching up his backup Ben Garland just in case. The Falcons had a lot of long, highlight-worthy runs this season, but that's tougher than usual against the Patriots. Atlanta ranked third in open-field yards per carry this season (i.e. gains more than 10 yards past the line of scrimmage), but the Patriots defense ranked No. 1 preventing such gains.

One of the interesting aspects of Kyle Shanahan's scheme in Atlanta is the way it goes against certain NFL trends. For a decade now, Football Outsiders has been spreading the gospel of shotgun almost always being more efficient than running a traditional offense with the quarterback under center. So it's a bit astonishing to see that the Falcons had the year's best offense while ranking dead last in use of the shotgun, just 40 percent of plays. The Patriots actually faced shotgun on 73 percent of plays, more than any other defense in the NFL. But the bad news for the Falcons: in part because of the strength of the run defense, the Patriots ranked fourth against plays with the quarterback under center compared to 19th in DVOA against the shotgun.

The way to take advantage of that run defense is with lots of play-action. Oddly, opponents used play-action against the Patriots on only 13 percent of passes, near the bottom of hte league. However, the Falcons used play-action on a league-leading 26 percent of plays and averaged 10.4 yards per play, second by a small margin behind Washington. And the Patriots defense gave up some pretty big plays against play-action, allowing 8.0 yards per pass (20th in the NFL) compared to just 6.1 yards per pass otherwise.

Despite the weakness of the Patriots pass rush, they will get to Ryan on occasion. The Falcons' offensive line is better at run-blocking than pass-blocking, and Ryan can be sacked. The Falcons ranked 23rd in adjusted sack rate, as Ryan took 13 more sacks than the three Patriots quarterbacks combined, and they ranked 18th in pressure rate allowed. And the best way to get to Matt Ryan seems to be blitzing him, something the Patriots defense has done more of in recent weeks.

Passing by No. Pass-Rushers, 2016
4 Pass-Rushers 5+ Pass-Rushers
Freq Yd/Pass Rank Suc Rate Rank Freq Yd/Pass Rank Suc Rate Rank
Matt Ryan 62% 9.5 1 59% 1 27% 7.6 13 54% 15
New England defense 54% 6.2 12 52% 10 20% 6.9 17 57% 9

The Patriots send three pass-rushers and drop eight into coverage more than any other defense in the league, so why are those numbers missing from the table above? Unfortunately, there is a huge disagreement between SIS charting and ESPN charting regarding how often teams sent only three pass-rushers at Ryan and how good he was on those plays. SIS lists Ryan with 10.4 net yards per pass on these plays, the best figure in the league. ESPN Stats & Info lists Ryan with 6.1 net yards per pass on these plays, which is near the bottom of the league. The numbers are so different that I'm not sure what to do without the time to review every single play involved. I'm not sure if this is a good strategy against Matt Ryan or a bad one.

However, both sources of charting agree that Ryan is only a little bit above average against the blitz, compared to his otherworldly numbers this year when opponents rushed the standard four men. Ryan's performance against defensive back blitzes is particularly interesting. It's a high-risk, high-reward strategy, as Ryan averaged 8.7 yards per pass (10th) but 5.5 net yards per pass (25th) according to ESPN Stats & Info. Ryan was sacked 9 times in 52 plays, but also threw six passes of 20 yards or more, plus another two passes of 19 yards and three shorter touchdowns. The Patriots allowed 6.8 net yards per pass with a DB blitz, about the league average. However, they started to run a particular cornerback blitz at midseason when Logan Ryan was moved into the slot, and Ryan has gotten a sack and four hurries off that play. It wouldn't be surprising to see one Ryan go after the other on a couple of slot corner blitzes Sunday night.

There's one other thing that seems to drop the Falcons offense from "historically awesome" to "sort of above average," and that's getting them into the red zone. It's a problem for the Falcons that could definitely be the difference in a close, one-score game. Although the Falcons led the NFL in offensive DVOA, they ranked only 14th in the red zone. Fitting the legend of Bill Belichick defenses as "bend but don't break," the Patriots defense ranked better in the red zone than overall, and it's been even better in the red zone in recent weeks. Running the ball in the red zone is going to be a particular difficulty for the Falcons, as they ranked 19th on red-zone runs while the Patriots defense was second. The Falcons are better passing the ball, where they rank 10th in the red zone. The Patriots were 24th against passes in the red zone for the season -- but if we only look at the defense since Week 13, they would rank eighth against red-zone passes.


For the regular season, Atlanta and New England were neck-and-neck on special teams, finishing seventh and eighth, respectively. However, one rating missing from the boxes at the top of this preview is weighted DVOA for special teams. If we lower the importance of games from the first two months of the season, the gap between the teams becomes much larger, with New England at 5.5% DVOA and the Falcons at only 0.3%.

For the Falcons, the difference between total special teams and weighted special teams can probably be ignored. The only aspect of special teams that has clearly declined is kickoff coverage, and a lot of that is related to Devin Hester's last hurrah in the Seattle playoff game. However, the Patriots' improvement on special teams is absolutely real. After early struggles, they have returned to their usual spot near the top of the league for a few reasons. First, whether it was a back issue or just the yips, Stephen Gostkowski really struggled in the first two months of the season. Since New England's bye, however, he has gone 21-for-23 on field goals and 30-for-31 on extra points. Second, punt coverage has been better in the second half of the season. Third, the Patriots stopped putting Cyrus Jones on the field to muff punt returns.

Gostkowski's return to a high level since midseason means both of these teams sport a trustworthy kicker, although the Falcons are more likely to depend on Matt Bryant for long kicks over 50 yards. (He was 7-for-8 this season, while Gostkowski was 2-for-4.) Gostkowski was also fantastic this year at pinning opponents deep with kickoffs that came down around the goal line, and the Patriots coverage prevents good returns. If they need a touchback, Gostkowski can give them that too. On the other hand, Matt Bosher's kickoffs for Atlanta were below average in net yardage despite more than half of them coming indoors. Eric Weems is pretty average returning kicks for Atlanta. The Patriots had a revolving door but used Dion Lewis in the first two playoff games. He had that touchdown against Houston, but also a fumble that the Texans recovered.

Both the Ryan Allen and Bosher had strong years as punters but the Patriots punt coverage is better than the Falcons punt coverage. However, the Falcons got better punt returns than the Patriots. Weems is better on punt returns than kickoff returns, with six returns of 15 or more yards this season. Again, the Patriots had a revolving door here, but it will likely be Julian Edelman returning punts in the Super Bowl.


We've been blessed over the past few years with a lot of close Super Bowls. This one is no exception. There's a reason the line is only three points. The Atlanta Falcons have a really, really good offense, and the Patriots can only hope to slow it down. They won't stop it completely. It's pretty easy to imagine the Falcons pulling off an upset with a ton of points and a couple of serendipitous bounces of the ball on defense or special teams.

And yet, the Patriots were clearly the better team this year. If we ignore the four games without Tom Brady on the field, these offenses are about equal. So are the special teams. Every bit of evidence we have says that the New England defense is better than the Atlanta defense. They've been better over the entire season. They've improved much more gradually. Atlanta's suddenly strong performance in the last few weeks doesn't seem to be tied to any concrete changes. It seems more likely to represent a general swing in performance, the random variation that's just always present in sports. One big game for the Falcons defense against a Packers offense riddled with injuries stands out a lot more when you are talking about just four or five good games. The Patriots have had some good fortune in opponent injuries, such as Le'Veon Bell, but that stands out less when we're talking about steady improvement over three months of games.

The Patriots are more likely to slow down Matt Ryan than the Falcons are to shut down Tom Brady. There are just too many matchups that specifically favor the Patriots. The Patriots have better cornerbacks. The Patriots have better run-stoppers in the front seven. The Falcons have that one awesome pass-rusher in Vic Beasley, but the Patriots have better pass protection than the Falcons. Tom Brady is going to get time to throw. He's going to read the coverage and find open receivers, especially in the slot. He's going to take advantage of a mismatch on White or Lewis for a couple huge gains. If the Patriots take a lead into the fourth quarter, LeGarrette Blount should be able to run out the clock with ease against a porous Falcons front.

Falcons fans are dreaming of a win that looks like the Giants beating the Patriots in Super Bowl XLII or Super Bowl XLVI. But this Atlanta team doesn't look anything like those Giants teams that took down the Patriots juggernaut. I'm expecting a game that looks more like the Patriots' third win, when they beat the Eagles in Super Bowl XXXIX. The Patriots clearly had control of that game the entire time, but they never really pulled away. There was always a chance of the Eagles completing the comeback with one last touchdown. I'm expecting something similar, only with more back-and-forth scoring this time. Also, Matt Ryan won't need eight minutes to run a two-minute drill if the Falcons get one last possession in the fourth quarter. But after facing a 60-yard, clock-killing drive full of Blount Force Trauma, the Falcons won't have eight minutes to waste.


DVOA (Defense-adjusted Value Over Average) breaks down each play of the season and compares it to the NFL average based on situation and opponent. You'll find it explained further here. Since DVOA measures ability to score, a negative DVOA indicates a better defense and worse offense, and a positive DVOA indicates a better offense and worse defense.

Team DVOA numbers incorporate all plays; since passing is generally more efficient than rushing, the average for passing is actually above 0% while the average for rushing is below 0%.

SPECIAL TEAMS numbers are different; they represent value in points of extra field position gained compared to NFL average. Field goal rating represents points scored compared to average kicker at same distances. All special teams numbers are adjusted by weather and altitude; the total is then translated into DVOA so it can be compared to offense and defense. Those numbers are explained here.

Each team is listed with DVOA for offense and defense, total along with rush and pass, and rank among the 32 teams in parentheses. (If the DVOA values are difficult to understand, it is easy to just look at the ranks.) We also list red zone DVOA and WEIGHTED DVOA (WEI DVOA), which is based on a formula which drops the value of games early in the season to get a better idea of how teams are playing now (explained here).

Each team also gets two charts showing their performance this year, game-by-game, according to offensive and defensive DVOA. In addition to a line showing each game, another line shows the team's trend for the season, using a rolling average of the last five games.


36 comments, Last at 06 Feb 2017, 1:16am

1 Re: Super Bowl LI Preview

My gut keeps telling me this isn't going to be quite the shootout everyone expects. It just seems like the Super Bowls that are supposed to be shootouts usually aren't.

So, I took a quick look at past over/under totals. This game has the highest O/U ever, at 59, but there have been 9 previous Super Bowls with an O/U of at least 50. The under has prevailed in 6 of those games, including the last 4 of them.

I don't know the explanation, although I suspect it might have something to do with extra adrenaline leading to sloppy offensive execution, but that's just a wild guess.

Anyway, with that in mind, I expect a close, moderately high scoring game:

Patriots 28, Falcons 24

2 Re: Super Bowl LI Preview

Or maybe the whole game will be like the first quarter of the SEA@ATL playoff game, with the teams ripping off 7+ minute drive after drive. So high scoring efficiency but a (relatively) low score because of reduced number of drives.

4 Re: Super Bowl LI Preview

If I was betting the over that would be what scared me.
To really put up points you like turnover shortened fields and long yardage plays.
The Pats defense isn't the steel curtain but they do limit big plays while the offense limits turnovers.

And on offense if the Patriots use the type of superbowl game plans that the Parcells/Belichick Giants used against the K-Gun or the Pats used against the GSOT Rams then that will involve deliberately killing clock.
That's less of a sure thing though since the current Pats are a lot less scared of shootouts then those teams would have been. It may even be that the Pats use a significant amount of hurry up themselves if it works well against Atlanta.

3 Re: Super Bowl LI Preview

Blount with 25 touches and 125y makes a lot of sense.

NE clearly has the stronger team, and little Bill did once work for, and win a lot of games with, big Bill.

5 Re: Super Bowl LI Preview

Fakcs could win, Pates could win,. who knows. will possiibl y buy Chinese food before gaem or eat something already have at home. maybe drink some rum. have picked Falcs 23-22 but could see Pates winjning 26-20. just so tired of Pates in super bowl. not enthused about game. maybe that will change come sunday evening

6 Re: Super Bowl LI Preview

Certainly citizen Joe you are at least curious what each team's tactics will be, and whether they'll be able to pull them off?

There is irrefutable truth to football, at least by the end of the game. That's part of what makes it beautiful.

9 Re: Super Bowl LI Preview

generally hate super nowl hype. these 2 weeks are sooo slow. do like atl falcs espeiclaly the RBs. so will surely get ibnto it. just been paying little less arttention than normal. am fan of b belichick evemn though do not like the team. Liked Belichick with Browns. thought was good there. D. halberstam book on him is gerta. if I ran a football tema it would be run similarly to Pates. always did that when played madden and other football games with any type of long term deal going on . certainly do it in fantasy sports. I get rid of guys left and right. no loyalty aside from top 1 or 2 or so stars. the way Pates churn roster is up my alley. just not excited seeing same uniforms in there and all this Bardy tlak all the time. aty least with m. ryan there are new thongs to talk about.

7 Re: Super Bowl LI Preview

Interesting to compare and contrast this with the Ringer's four alternative takes on the game:

8 Re: Super Bowl LI Preview

I get that ATL is bad against early runs. But if NE goes run heavy, isn't that a win for ATL?

NE's pass offense is better than ATL's pass defense by more than NE's run offense is better than ATL's run defense. Don't let ordinal rank seduce you.

12 Re: Super Bowl LI Preview

I just can't get past New England's lousy schedule and the fact they lost to the best team they played this year (Seattle). Even that win over Houston to start the playoffs reminds me exactly of the kind of game Peyton Manning would lose, only Brady has enough defense and the Texans a terrible enough offense it didn't matter.

The win over Pittsburgh was very solid, but to me that felt like the ceiling as to how well the Pats can play, and it still felt like the Steelers left opportunities on the board despite playing one game more in the playoffs and having a day less to prepare for a road game.

Meanwhile the Falcons crushed a red hot Green Bay team and beat one of the best defenses in the league in Seattle by 16.

I just think Atlanta has the edge and wins by 7, though I think the total points will be on the under side.

13 Re: Super Bowl LI Preview

Atlanta certainly can win, but these aren't particularly strong arguments.

"I just can't get past New England's lousy schedule"

NE has faced some terrible offenses, but Atlanta faced a lousy slate of defenses, exposing them to the same line of questioning.

"they lost to the best team they played this year (Seattle)"

In the aftermath of trading a key defender and before NE settled on a new LB rotation. Seattle specifically exploited a guy in pass coverage who doesn't do that much any more. It also needs to be recognized that that was nearly three months ago and Pitt is better than Seattle by most measures.

"it still felt like the Steelers left opportunities on the board"

Of course, so did both Seattle and Green Bay.

"Meanwhile the Falcons crushed a red hot Green Bay team"

Yes, and a Green Bay team nearly crippled in the secondary and almost as bad at the skill positions.

"...and beat one of the best defenses in the league in Seattle by 16."

After losing Earl Thomas, Seattle became one of the worst passing defenses in the league.

All the questions about NE are legit. The defense is still somehow unproven despite playing 18 games, the weapons would be a lot more imposing if Gronk were around, the middle of the OL is inconsistent. But, as I said before, the defenses faced expose Atlanta's offense to a similar line of questioning. Even more so during the playoffs when you factor in opponent injuries.

20 Re: Super Bowl LI Preview

You make some good points, but DVOA disagrees with your claim that "...Atlanta faced a lousy slate of defenses..."

From the second paragraph: "Only Washington played a more difficult schedule of opposing defenses than Atlanta did this season."

21 Re: Super Bowl LI Preview

I stand corrected. Earlier this week I saw this chart and presumed the DVOA rankings would be similar.

23 Re: Super Bowl LI Preview

Yes that is certainly not true. Atlanta faced Denver, Seattle (Earl Thomas version), Philadelphia, Arizona, San Diego, and Carolina (x2) during the regular season - all top 10 by DVOA. That is obviously reflected in their offensive DVOA. But the fact that Matt Ryan posted the 4th highest ANY/A ever against this schedule emphasizes how good/explosive this offense is.

24 Re: Super Bowl LI Preview

It should be noted that when they had their first game against the Panthers, Carolina was not playing like a top 10 defense. In particular, the secondary was in shambles.

However, to give credit where it's due, the Falcons did exactly what a great offense should do against a bad defense. They absolutely destroyed them, with Ryan throwing for over 500 yards.

15 Re: Super Bowl LI Preview

Here's how I see it. Atlanta's defense is somewhere in the Cleveland - Buffalo range. New England dropped 41 on Buffalo, 33 on Cleveland. I don't see them scoring less than around 31. I might believe 28, but then New England scored 28 on Seattle two years ago in the Superbowl with a worse offense against a MUCH better defense.

The question is, how confident are you in Atlanta scoring 35+ against New England? I am not very confident in that. The most points New England gave up this year is 31, vs. Seattle. It _could_ happen, but then New England could easily score 35+ also.

16 Re: Super Bowl LI Preview

It seems doable.

The defense closest in performance to weighted NE is Carolina. Atlanta put up 48 and 33 on Carolina.

They played Seattle twice. They put up 24 @SEA and 36 @ATL. Atlanta has outscored NE on average in every game played against a common foe.

18 Re: Super Bowl LI Preview

You could mention the 30 points Atlanta put up on SD as well. Their distribution isn't quite as similar as Carolina, but it's not too far off and the weighted overall is almost identical to NE's.

17 Re: Super Bowl LI Preview

I just checked and was surprised to see how much better this year's offense graded out as being. My first instinct would be that the difference is due to an improved running game, but that isn't even the case; DVOA just has the 2016 passing offense as being significantly better than the 2014 unit.

Despite that, I think some of the advantage is offset by Gronk's absence. I understand the numbers already account for that, I just think he is a particularly useful chess piece against this particular style of defense. Had Gronk not gotten hurt, I'd feel extremely confident about a NE victory. As is, it feels more like a toss up.

25 Re: Super Bowl LI Preview

"I just can't get past New England's lousy schedule and the fact they lost to the best team they played this year (Seattle)."

It's better to lose to the Chargers? At home?

Falcons have 5 losses, Pats have 2, one of which happened when they ran out of healthy QBs.

I wouldn't overthink this too much.

And, FWIW, PIT rates higher than Seattle did.

22 Re: Super Bowl LI Preview

In five of the last six Superbowls, the team with the lower DVOA (in 4 of the 5, significantly so) won. Nobody knows anything. H/t William Goldman.

Not that the analysis isn't fun, and it definitely helps me understand the game better (which is the whole point of the exercise). But it's a single game. Even if a pan-dimensional species of white mice invented a galactic computer to calculate the odds with unerring precision, in a single trial it would feel indistinguishable from a toss up.

26 Re: Super Bowl LI Preview

I'm rooting for Atlanta (new team, better story). No one has really stopped New England this year, and the Falcons' defense isn't particularly impressive, but, as the poster above notes, nobody knows anything

27 Re: Super Bowl LI Preview

Can Atlanta win? Sure, would be only a mild upset. But that's just it; about 6-7 times out of 10 I'd expect Darth & minions to prevail, and there isn't anything about the spread or money line to make this an especially attractive opportunity. Tap-ins, like last year, when you could get the Broncos and 5 points, aren't frequently available.

(Edit) It's funny how confirmation bias works, however. Looking back, I see I expected Carolina to pull away in the 2nd half, so I didn't see the tap-in at the time. What I didn't anticipate was how thoroughly the Broncos would outcoach the Panthers, especially if we think of Peyton mostly being a coach at that point. The Broncos got an early easy score when the Panthers gave Peyton a bunch of super-easy blitz reads, and as far as Bum Phillips' son vs. Don Shula's son? Well, baby seals have had better days against guys with baseball bats. That was the game.

I still think the Falcons have too many young guys on defense who are going to be asked to out-execute an opponent that puts more execution stress on defenses than anyone. If Quinn has gotten his guys ready for the challenge, it'll be one of the better Super Bowl coaching jobs.

29 Re: Super Bowl LI Preview

The offenses are a wash, the defenses are not. Atl's defense has played off their offense all year. In a nip and tuck game, they don't have the talent for a straight up fight. In a way, neither does Ne, but Bb is like a 12th, 13th, 14th, and 15th defender out there.

33 Re: Super Bowl LI Preview

1st and 10 at the NE 22, up 8 with 4:47. It's almost impossible to lose from that position. I'd love to see the win %. 3 runs and 37-40 yard FG, leaves the Pats down 11 with maybe 3:50 and no TOs. The way Atlanta was running it, they might even have got the first down. Instead they go run, freak out at the bad result, sack, holding, incomplete. Pats get the ball with 3:30, down 8, still alive. Un-freakin-believeable. Also, the challenge of the Edelman "Tyree catch" was incredibly stupid.

Credit to the Pats (although Brady was sketchy most of the game), but something about them leads their opponents to play like total idiots.

34 Re: Super Bowl LI Preview

I think it's fair to say Kyle Shanahan's play-calling there was atrocious.

Some coaches are great at design and terrible at situational understanding. I'm baffled at how the Falcons didn't get at least 3 points there.

35 Re: Super Bowl LI Preview

I can only attribute it to disorganized thinking. This is something that Belichik is rately guilty of. It was the same factor that helped the Patriots win their last Super Bowl. Carroll was disorganized in his thinking in the last two minutes of that gamee, and it made a huge difference

36 Re: Super Bowl LI Preview

Ugh, another gut punch to the stomach for someone rooting so hard against ne. Hey, might as well be up front about it right? Nothing personal Pats fans, I'm not delusional about their greatness, but man - 13 years ago I started watching football and this team snatched my heart right out of my chest. 13 years later - ugh, the machine never dies!

One of my closest friends is a die hard pats fan and even she refrained from drinking my tears. Ah...they win so damn much!