“Establishing the Run” Is A Myth (Part I)

What is “establishing the run”? The concept has been long instilled in NFL history. Yu run the ball to set up the pass. Running down a team’s throat forces them to put more defenders in the box to stop you, leaving less defenders further down the field to defend the pass. The concept of run, run/run/play action pass should be a fairly common thought. But does establishing the run actually achieve what it hopes to? I wanted to figure out this enigma and tracked every single Texans first down, searching for an answer. The findings shocked me.

Throughout Bill O’Brien’s time with the Houston Texans, he has garnered the reputation of a “run heavy” offense, especially on first down. I will admit to hating this about his scheme, pleading for less first down runs and more passes instead. I, and many other Texans fans, have been irate with the predictability of the first down offense, but have our concerns been warranted? While I simply couldn’t torture myself and watch every single first down since BOB was hired in 2014, I did chart every Houston first down play from the 2019 season.

The first question I wanted to uncover was the first down run/pass split. The answer surprised me. On first down, Houston ran the ball 51% of the time and thus passed the ball 49% of the time. This near perfect 50:50 split just couldn’t be true, right? I’m afraid it is. We passed on first down a whole lot more than you would think.

Diving further into this, out of Houston’s 497 first downs, they ran it 255 times and passed 242 times. A 13 rep disparity, a small enough number that we shouldn’t even be able to diagnose a difference. Yet the consensus is Houston run an unreasonable amount on first down. If that theory doesn’t match reality, why do we believe it? I thought it was because of a lack of efficiency from these runs, but statistically, that’s just not the case.

I tracked Houston’s first down Yards per Carry (YPC) and it was an astounding 4.72! That’s a ridiculous number and should set the Texans up for great success on the remaining downs. In comparison, Houston’s first down Yards per Pass (YPP) was just 7.34. While that is clearly a larger number, at least from watching the Texans’ games last season, one would think there would be a much larger difference between yards per carry and yards per pass.

So we’ve determined that the Texans’ run/pass tendencies are a lot more reasonable than we initially thought and that the efficiency of first down runs was eye opening. More work needs to be done to truly understand “establishing the run,: As I mentioned in the opening paragraph, the whole idea is to force defenses into higher box numbers, to take advantage of that by passing instead of running. When running the ball on first down, the Texans saw an average of 6.8 defenders in the box versus 6.3 when passing. Therefore, even though we were an efficient first down running offense (4.72 YPC), we didn’t scare defenses into loading the box to try and stop us. There was actually a negative effect. Furthermore, the average box numbers didn’t vary much on a quarter by quarter basis, as seen below.

What does this really mean? Well, one would expect that with such an efficient first down run game, defenses would look to stack the box more and more throughout the game. But they did not care to do so against the Texans. There was no fear when opposing teams saw Carlos Hyde up the middle. They actually put less defenders in the box, inviting the run.

Not convinced “establishing the run” is a myth? That’s fine; I wasn’t at this point in my research either. I looked further into Houston’s passing game to see if the run game truly had any effect. I tracked the completion percentage of regular passes and play action passes, and this is where I was truly amazed. A big piece of “establishing the run” is being able to play action fake, suck in the defense, and burn them over top with a pass. Were we able to do that? Yes, but how much of that was due to the run game?

The Texans’ first down completion percentage of regular passes was 69%, while our play action completion percentage was just 63%. With such an efficient first down run offense, you would expect our play action game to be much more effective; just think about how greatly it helped the Titans and Ravens. While the completion percentage between regular and play action passes actually went down, the yards per attempt (YPA) wasn’t even close. Regular pass YPA was a putrid 5.95, while play action YPA was 14.0. So the completion percentage and YPA tell different stories, but the tape showed me all I needed to know.

The completion percentage was lower for play action simply because majority of the time, defenders wouldn’t bite on the play action; they had no fear of the run game. Look on this play how the linebackers barely take a small step forward on the fake and then are immediately looking for the pass.

The play action pass results in a completion and 15 yard gain, but the run fake did nothing to help. This exact play helped me understand why our play action completion percentage was lower while the yards per attempt was so much higher. It’s simply a scheme thing. Our play action passes were aimed towards the intermediate and deep parts of the field, whereas our regular passes were looking to get the ball out quick and thus had shorter routes. Look at this play against the Chiefs, where we only have one route that goes past five yards.

This quick passing offense was a theme for much of the year. As painful as it was to watch, it was also quite inefficient, with just 5.95 YPA on first down. If Houston is only getting about a yard more (5.95 YPA - 4.72 YPC) from these regular passes versus a run, why not just run the ball, where there is less risk of interceptions or sacks?

A big reason for this inefficiency was a lack of yards after the catch. Houston’s receivers weren’t good in that department, but getting the ball to Duke Johnson or David Johnson would be wiser going forward. Duke consistently got yards after the catch and was a monster in space. Hoping for the same with David.

Ideally, the Texans would be wise to trust in Deshaun Watson, the offensive line, and the wide receivers to move to a vertical passing offense, targeting more of the intermediate and deep parts of the field even when they aren’t running play action. I have confidence O’Brien will do this in 2020 because we saw this type of offense whenever Will Fuller was healthy. It disappeared when he was injured, but with Brandin Cooks on the roster now, there’s an insurance policy. If O’Brien would just trust Kenny Stills to play Fuller’s role, there would be two options to stretch the field.

Check back tomorrow for Part II.