Earlier in the offseason we looked at what Matt LaFleur’s offense in Tennessee was likely to look like based on his previous NFL stops in Washington, Atlanta, and Los Angeles. If you missed it, you can check that out here.
Today we will breakdown elements of both Dean Pees’ defenses in New England and Baltimore as well as Mike Vrabel’s defense in Houston to get an idea of what they might bring to the Titans.
A few of the specific components that I wanted to examine are performance, personnel usage, and scheme. For the latter two categories we will look at similarities and differences between Vrabel and Pees as we try to build a framework for what the defense might look like with them working together.
Performance
It’s no secret that the Texans defense struggled statistically in 2017, Vrabel’s first and only year as defensive coordinator. There are plenty of reasons for why those struggles occurred — we will get to those later — but for now let’s look at the numbers. All rankings below are measured from best to worst so a ranking of 1st is good and 32nd is bad.
The raw stats are mostly ugly, particularly when it comes to pass defense.
Points Per Game Allowed: 27.3 (32nd)
Yards Per Game Allowed: 346.6 (20th)
Yards Per Play Allowed: 5.7 (28th)
First Downs Allowed: 301 (12th)
Net Yards Allowed Per Pass Attempt: 7.1 (30th)
Yards Per Carry Allowed: 4.0 (10th)
Turnovers Forced: 16 (27th)
Sacks: 32 (23rd)
The advanced stats are slightly more kind, but still not great.
Football Outsiders DVOA: +5.6% (23rd)
Pass Defense DVOA: +19.1% (25th)
Run Defense DVOA: -9.9% (12th)
Opponent Drive Success Rate: 0.678 (13th)
Explosive Pass Plays Allowed: 53 (20th)
Explosive Run Plays Allowed: 51 (20th)
Football Outsiders also tracks a weighted DVOA which gives additional weight to more recent results in their formula. The weighted defensive DVOA for the Texans was +13.4%, good for 31st in the NFL, which shows just how much the defense cratered down the stretch as the injuries and losses started to pile up.
The overall numbers tell a pretty clear story. Houston was pretty tough against the run, but really struggled to stop opposing passing games in 2017. There are a couple simple explanations for that.
First, losing J.J. Watt and Whitney Mercilus in the 5th game of the season really took some sting out of their pass rush and allowed offenses to give extra attention to the only remaining threat, Jadeveon Clowney. Imagine what the Titans 2017 defense would look like if you took Jurrell Casey and Brian Orakpo away.
Second, they allowed their top cornerback, A.J. Bouye, to walk in free agency before the season. Johnathan Joseph (33) and Kareem Jackson (29) — the top two corners remaining after Bouye left — have both lost a step. Former 1st round pick Kevin Johnson continues to struggle to stay on the field and isn’t that effective when he’s out there. Behind the corner group, safeties Andre Hal and Marcus Gilchrist were fine, but not good enough to cover for the below average corner play. The Texans secondary simply lacked playmakers in 2017, something the team clearly felt as well based on their signings of Tyrann Mathieu and Aaron Colvin.
Vrabel’s defense also got saddled with a terrible Texans offense that often left his unit with a short field to defend, especially late in the season. Houston’s opponents started on average at their own 29.5 yard line — 7th worst in the NFL — largely caused by 28 turnovers by the offense.
Of course its also fair to point out that one of the biggest differences between the Texans excellent 2016 defense (9th overall in DVOA) and the poor 2017 version was the switch in defensive playcallers from Romeo Crennel to Vrabel. Crennel was also without the services of Watt for much of that campaign, though he did have Bouye and Mercilus as well as younger versions of Joseph and Jackson to work with. It’s almost impossible to separate team performance from coaching performance so it’s hard to say how much of the Texans defensive failures should be blamed on Vrabel or how much should be blamed on the personnel changes or bad field position or any other potential cause.
While he will certainly have plenty of say in the overall design of the defense — and perhaps even specific calls from time to time — Vrabel won’t be the primary defensive playcaller for the Titans. He’s chosen instead to entrust those duties to veteran defensive coordinator Dean Pees. Pees has 10 years of experience as a defensive playcaller at the NFL level comprised of a 4 year stint in New England — where he also served as Vrabel’s position coach — and, most recently, a 6 year run in Baltimore that ended with his very brief retirement after the 2017 season.
Pees collected two Super Bowl rings during his coaching career. One as the linebackers coach for the 2004 Patriots and one as the defensive coordinator for the 2012 Ravens. His defenses have ranked in the top half of the league in 7 of his 10 seasons as an NFL DC.
2006 Patriots: -9.2% (7th)
2007 Patriots: -5.8% (11th)
2008 Patriots: +3.6% (17th)
2009 Patriots: -1.1% (14th)
2012 Ravens: +2.2% (19th)
2013 Ravens: -8.7% (7th)
2014 Ravens: -4.6% (8th)
2015 Ravens: +5.1% (20th)
2016 Ravens: -9.9% (6th)
2017 Ravens: -13.9% (3rd)
For some context, the Titans haven’t finished higher than Pees’ worst season (20th) since 2011 when Tennessee finished 15th in defensive DVOA. If Pees is able to coax even an average season (for him) out of the 2018 Titans that will be a massive upgrade.
Pees’ “bad” seasons — 2008, 2012, and 2015 — were all severely impacted by sudden departures of key talent on the defensive side of the ball. The 2008 Patriots defense lost 2007 First Team All-Pro cornerback Asante Samuel to free agency before the season (the Pats starting corners were Ellis Hobbs and a 31 year old Deltha O’Neal for the majority of 2008, not good) and then lost pass rusher Adalius Thomas and safety Rodney Harrison to season-ending injuries during the season. The 2012 Ravens similarly lost superstar pass rusher Terrell Suggs for half the season and Ray Lewis for most of the season due to injuries. The 2015 Ravens lost nose tackle Haloti Ngata in free agency prior to the season and once again lost Suggs for the majority of the season. All defensive coordinators would struggle with that kind of loss in talent.
In 2017, Pees put together his best statistical defense of his career. Here are some of the stats that his Ravens defense was able to put up last year. Again, all rankings are best to worst (1st is good).
Points Per Game Allowed: 18.9 (6th)
Yards Per Game Allowed: 325.0 (12th)
Yards Per Play Allowed: 5.0 (8th)
First Downs Allowed: 296 (9th)
Net Yards Allowed Per Pass Attempt: 5.6 (5th)
Yards Per Carry Allowed: 4.1 (19th)
Turnovers Forced: 34 (1st)
Sacks: 41 (11th)
Football Outsiders DVOA: -13.9% (3rd)
Pass Defense DVOA: -15.4% (2nd)
Run Defense DVOA: -12.0% (9th)
Opponent Drive Success Rate: 0.651 (6th)
Explosive Pass Plays Allowed: 48 (12th)
Explosive Run Plays Allowed: 39 (6th)
That’s a pretty dominant defense. To put that 34 forced turnovers number in perspective, consider that the Titans have forced just 39 turnovers in the past two seasons combined.
The primary complaint that you’ll hear about Pees as a defensive coordinator is that the Ravens had a habit of blowing leads late in games due to Pees going to a soft prevent defense. I searched Pro-Football-Reference.com for all games since 2012 where a team has blown a lead of 7 or more heading in to the 4th quarter and found 86 results across the NFL during that time span. The Ravens had 4 of those 86 which is above average, but nowhere near other teams like the Chargers (7) or Browns (6). Interestingly enough 3 of those 4 instances came against the hated Steelers, which may be a reason that Ravens fans are so fixated on that specific issue.
I’m not terribly concerned with the prevent defense issue personally. Pees consistently churns out good defenses and if the prevent becomes an issue I would fully expect Vrabel to step in and demand an adjustment there.
Personnel Usage
Personnel usage is always one of the more fascinating aspects of a new coaching staff to me. How often do they rotate defensive linemen? Do they set their corners as right and left or field and boundary? Do they use specific corners to shadow specific receivers? Do they like to use a “moneybacker” (safety playing in a traditional linebacker role)?
The Texans were very limited along the defensive line after losing J.J. Watt and Christian Covington early in the year. That led to nose tackle D.J. Reader getting more snaps on pass rush downs than you would typically expect. Reader ended up being the only Houston interior defensive linemen to play over 50% of total defensive snaps in 2017, checking in at 51.2%. It’s hard to get a true read of how Vrabel would like to rotate this position in an ideal scenario based on how banged up the Texans were last season.
Dean Pees’ defense also had his best defensive lineman injured for part of the year when Brandon Williams went down, but he only missed 4 games. Those 4 missed games reduced Williams’ snap total as he finished 3rd on the team being in on 43.4% of total defensive snaps. Two other interior defenders finished with a higher percentage: Willie Henry (54.6%) and Michael Pierce (54.3%). Even if Williams had been healthy for all 16 games, he likely wouldn’t have finished with more than 55-60% of total snaps. In 2016, Williams and Timmy Jernigan both topped out around 60% of total snaps as well so I believe that to be Pees’ preferred ratio given optimal conditions. However, my expectation is that Jurrell Casey will be treated differently than those players given his level of play and ability to contribute on all three downs. Last season Casey saw a snap percentage of 79.5%, with no other defensive linemen finishing above DaQuan Jones’ 40.1%. While I don’t think Casey being off the field is ever a good thing, I wonder if the Titans might not try to ease that back towards 70% in 2018 by resting him on run downs given the strength of Jones, Austin Johnson, and Bennie Logan as run defenders.
Similar to the situation at defensive line, an injury sabotaged Vrabel’s preferred strategy at edge rusher last year as well. Jadeveon Clowney was heavily featured, checking in at 87.4% of total defensive snaps, but instead of seeing a similar snap count for Whitney Mercilus, the Texans ended up relying on a committee approach at the opposite edge position in 2018 with many players taking turns opposite Clowney. Often, the Texans would opt to keep three inside linebackers in the game rather than the more traditional 3-4 set up with just two. The starters at inside linebacker — Bernardrick McKinney (93.7%) and Zach Cunningham (79.4%) — almost never came off the field under Vrabel.
The Ravens under Pees had a more traditional 3-4 snap distribution with starters Terrelle Suggs (77.2%) and Matthew Judon (72.1%) handling the lion’s share of the work, while backup Za’Darius Smith (48.5%) also got a lot of work. That distribution lines up pretty closely with what the Titans did with Brian Orakpo, Derrick Morgan, and Erik Walden in 2017. It seems to me that would be the most obvious deployment for Orakpo, Morgan, and rookie Harold Landry this year barring an unexpected training camp development. At inside linebacker C.J. Mosley checked in for 98.4% of snaps. Across from Mosley, the Ravens typically used undersized coverage linebacker Patrick Onwuasor (59.2%) which is interesting. The inside linebacker rotation is among the toughest spots to project on the Titans roster right now. Most expect first round pick Rashaan Evans to step right in to a 3-down role, but it may be tough to unseat Wesley Woodyard if the veteran plays anything like what he did last season. At least three of Evans, Woodyard, Will Compton, and Jayon Brown figure to see a lot of action, but it’s hard to say who will be the early down “starters” and who will be the odd man out at this point in time.
Lastly, we have the defensive backfield. Neither Vrabel nor Pees used their cornerbacks to shadow receivers often in 2017. According to PFF, the only game that Houston used a shadow was in their Week 10 matchup with the Rams when Johnathan Joseph attempted to follow Robert Woods with disastrous results — Woods caught 5 of 7 passes thrown his way with Joseph in coverage for 138 yards and a touchdown. The Ravens didn’t use shadow coverage in any games last year per PFF’s analysis. That may change a little bit with different personnel in Tennessee, but it seems unlikely that it will be used as a week to week staple.
Safety usage between the two coaches was a little different, as was the caliber of safety they each had to work with. Vrabel leaned heavily on Andre Hal (91.7%) and Marcus Gilchrist (79.5%), but he also put Eddie Pleasant (29.9%) to work as a 3rd safety/jumbo slot corner from time to time. Pees almost never took his extremely talented safety duo off the field as both Eric Weddle and Tony Jefferson played 99.1% of the Ravens defensive snaps in 2017. The Ravens did use looks with 3 safeties on the field quite a bit as well as Anthony Levine (23.9%) got on the field for nearly a quarter of all defensive snaps.
Obviously the personnel usages will likely vary in Tennessee as the Titans have a different defensive roster makeup than either the 2017 Texans or 2017 Ravens, but it’s a decent starting point for how Vrabel and Pees might like to rotate their defense in 2018.
Scheme
Vrabel and Pees come from a similar background so it should come as no surprise that their defensive schemes align pretty closely. Obviously both run base 3-4’s, but the similarities go deeper. Let’s start with a pass rush look that Vrabel is largely associated with: the diamond front.
So what is the diamond front? You can check out a couple good reads on the concept here and here, but in general it’s not a terribly complicated concept. Its basically a variation of a zone blitz where the defense starts with 5 defenders on the line of scrimmage across from the offense’s 5 offensive linemen, dictating one on one blocking assignments pre-snap. After the ball is snapped, 1 or 2 of those defenders may drop off in to coverage while the others rush. They could also elect to bring all 5, or they could even drop 1 and replace him with a blitzing linebacker, corner, or safety. There are lots of options that can be built off this very basic look, but the beauty of the design is the ability for the defense to dictate blocking assignments to create mismatches in their favor.
The diamond front was a 3rd down staple for the Texans in 2017. Let’s start with its most basic application. Pre-snap the Texans have 5 lined up on the line of scrimmage — edge rusher Brennan Scarlett, defensive tackle D.J. Reader, inside linebacker Bernardrick McKinney, defensive end Christian Covington, and edge rusher Jadeveon Clowney — with free safety Andre Hal and inside linebacker Zach Cunningham playing on the second level behind them. Each of the 5 guys on the line of scrimmage rush which dictates one on one blocking assignments across the board. You can imagine this being a much scarier set when the Texans had J.J. Watt and Whitney Mercilus in the game instead of Reader and Scarlett, but it’s still effective here as Clowney destroys Seahawks left tackle Rees Odhiambo and gets the strip sack.
Here is another variation that can be run out of this front. This time the Texans line up Watt, Clowney, Mercilus, Cunningham, and McKinney on the line of scrimmage. You also see the Texans 3rd safety, Eddie Pleasant, lurking at the line, lined up over the tight end. At the snap, Cunningham, Pleasant, and McKinney bail in to zone coverage while Watt, Clowney, and Mercilus all rush on the offense’s right side with Clowney in particular getting a good push in to Dalton’s lap. Joining those three is blitzing slot corner Kareem Jackson who beats the running back’s block — the back is late to get over due to his pre-snap alignment — to get the sack. You can see the numbers advantage that Houston created pretty clearly here. The Texans are rushing 4 players from the center to the right of the offensive line, but the Bengals only have 3 — including the center — in to block on that side. That’s a problem for the offense.
Here is a clip of the Ravens running something very similar. Pre-snap, they are lined up with edge rusher Za’Darius Smith, defensive tackle Willie Henry, inside linebacker C.J. Mosley, edge rusher Matthew Judon, and edge rusher Terrell Suggs on the line of scrimmage. You also have safeties Eric Weddle and Anthony Levine in the box. Just before the snap, the two safeties move inside to blitz. The Texans do a nice job picking up the blitz, but you can see some of the problems that these looks cause for offensive linemen when you look at the right tackle. He sees the safety sneaking inside and looks to help the right guard with Judon first, but that gives him pretty much no shot at slowing down Terrell Suggs who very nearly gets the sack.
Another common thread between the two defenses was the use of a 3rd safety at the line of scrimmage on 3rd down plays. The Texans Eddie Pleasant and the Ravens Anthony Levine — #41 in the gif below — were both used in a very similar fashion. On this play, Levine shows blitz early, but then bails out to cover the H-back man to man while another safety, Tony Jefferson, blitzes from the other side.
That brings me to yet another similarity. Both Vrabel and Pees were masters at deceiving quarterbacks pre-snap, particularly on clear passing downs. This next gif is a second view of the play above. You can see the Ravens start with a clear Cover 2 look before rotating late to a single high look just before the snap. On this snap, Kizer is reading from right to left. The deep comeback is taken away by great coverage from Brandon Carr and then the dig from the backside is taken away by Tyus Bowser dropping off the line of scrimmage in to a robber position, leaving Kizer with nowhere to go with the ball.
Here’s an example of pre-snap deception from Vrabel’s Texans. If this looks familiar, it’s probably because Justin Graver just did an incredible breakdown of the play for his Mariota interception series. The Texans show a single high look pre-snap, but then roll in to a Cover 2 zone. It’s essentially the same defensive call as the example from the Bengals game above. The Texans have 6 defenders lined up on the line of scrimmage, but after the snap the 3 to the right of center bail in to coverage while the 3 to the left of center rush with a defensive back joining in to create a 4-on-3 numbers advantage for the defense. The Titans offensive line does a nice job picking it up, but Mariota’s throw floats in to the path of the safety who picks it off. I don’t know that Mariota was necessarily fooled on this play by the shifting coverage — I tend to think this ball floated on him due to the hamstring that he had injured on the previous series holding him back from being able to truly step in to the throw — but you can see how Vrabel liked to disguise his coverages and blitzes using various looks before the snap.
While the play above shows Houston’s defense in zone, both the Texans and the Ravens were among the most frequent users of man coverage in the NFL last year. Here’s an example of the Ravens in Cover 2 man on a 3rd and 4 against the Browns. They are in what I would call a “big nickel” personnel package with 4 defensive linemen/edge rushers, 1 inside linebacker, 3 cornerbacks, and 3 safeties on the field. This was a very common personnel package for the Ravens on 3rd down in 2017.
It’s a pretty straightforward defensive call here, with the Ravens matching up in man coverage underneath with Weddle and Jefferson hunting for routes at the sticks on either side of the field. You can see Weddle’s otherworldly anticipation here as he is breaking on the out route to Duke Johnson out of the backfield before Kizer even goes in to his throwing motion. Johnson mishandles the pass and Weddle is able to come up with a pick on the deflection, but even if this ball is caught Weddle would have been able to make the stop well short of the line to gain. Weddle is one of the safeties that Kevin Byard has studied extensively since coming in to the league and I’m sure Pees envisions him playing the Weddle role in his defense in 2018.
On early downs, both Vrabel and Pees run what I would consider the closest thing to a true two-gap 3-4 base defense in the modern NFL. The two defensive ends typically are asked to stack the lineman across from him and maintain control of the gap on either side while the nose tackle attempts to absorb a double team, effectively keeping the two inside linebackers clean to flow to the ball and make the tackle as C.J. Mosley does on this play.
While this is considered their “base” defense, it’s worth pointing out how infrequently teams are really in base these days. When you consider that both the Ravens and the Texans had multiple run blitzes and defensive line calls to go with the base front, they really only had a true two-gap “base” call on a handful of snaps per game.
When the Ravens went to nickel packages, strong safety Tony Jefferson’s role really stood out to me. He could often be found near the line of scrimmage acting as a de facto 3rd linebacker as he is in the clip below. On pass plays, he usually was responsible for man coverage on the opposing team’s tight end. The way Jefferson was used under Pees really fits Johnathan Cyprien’s skill set.
All in all, I don’t think the overall schemes that Vrabel and Pees are likely to use in Tennessee will look drastically different from what Titans fans have seen from Dick LeBeau the last couple years. The bigger changes will be in the details of how they want specific responsibilities to be carried out on each play and the game day playcalling.
The things I’ll be interested in early will be how they plan to handle certain roles within the defense. Vrabel’s Texans opted to leave two inside linebackers — Bernardrick McKinney and Zach Cunningham — on the field for almost every snap last year, while Pees’ Ravens preferred to sub one off for a 3rd safety. I think the smart approach — and the one they’re most likely to take — will be to focus on the best 11 regardless of position label.
That could lead to some interesting cross position battles. For example, we could see a battle between Jayon Brown and Dane Cruikshank for a spot in nickel/big dime packages. You have to assume that Jurrell Casey, Brian Orakpo, and Derrick Morgan will be on the field for nearly every 3rd down, but who will join them? I would expect Harold Landry to be the 4th pass rusher, but could Rashaan Evans take that role if the Titans feel more comfortable with a Woodyard-Brown combination at linebacker? I think a diamond front featuring Orakpo, Casey, Morgan, Landry, and Evans would be hell for opposing offensive lines and I can’t help but think that’s something we see quite a bit of in 2018.
The good news is that the Titans defense features some good depth with versatile skill sets so there are quite a few combinations that could work well. While all eyes will certainly be on the new look offense under Matt LaFleur early, I actually think this defense could carry the team if needed for stretches. That’s the first time I’ve thought that about a Titans defense in at least 7 or 8 years.