The third in an irregular series of posts about the Tennessee Titans. Perma-disclaimer: I write for Football Outsiders, whose statistics, rate-stat DVOA and counting-stat DYAR, will be referenced throughout this article. Some numbers may be through Sunday’s games; opponent adjustments from the Texans’ play will have slightly affected some numbers.
There has been a great deal of wailing and gnashing of teeth coming from Tennessee-following quarters, including from yours truly, about the quality of the Titans’ offense this season. It came Week 1 after the loss to Minnesota and has continued after most weeks. The wins against the Dolphins and Browns quieted the cries somewhat, but they rose again after Sunday’s loss to the Colts.
DVOA does not agree. DVOA actually rates the Titans’ offense fairly highly-currently 11th at 4.2%, right behind San Diego and Green Bay (9/10, 4.7%) and ahead of Detroit (12th, 4.2%), where Matt Stafford has gotten articles declaring him an MVP candidate. If they keep that up for the entire season, 4.2% would be their best offensive DVOA since 2009 (also 4.2%), and an 11th-place ranking would be their best since the halcyon days of Steve McNair and 2003’s third-place ranking. And it’s not just the DeMarco Murray-led run game that has been driving the offense. DVOA currently rates the pass game as pretty average, 15th at 16.0%, while Mariota is rated at almost exactly league average (-0.7%; 0% is league average).
This presents to us a puzzle. DVOA thinks the Titans’ offense is actually pretty average. The emanations from Titansland, including from me, say it is not average. I like DVOA quite a bit and believe it can be used to tell us interesting and powerful things. Why am I disagreeing with it?
1. The Titans are a consistent but not high-yardage offense.
DVOA places a great value on consistent success. This makes a great deal of sense. And this year’s Titans offense ranks highly by success rate-third at 48.4%, behind only Dallas and New Orleans, and well ahead of the league average of 43.3%.
At the same time, the Titans rank just 14th in yards per play, so their consistent successes are mostly only minor successes. My guess is a statistic like Bill Connelly’s IsoPPP for NCAA stats that measures explosiveness would rate the Titans quite poorly. Unfortunately, I’m not aware of an NFL statistic that captures quite that. (They’re 14th in FO’s open field yards, which measures rushing yards over 10, and in a 14-way tie for 13th with 3 passes of 40+ yards, for what those are worth.)
2. The Titans are good at moving the ball without getting first downs before third down.
Thing I believe is probably “true” but haven’t done the research to convince you of: first down is the most important down by a significant margin. Distance matters a fair amount for third down conversions, and the best way to get a first down is to not get to third down in the first place. (Along similar lines, I prefer the quarterback with more second quarter wins to the one with more fourth quarter wins.)
One thing the non-explosive Titans do not do is go from first down to first down. They had zero first down conversions in the first 58 minutes of Sunday’s game against the Colts, and that’s been a problem that’s plagued them for most of the season. In the first three quarters of games, the Titans get a first down on just 13.1% of their first down plays. League average is 21.1%. The Packers are second-worst at 16.2%. The Falcons lead the league at 29.8%. Yes, Atlanta is more than twice as likely as Tennessee to get a first down in first down in the first 45 minutes of a game. 31.7% of such Titans plays are successful but do not produce a first down. That’s the highest total in the league, and well above the league average of 21.9%.
The situation on second down may be nearly as bad, and is even worse relative to the league average. The Titans are successful but do not gain a first down on 21.1% of second down plays in the first three quarters. League average is 14.0%, and the Saints are second at 18.1%. Add them together, and the Titans have such plays on 27.0% of their first and second downs, compared to a league average of 18.5%, and well ahead of the second-place Saints at 23.5%.
Hey, what about the fourth quarter?
The fourth quarter is different. The basics are straightforward: in the first three quarters, 37% of successful first down pass attempts by the Titans result in a first down. In the fourth quarter, 85% do. My guess is this is likely related to the number of passes that come in two minute situations, and you’d see an even clearer trend if I looked to just those at the end of each half. But this post is enough work without me bothering to prove or disprove that.
3. Marcus Mariota has not been good outside of the red zone.
General DVOA note: player stats cannot be separated from the context of the team and particularly the offense. By “Marcus Mariota” in the above, I really mean “Marcus Mariota, throwing to Tajae Sharpe, Delanie Walker, and company, with Mike Mularkey and Terry Robiskie designing the offense and calling the plays, and with DeMarco Murray running the ball, while playing behind the Taylor Lewan-…-Jack Conklin offensive line, paired with Dick LeBeau running a defense with Jurrell Casey and company” but it’s too long to write that out every time.
Mariota did quite well in the red zone last year, posting a DVOA of 49.4%. This year, he has been even better, ranking second in DYAR (behind Drew Brees) and DVOA (behind only Tom Brady among qualifiers). But DVOA rates Mariota as just an average passer overall, even with that fine red zone DVOA. Why?
Simple: Mariota, as good as he’s been in the red zone, has not been good outside the red zone, which makes up 80% of the length of the field. From his goalline to the plus 20, he has posted a DVOA of -13.5%, which ranks 25th among the 32 qualifying quarterbacks. For comparison’s sake, that puts him just ahead of Case Keenum and just behind Ryan Six-picks-trick.
Being good in the red zone is great, but being good in the red zone is not something seems to be particularly consistent.
So what if the Titans aren’t quite as good in the red zone?
I think it’s useful to look at this through the prism of drive stats. The Titans currently rank 16th with 1.92 points per possession while scoring 6.0 points per red zone possession. If they instead scored the league average of 4.9 points per red zone possession, they would score about 0.3 fewer points per possession, which would be around 24th in the league.
4. Marcus Mariota has not been good in the under center pass game.
One of the big questions with Mariota coming out of Oregon was how he would adapt to taking many of his snaps from under center instead of from the shotgun. The answer we got in 2015 was “quite well.” He was actually better under center (1.6% passing DVOA) than he was in shotgun (-18.4%), and that was true under both Ken Whisenhunt and Mike Mularkey (cited numbers season-long). This year, that is not the case.
Among the 32 quarterbacks with 105 passes (the current DVOA QB leaderboard table cutoff), Mariota ranks 30th in both passing DYAR (-23) and DVOA (-16.8%) from under center. That’s … not so good. One of the players behind him is Cam Newton, and Mike Shula has had Cam attempt just 8.2% of his passes from under center this year, while Marcus has attempted 24.1% of his from under center. (The other is Brock Osweiler, whom you may have heard got $72 million over four years and is not playing well.)
5. Marcus Mariota has not been good on first downs.
Mariota ranks last in the league in both DYAR and DVOA (among 32 qualifiers) in first down passing DVOA. That is true even though he has a decent success rate. Turnovers seem like an obvious culprit, and they’re indeed part of the story. But only part of it. Without turnovers, his rank by DVOA jumps from 32nd all the way to 29th (among qualifiers). That’s still not good.
6. The Titans have been significantly better on third down then on first or second downs.
As much as I denigrate third down, third down performance is extremely important. And good news: led by Mariota, the Titans have been very good on third downs. He ranks fourth in DVOA (among qualifiers) and fifth in DYAR, and the Titans as a whole rank third in offensive DVOA on third down.
The bad news comes in two parts. First, notwithstanding their pretty good success rates on first and second downs, the Titans are not good on third downs because they’re ending up in particularly favorable third down situations. They are in fact almost dead average when it comes to third down situations-basically, a team that was precisely league average would have had two more third-and-mediums (4-6 yards to go) and one less third-and-short (1-3 yards) and one less third-and-long (7+ yards). In terms of average to-go distance, the Titans are pretty much right at league average (7.1 v. average 7.0).
Second, third down performance, especially third-and-long performance, tends to be pretty variable. The 2016 Tennessee Titans are a great example of this. They had 18 third-and-longs against the Vikings, Browns, and Colts, and got 10 conversions, an outstanding performance. They had 18 third-and-longs against the Lions, Raiders, and Dolphins, and got exactly 0 conversions, a quite lousy performance. That’s enough to rank quite highly for the season (currently third), but not enough to make me feel better about the offense.
The Titans currently have the second-biggest difference between their third down DVOA and their offensive DVOA on first and second downs, behind only New England’s unwanted experiment into the effect of quarterback quality on team performance. That gap is more likely to close over the course of the season than it is to increase or stay the same.
Hey, what about the run game?
The run game has been fine. I don’t quibble with DVOA putting the run game 9th. That’s probably about where I would have put it in general, though I admittedly haven’t really studied teams around the league enough to say that with confidence. The puzzle I was trying to solve was about the pass game, and why DVOA and I thought (and still think) different things about it. Also, passing is generally more productive and more variable than running, so more worth looking into even if things were equal, which they weren’t.
So, the Titans are running two offenses?
Eh, not necessarily that much more than most NFL teams do. The league as a whole runs a pass-oriented offense out of shotgun that is somewhat more efficient overall than the run-oriented offense they run from under center.
What about opposing defenses?
DYAR is adjusted for quality of opposing defense; that’s what the “D” stands for, so I haven’t touched on this subject yet.
Even though they opened with the Vikings, FO’s #2-ranked defense, the Titans have faced the league’s easiest schedule of opposing defenses thus far, including the 32nd-ranked Lions, 31st-ranked Colts, 30th-ranked Browns, and 28th-ranked Raiders. The Titans’ offense, both the run game and the pass game, has been worse than it “looks” so far, and is likely to decline going forward as the Titans face more accomplished defenses.
I believe this is particularly a problem for the Titans, as their lack of explosiveness and playmakers means their offense works best against the worst defenses, which tend to have more execution issues.
TL;DR?
DVOA likes consistently successful offenses, which the Titans are. But the Titans have too many short gains. They therefore must consistently and repeatedly execute to move the ball down the field to score points. To do this, they need to be great in high-leverage situations, particularly third downs and the red zone. They have been so far, which is a big reason DVOA likes them so much more than I do. Performance in high leverage situations tends to regress, so the Titans are not likely to be as good in those areas moving forward, plus they have not been as good as they look because of an easy schedule. They are therefore likely to score fewer points going forward, unless they start playing better.
Reading and Thinking Football 
I just wish Marcus Mariota could be set up for success in this offense. Unfortunately, I do believe he will continue to struggle as long as Mularkey sticks to his offensive scheme.
No comments? I’m a longtime FO reader and I absolutely love the way you guys break down film, create the advanced stats, and most importantly use the advanced stats in accordance with your judgement. And I think your complaints advanced above are fair.
Still, I don’t always like pitting advanced stats vs. advanced stats, as you’re doing above (i.e. ‘let’s look at how many first down plays beget more first down plays’ vs. DVOA’s consistency measure). The reason why I don’t like it is that you guys – and heck, even me as a rank amateur – at this point have watched enough football to look at the tape and see what things could conceivably improve during the season and which things are a serious problem that DVOA cannot “see.” For example, even an idiot like me can see awful throwing mechanics or “happy feet” or undersized linemen or a tendency to always run to the outside. We can tell what’s fixable in a season, what’s not. DVOA cannot really work with those observations except in the most remote way. Also, I’m not clear that discrediting a team’s gains because of luck or favorable situations is a good idea. You guys taught me that there’s a value to “garbage time:” sometimes, it tells you whether a defense or offense has adapted to what they’ve faced. And you guys have taught me that “stomps” matter more than “guts,” that what matters is taking advantage of weaknesses or favorable opportunities. Yes, that means not squeaking by for the best teams. What does it mean for teams that are improving or not improving?
To be fair, DVOA sees a lot of things we don’t see. It has tremendous predictive value because it measures consistency, takes into account strength of schedule, and does account for how teams are actually winning games, albeit indirectly. I know, I sound like I’m preaching to the choir, but I’m really preaching to myself. DVOA has been a very good tool for me in understanding what I’m seeing on the field. It helps me better see which teams are coming together to play at a higher level, which ones are coming apart. It helps me see strengths on otherwise weak teams.
So to me, I think you can ask “hey, can the Titans do X or Y better right now,” and ask if DVOA might be leaving room for that. You mentioned Mariota’s efficiency under center and in the shotgun. What’s going on specifically during those plays? Deadspin (!) beautifully broke down the Lions offense recently, talking about how putting Stafford in the no-huddle and spreading the ball around has made them much quicker than the defenses they’ve faced. Would a similar “fix” work for the Titans, or would they compromise something essential to the run game? Why isn’t an above average run game creating second and short? Why is it not creating playaction and making the pass game deadlier?
If the answer to these sorts of questions (I’m not saying I have the exact questions right) turns out to be “none of these things are fixable within a few weeks,” then DVOA in this case is up a creek without a paddle. It is not functioning as a predictor of consistency, it’s just a yardstick measuring what’s put into it. But if there are small fixes that make the Titans offense deadlier, I think it’s reasonable to assume that coaches not named Norv Turner can put them into play.
I dunno. I’m a rank amateur, I’m out of my league, but this is how I use DVOA for better or worse. I’m grateful to you guys for sharing your insights and knowledge, you’ve taught me to see the game better. Whether I can write anything but a rambling nonsensical comment remains to be seen.
Of course we all know that it is not all about Mariota as a whole. Sure he needs to gain more confidence within the offensive scheme and he needs to think more about his play. But it is also about receivers being quicker on the average and creating space. I do not agree, despite the stats that the running game is fine . Murray and company have to many no gain and minus yard situations which force those third down situations. Then there is the inconsistent defense, which is another story.
Given Mariota’s rookie performance, it seems a safe bet that while his red zone and 3rd down performance may be more impressive than they ‘should’ be, his performance outside of the red zone is likely a little worse than it ‘should’ be. I see your point that his performance in red zone and 3rd down may be illusory because those usually revert to the mean YOY as they are naturally smaller sample sizes. But, we’ve only played 7 games so far, so Mariota’s performance outside of the red zone is also somewhat of a small sample size. Which is to say, Mariota is clearly not the second coming of Tom Brady right now, but he may still be a better quarterback than his non-redzone/3rd down work suggest year to date.
What do you think the ceiling is for this offense? While the Titans aren’t anything epic, it seems a low-variance, ‘highly-successful in small chunks’ offense could still be massively effective if everything was firing on all cylinders. In fact, it seems less susceptible to flaming out in the playoffs like other great offenses are wont to do that rely on the rare splash plays to come through.