Fooled By Randomness: Derrick Henry Isn’t Teaching The Lesson You Think
Fooled By Randomness: Derrick Henry Isn’t Teaching The Lesson You Think
I know you have heard it. “DERRICK HENRY MATTERS!!!!” “Turns out some Running Backs DO matter, nerds!”
As the Tennessee Titans shut down the juggernaut Baltimore Ravens with an emphatic victory in the divisional round of the NFL playoffs (after thoroughly dunking on the New England Patriots by scoring 14 points on offense), the anti-data crowd has reached a fever pitch of confidence that they were right all along and that the analytics community never know what they were talking about. However, as always, the data (and what it shows) does not change as a result of one or two games.
You might be asking “Why talk about randomness and RB value on a fantasy football website?” to which the answer is pretty clear. The more fooled by randomness you are, the worse you are going to be at predicting the future. Creating projections for anything (stocks, golf, weather, ANYTHING) is all about finding ways to eliminate noise and randomness and focusing on matters. There are lessons to learn from this wild run that the Titans are on, but it is very much not the lesson that the anti-data crowd would have you believe.
“Mild success can be explainable by skills and labor. Wild success is attributable to variance.” – Nassim Taleb
The best place to start is with a simple fact: Derrick Henry has always been the player that he is at this very moment in time. He did not just wake up and become a new player in Week 10. In Henry’s career before this, the Titans have finished in points scored: 25th, 23rd, and 11th (2016 in which DeMarco Murray had 293 carries to Henry’s 110). In Henry’s career, the Titans have ranked 5th, 9th, 12th and 3rd in Pro Football Focus’ run-blocking grades. Henry’s career yards per carry is 4.8 and he has never had a season below 4.2. His success rate on rushing attempts is over 50% for his career and he has never had a season with a lower than 48% rushing success rate. What I am saying should be remarkably clear: Derrick Henry has always been a good-to-great rusher running behind a good-to-great offensive line.
So why is it that the Titans have never been able to vanquish foes like the elite Ravens and Patriots before this season? The first reason is what you all know: just being able to run the ball is not that important. Passing the ball is more efficient than running the ball and teams that run the ball very well but have bad or average quarterbacks are going to be bad to average NFL teams. It is not a surprise that the Titans’ best offensive year with Henry on the team was the year that Marcus Mariota played well. The change in the Titans offense between Derrick Henry’s first 550 NFL carries (with Mariota) and his next 200 was….at quarterback. It wasn’t Derrick Henry becoming the Wolverine overnight. Derrick Henry has always been a Grown Ass Man.
Derrick Henry yards *before* first contact (per PFF)
Marcus Mariota starting: 0.5 yards (51st of 57 RBs)
Ryan Tannehill starting: 1.2 yards (30th of 55 RBs)
— Pat Thorman (@Pat_Thorman) January 15, 2020
So if we use our brains, we can probably imagine why the shift from Mariota to Tannehill made the Titans good on offense. I think a big issue with understanding what is powering the Titans right now is that Tannehill was not a good NFL quarterback with the Miami Dolphins. He took an insane amount of sacks and never led an above-average offense. He came to Tennessee and all of a sudden, he is leading the NFL in YPA. While the answer to “How did Ryan Tannehill become great all of a sudden at the age of 30” is incredibly difficult to answer, it is not as hard to demonstrate that the Titans and Henry’s sudden success is more related to how well Tannehill has performed than any other variable.
Our friend Kevin Cole from Pro Football Focus did an extremely deep dive into the Titans offense in a recent piece and discovered that in terms of points added, Henry’s playoff performance may not even be as special as it appears. As Kevin demonstrated in his recent piece on PFF “If you look at the top-12 most valuable offensive plays for the Titans over the first two playoff games, dropbacks make up the majority, even though the Titans only dropped back 39 times versus 74 designed runs...Only three Henry designed runs make it into the Titans’ top-12 plays. Henry does appear as a passer once and a receiver once, but those aren’t due to his efforts in the running game.”
So let us think about what this means just in the context of the two massive playoff wins for the Titans. Three of their four biggest plays were Ryan Tannehill passing touchdowns with (no surprise) their second-biggest play being the 45-yard play-action passing touchdown to Kalif Raymond. Without that play, it is hard to even imagine the Titans beating the Ravens. Looking at Henry’s playoff EPA (expected points added) specifically, Cole notes “Henry’s 30 carries against the Ravens generated 3.5 EPA, with 5.5 coming on a single 66-yard run. In other words, outside of that one run, his rushing EPA was negative on the day, as typical rushing attempts offer slightly negative EPA.” Again, the football-consuming public is misunderstanding the SPECTACULAR for the predictive. I am not here saying that Derrick Henry has not been awesome: he has been awesome.
The issue, as it almost always is when discussing running back value and the relationship between passing and rushing is that even in the top 99% of outcomes (that Henry is experiencing), passing is still better than rushing. In fact, a commentor on Chase Stuart’s excellent Football Perspective blog may have explained this conundrum best: “If we’re asking if the RB position matters, the answer is a definitive yes. If we’re asking if individual RB’s matter, the answer is very little.” The Titans are a huge flashpoint for this debate because of how much they use one RB versus how little they use their other RB WHILE winning games. However, even factoring in how good Henry has been as a whole and not per play, the Titans have gotten incredibly lucky to score as many points as they have in the playoffs (and to have only required 14 points to beat the Patriots).
JJ Zachariason (aka The LateRoundQB) made this point better than I could hope to on twitter, summarizing the entire Tannehill experience: Since Ryan Tannehill took over in Week 7 (including the playoffs), the Titans have scored a TD on 34.6% of their drives (2nd-most in football). They’ve kicked a FG on 3.1% of them (32nd). They’ve scored 31 red zone TDs (most). They’ve had 1 red zone FG (least).
Red zone touchdown to field goal ratios over this time period (this is legitimately insane): pic.twitter.com/dILSwBSVTL
— JJ Zachariason (@LateRoundQB) January 12, 2020
The concept of a “heater” should not be lost on anyone. If you’ve been involved with gambling or games of chance/skill long enough, you how it feels to go on a heater. We have also seen examples of NFL offenses/quarterbacks going on these heaters before. Nick Foles had a season where he lead the NFL in TD%, yards per pass attempt, and yards per completion. He would never again top 7.0 yards per adjusted attempt and is pretty much an all-time warning for small sample size theatre with NFL players. Tannehill has performed in a much different way than Foles did during that run (more reliance on rushing) but with enough luck on deep passing (544 yards and 3 touchdowns on only 34 deep attempts) that I am not ready to announce Tannehill as the savior of the Titans franchise though it is in the range of outcomes.
Why Derrick Henry Is A Lesson In Randomness, Not Re-Defining Football
Football twitter is a dramatic place and the world of football analytics is brand new. I can definitely acknowledge that some of the gatekeepers/believers of the new “analytics” put people off with how information is presented, but that doesn’t change the truth in the information itself. The individual skill of a running back really does not matter much on a play-by-play basis and really does not matter at a team level at all. Consider how often starting running backs vs their backups saw 8+ men in the box in 2018 and 2017 from Stathole Sports:
Opposing defenses are not treating Ezekiel Elliot any differently than they are treating Alfred Morris or Tony Pollard. If opposing defenses do not care who is in the backfield, then we probably shouldn’t either. As it pertains to the Titans, we also know that a huge amount of their success has been related to play-action passing. Advocates for #AllRBLivesMatter will give you a quotation on how “running sets up the pass”. This is not reflected in the data. Ben Baldwin’s work on Football Outsiders concluded that 1) previous rushes had no impact on future success of PA passes and 2) PA success remained the same regardless of which RB was in the backfield, stating “Between 2011 and 2017, 93 percent of play-action passes occurred when the offense had between one and four rushes in the previous five plays. In this range, yards per play and its standard deviation are remarkably similar for all values of previous rush attempts.”
What this all boils down to is a spike in efficiency from an incredibly physically gifted player having the best games of his career in subsequent order. This graph from Rotoviz is Derrick Henry’s splits from the last two seasons where he was the unquestioned starter for the Titans:
The 19 games played without Tannehill look a lot more like what you would expect from a running back with a good offensive line (remember in his Titans career, the offensive line has never been ranked worse than 12th in run blocking per Pro Football Focus) but who struggled to create value on his own. In fact, this season the Titans are fourth in the NFL in adjusted line yards and fifth in the NFL in “power rank” from Football Outsiders. It cannot be overstated the dramatic effect that Ryan Tannehill has had on Derrick Henry’s career. Another bit of randomness contributing to how effective we think Henry is: his performance in the second half of seasons has been better than in the first-half seasons so are minds are attributing this to some sort of “late in the season” narrative which is just that, a narrative
The last two games (and really, the entire second half of the season) have been a perfect randomness stew. The Titans faced an absurdly easy schedule, beating only two playoff teams and facing a decrepit Patriots team in the divisional round. There is no skirting the Raven’s performance as one of the most impressive of the year, though even the largest Titans/running matters truthers would admit that the Ravens ran at about 1% of expectation with fourth downs and turnovers combined. AJ Brown has averaged over 12 yards per target and scored nine touchdowns on 87 interactions, a ratio that would make Jerome Bettis blush. The team turned the ball over only 17 times TOTAL (including Mariota’s five) which was far better than league average.
All of this compounded with a running back who never has to take a breather. If the “HENRY MATTERS” people wanted to make one case that could be backed with data, the fact Henry never has to come off the field is advantage but it also brings all of a teams’ rushing numbers to one player (similar to how Dallas operates) which has a sneaky impact on the perceived value of that teams rushing game. No one argues that Tevin Coleman or Raheem Mostert is the reason for the 49ers’ success, yet every football fan or pundit is quick to point out how special and successful the 49ers running game is.
Has Derrick Henry had good games? Yeah. If the Titans had Ryan Tannehill, turnover + 4th down luck and Bo Scarbrough, could they have beaten the Patriots and the Ravens? The answer likely doesn’t change based on who the running back for the Titans is. A good rushing offense does not have the ability to be as good as an explosive passing offense. This is why, in the biggest sports market that exists in the whole world (NFL playoff games), the Titans have traded as underdogs to good passing offenses. They are 7.5 point underdogs to the Kansas City Chiefs who rank 20th in the NFL in yards per carry and 23rd in total rushing yards because they have the best passing offense that we may have ever seen.
While it feels counter-intuitive to watch Henry stiff-arm safeties and never be tackled on the first try and say “this outcome is random”, that is true of many things in life. Our brains are designed to see patterns, to see the spectacular and believe. When we dive a little deeper, however, we can attribute much of this insane Titans run to Ryan Tannehill, play-action success, turnover luck, and above-average defense.