Looking at the numbers behind the Titans’ hyper-efficient day on offense

Happy Thanksgiving everyone! If you’re joining me for some pre-turkey reading, I’m happy to distract you from your family members. I want to say, I’m extremely thankful for Derrick Henry.

Anyway, onto the write-up...

Since Ryan Tannehill became the starting quarterback for the Titans, I’ve been doing these efficiency report articles looking at the advanced statistics for Tennessee’s offense.

Mike wrote on Monday about how well the offense has performed over the last five games, and I encourage you to check that out here if you missed it. He also published his weekly All-22 on Wednesday, which you should read as well. In this article, I’ll look to compare the numbers from my previous reports and continue building the sample size for Ryan Tannehill’s version of this offense. We’ll let the numbers do most of the talking.

The usefulness of this information is up for debate, but to me it provides a glimpse at the sustainability of the offensive success. Higher efficiency outputs over larger sample sizes mean there are less “fluky” plays influencing an offense’s performance. The good news before we start is that the Titans offense continues to perform at a more and more efficient level with each passing week. As the sample size is growing larger, the offense is becoming more efficient.

That’s what we call improvement, plain and simple. So without further ado, let’s dive in...

Drive Success Rate

Let’s start where we always begin, with drive success rate. If you’re unfamiliar, “drive success rate” measures the percentage of 4-down series resulting in either a first down or a touchdown. Here are the Titans’ success rates broken out.

The Titans saw their most efficient outing of the season from a DSR standpoint in Week 12, converting 20 of 25 down series into first downs or touchdowns. Here are their 5 non-conversions:

• Tannehill strip sack in the red zone
• Derrick Henry’s fumble
• Incomplete pass on 3rd and 8, punt
• Incomplete pass (the miss deep to Corey Davis) on 3rd and 5, punt
• A failed 3rd-and-11 run that came up 3 yards short (after one of just three offensive penalties committed in this game wiped out a first down) late in garbage time with the game well out of hand, leading to the Titans’ third punt.

That’s it. The failed 3rd-and-11 was the Titans only three-and-out of the game (more on that later).

The average DSR continues to climb from Weeks 7-9 to Week 10 to now Week 12. Tannehill’s emergence is one of many factors; another is the gelling of the offensive line, and yet another is Arthur Smith’s ability to improve as a gameplanner and playcaller on a seemingly weekly basis.

Yards Per Drive

Here are the Titans’ average yards gained on each drive, broken out for this season:

This was another stat where the Titans crushed their previous season high. A 74-yard touchdown run by Henry and a 65-yard touchdown pass to A.J. Brown sure help this metric, but 475 yards gained on 11 drives isn’t just the result of two big plays. The Titans moved the ball at will in this game, and if it hadn’t been for a couple of fumbles and those two stalled drives, this already insane offensive output could’ve been even more extreme.

As Mike pointed out already, the Titans set a franchise record with 9.1 average yards per play in this game. The 471 total yards (475 on 11 drives minus 4 yards on two kneel-down drives) is the team’s most in a game since 2017.

Three and Outs

Over the first six games, three-and-outs were a major problem for the Titans offense. We didn’t see a ton of improvement in this category over weeks 7-9 outside of one great performance against Carolina... But here’s the updated numbers on three-and-outs per drive, broken out:

The Titans had only one three-and-out on 11 drives in this game, just shy of their best week in this category when they had 0 three-and-outs in Week 7. The Titans have continued to eliminate negative plays from their offense over the past few games (although they still lead the league in that category), which helps limit three-and-outs tremendously.

Negative Plays

Henry had only one negative run in this game, his first negative carry since Week 8, and Tannehill was only sacked one time (the strip sack). The reverse attempt to Khalif Raymond, which lost 5 yards, was the only other negative play for the Titans offense in this game (excluding four Tannehill kneel-downs).

Courtesy NextGenStats.

The Titans only committed three penalties on offense the entire game (one coming on that final three-and-out drive in the fourth quarter). These penalties all wiped out first downs (or touchdowns), which is not good obviously, but the general lack of false starts and drive-killing penalties allowed the offense to stay in third-and-manageable, thus limiting the three-and-outs.

The offense was also able to overcome some negative plays, which is another improvement for them. They eventually scored a touchdown on the drive after Henry’s one negative carry, and they were able to convert a first down after the negative play to Raymond (they actually scored a touchdown on the very next play, but it was negated by a holding penalty — not a good sequence there).

The Titans’ lack of offensive penalties is especially impressive considering the cheap shots taken by Yannick Ngakoue and Calais Campbell in this game. The Titans, particularly Taylor Lewan and Rodger Saffold, deserve credit for keeping their composure and not retaliating. That kind of discipline is a recipe for success against any team.

Red Zone Performance

The Titans entered this game red hot in the red area, scoring touchdowns on all 10 trips with Tannehill as the starting quarterback. Here’s an update on how the Titans red zone touchdown rate breaks down:

The perfect red zone streak ended when Tannehill was strip-sacked on the Titans’ first red zone trip of Week 12, but the team scored touchdowns on their next three trips. Through six games, they were the 19th in the league. They’ve since climbed to 1st overall with a season-long touchdown rate of 72.41%.

Play Action Frequency

In Tannehill’s first four starts, the Titans called play action on 20.9% of his dropbacks. As I mentioned on our podcast this week, Art Smith dialed up play action on 14 of 22 dropbacks against Jacksonville, good for 63.6%, more than three times the season average.

On those 14 dropbacks with play action, Tannehill completed 10 of 10 passes for 218 yards (21.8 YPA) and two touchdowns while adding three scrambles for 44 yards and two rushing touchdowns (the 14th dropback was Ngakoue’s strip sack). His passer rating off play action was 158.3... for the second straight game.

On his 8 non-play action dropbacks, Tannehill was 4/8 passing for 40 yards (5 YPA).

Play action was extremely effective in this game, but not just because of Henry’s performance. Whether or not Henry is running well is irrelevant to the success of play action (in fact, Ben Baldwin and others have found time and again that there is no correlation between rushing volume or success and play action passing success).

What matters more is simply the threat of running the ball. Henry has run all over the Jaguars for four seasons now, so it makes sense that they would load up to stop him. This may have opened up play action even more than normal for this match up.

Will play action continue to be as effective if the usage rate stays so high? Who knows, but I’d be willing to try it and find out...

Tight Window Throws

If you’ve missed the last few editions of this article, here’s a quick refresher about Next Gen Stats’s advanced metrics, the most interesting of which to analyze (in my opinion) are the passing stats for aggressiveness percentage and completion percentage compared to expectation.

Aggressiveness percentage refers to the percentage of pass attempts on which a defender is within 1 yard of the receiver at the time the ball arrives. This is a piece of analysis that can help assess anything from the quarterback’s decision-making to the offensive scheme to the receivers’ ability to get separation. I’m choosing to look at how the results affect the Titans’ offense and the sustainability of having a high percentage in this stat, rather than trying to understand why Tannehill’s percentage is so high (currently 5th in the NFL at 22.1%).

The “completion percentage above expectation” stat is defined as the difference between a passer’s actual completion percentage and their Expected Completion Percentage, which is calculated by Next Gen Stats using their Completion Probability formula to determine what a passer’s completion percentage is expected to be on any given play. This “formula” is based on numerous factors, including receiver separation from the nearest defender, where the receiver is on the field, the separation the passer had at time of throw from the nearest pass rusher, and more.

Through 9 weeks, Tannehill sat 4th in the league with a 21.4% aggressiveness rate on the season. I maintain the opinion that the passing offense cannot sustain success with such a high aggressiveness percentage unless you believe (A) that Ryan Tannehill is among the best tight-window throwers in the NFL and (B) that the Titans receivers are among the best contested catch-makers in the NFL.

Despite the concerns I have about the longterm sustainability of the Titans offense, Tannehill’s aggressiveness percentage actually shot up to 31.6% for Week 10 in a 35-point performance against the Chiefs. Tannehill had a fine day throwing the ball, completing nearly 70% of his passes for 9.5 yards per attempt and 2 touchdown passes, and his passer rating of 133.9 was excellent — but he also only completed 13 passes for 181 yards. It was a very efficient, although not proficient, day for the passing game, likely because nearly one-third of his attempts were thrown into coverage (and definitely because the running game was so effective as to render the passing attack unnecessary).

Against the Jaguars last week, Tannehill targeted tight windows on only 16.7% of his pass attempts, his second-lowest aggressiveness percentage since taking over as the starter. This drop in aggressiveness may be explained by the high play-action rate. I wondered if there might be an inverse correlation between aggressiveness percentage and play action usage...

The purpose of play action is to suck the linebackers forward and create open throwing windows behind them. The rise to a 64% play action usage rate does correspond with Tannehill’s drop in aggressiveness percentage; he only attempted 3 “aggressive” throws in this game on 18 attempts.

Play-action and aggressiveness rates don’t always have such a correlation (my samply size is much too small to determine anything, but we can see the weaker relationship in Tannehill’s other starts). Generally, a lower aggressiveness percentage will lead to easier completions with more room for yards-after-catch, so you’d expect a passing game to function at a higher rate with a lower aggressiveness percentage.

As a reference point, Patrick Mahomes finished his MVP season with a 12.4% aggressiveness rate, a shade lower than Tannehill’s lowest % in a game this season (12.8%). Lamar Jackson is currently at 15.1%.

The top five leaders in aggressiveness percentage for 2018 were Josh Rosen, Jeff Driskel, Jameis Winston, Ryan Fitzpatrick, and Sam Darnold (in that order). All of them were lower than Tannehill’s current 22.1% season-long rate.

Four of these passers were in the top-five for interception percentage last year. Three of them were in the bottom-five for “successful passing play percentage” (according to Sharp Football Stats).

Allow me to reiterate: passing offenses cannot sustain longterm success with such a high aggressiveness percentage. The Titans should strive to be closer to the 16.7% ballpark we saw on Sunday rather than Tannehill’s 5-game average (as a starter) of 23.2%.

A lower aggressiveness percentage isn’t always a good thing. That said, the three lowest aggressiveness percentages this season belong to Drew Brees (10%), Kirk Cousins (10.6%), and Patrick Mahomes (11.5%). What does this mean? You decide.

Here’s the full list if you’re curious to browse.

Completion Percentage vs Expectation

In Tannehill’s first start for Tennessee, his completion percentage over expectation was bonkers at +22.1% — that was always going to be an unsustainably high number.

This stat predictably dropped in Week 8, all the way to +1.7% over expectation. Against the Panthers, Tannehill’s completion percentage again fell way below the +22.1% threshold, this time checking in at +4.6% above expectation. Both numbers were well within the “average” range among all quarterbacks.

But in the past two games, the numbers have spiked again. Versus the Chiefs, Tannehill completed passes at a rate of 7.8% above expectation (2nd-highest in the league for Week 10), and he had the 3rd-highest differential for Week 12 at +11% over expectation.

A quarterback who is completing passes at a much higher rate than “expected” is either playing the game at a very high level in terms of ball placement and decision-making or receiving a boatload of support from his pass catchers making contested catches and coming down with balls that would generally be expected to fall incomplete (given the “completion probability formula”). Likely both.

Tannehill leads the entire NFL in completion-percentage-over-expectation through 12 weeks with a rate of +9.4%. Kirk Cousins and Drew Brees are next on the list, checking in at +7.2% and +7.0%, respectively. Russell Wilson and Dak Prescott round out the top 5. This is a much stronger group to be part of than the leaders in “aggressiveness percentage.”

And this isn’t the only stat where Ryan Tannehill leads all NFL passers. As Mike pointed out, Tannehill is also the league leader in yards-per-attempt at 9.2. And according to Gil Brandt, Tannehill is also leading the league in passer rating as a starter (which conveniently ignores his play in Denver, but that’s life I guess).

Outside of his fumbling issue, it’s hard to find much fault with how Tannehill has performed since taking over as the Titans starting quarterback. Ben Baldwin frequently plots Expected Points Added (explained in detail here), and he found that the Titans EPA-per-dropback is third-best in the league since Tannehill took over as the starter.

Conclusions

The Titans had an historic day of offense against Jacksonville. The volume output was obvious, and looking deeper, it seems the efficiency was on par with that overall production.

While the majority of the Titans points (28) were scored in a 6-minute-and-16-second span, the drive success rate and yards-per-drive numbers indicate a complete performance beyond the early third quarter success.

I first started these write-ups because I believed the Titans were scoring points despite their offensive performance. In Weeks 7 and 8, this largely proved to be true based on the efficiency statistics.

That has since changed. The Titans are scoring points in their last three contests with efficient offensive production. I was concerned the point totals achieved against the Chargers and Bucs would be unsustainable. That concern has shifted; I now wonder if the team can continue to perform at such a high level of efficiency.

Based on Arthur Smith’s improvement, the offensive line’s new-found cohesion, Derrick Henry’s December track record, and Ryan Tannehill’s performance thus far, I’m inclined to believe this type of output is sustainable.

Mike Vrabel has often spoke in his press conferences about a team’s need to improve throughout the course of the season. I think we’re seeing this Titans team improve each and every week.

Games like this one make a strong case for each of Ryan Tannehill, Derrick Henry, and Arthur Smith to return next season. If the offense can maintain this level of play, there will be no doubts.

James and I recapped our thoughts from this game on our latest pod, recorded Monday night. You could listen to your family argue about something (probably politics), or you could listen to us talk Titans!

Happy Thanksgiving!