Negative plays, the concept of first-down runs, and Bill O’Brien’s Texans

Your Houston Texans are having, compared to 2018, a highly effective season running the football. They’re averaging 4.8 yards per rushing attempt, which puts them fourth in the NFL. In 2018, that number was 4.3. Both numbers see a Deshaun Watson boost (Watson averages 5.7 yards per attempt for his career), but the Texans have clearly improved from 2018, where they struggled to run out of base formations for whole games at a time at the end of the season.

However, because of the shape of what running the football looks like for most teams, how much Houston cares about the running game — and in particular on first down — anchors their offense. The Texans improved from 15th to eighth in DVOA this week on the heels of Watson’s four touchdowns against a Patriots defense that was playing historically well. The ceiling of the pass offense has been improved by all the additional weapons that they’ve brought in. With a 30.3% pass DVOA, the Texans are seventh-place and not far out of the top five.

The Texans are one of two teams with a top-8 offensive DVOA that has a negative run offense DVOA. The other, Kansas City, has run the ball 269 times to Houston’s 326. Critics of Bill O’Brien tend to revert to the idea that he plays too simple of a pattern: run, run, slant. That’s not necessarily true. The Texans did run-run-pass last year only 18 percent of the time — a high amount, but not outlier high like Seattle was.

But this was a massive problem against New England, where Carlos Hyde had just 10 carries for 17 yards. Often, the concept of “staying on schedule” is invoked in NFL circles. For O’Brien, he most often harps in press conferences about the idea of avoiding negative plays and mistakes. “We don’t have a play for first-and-30,” to paraphrase something he said about the Jaguars game in London. Where we have a ways to go in the analytics world is finding a way to make coaches understand that failed rushes in a non-clock killing situation are mistakes.

Last week Bill Belichick posted up eight in the box 30% of the time for Hyde, a much higher number than he was used to seeing. (Over the full season, that number is 14.1%.) He played to keep the Texans from running on first down because it is fairly evident to anyone who watches games that this is what O’Brien likes to do. O’Brien changed nothing, barely ran anything that would occupy a defender like a read-option, and the run offense died on the table. It was only for the grace of Watson and Duke Johnson that Houston’s pass offense was able to save the fact that it was handed the ball on second down with an average of 8.75 yards to go for the entirety of the game.

The top five teams in terms of running the ball on first down are Baltimore, Oakland, Dallas, Houston, and San Francisco. Baltimore and San Francisco have unique, well-documented run games — one has Lamar Jackson and a pile full of Greg Roman schemes, while the other has Baby Shanahan’s impressive set of tactical advantages. The Raiders have Derek Carr at quarterback, so they have little choice but to run the ball. The Cowboys … well, I hope if you’re reading this you know that as an organization they have decided to prove that signing Ezekiel Elliott to a huge contract was worth it even though it wasn’t.

The other thing those four teams have in common: they actually run for positive DVOA on those first downs. That is something the Texans don’t do, and something that gets even more extreme when you split out recent games:

It’s also something that is borne out in multiple years of data. The Texans ran 278 times on first down in 2018 and had a -25.2% DVOA on those carries. (The only teams that ran more were Seattle and Baltimore, and each only by two rushes.) The Texans ran 267 times on first down in 2017 and had a -14.1% DVOA. (The only team ahead of them that year was Minnesota, at 276.)


I am not quite as anti-run as a lot of my fellow analytics disciples. At the end of the day, what you are trying to do in a football game is play to where defenders aren’t and play to where your strengths are. Sometimes, as it was against Kansas City’s woeful run defense, that will mean you run the ball a lot more.

But the Texans have been donating downs away under O’Brien’s watch for seasons, and since we just spoke Duke Johnson targets into existence last week, let’s speak this one into existence too: Those donated carries don’t always have as much negative yardage as a sack does, but they continue to put the passing offense in situations that are trickier than they should be. Worst of all, the runs themselves are not particularly interesting. I make it a point to pick out O’Brien’s exotic designs with Watson and satellite receivers and praise them. They were a big part of the Kansas City game plan. Whereas the Ravens and 49ers are running on first downs with a big schematic advantage, a lot of the donated downs Houston runs are just simple inside zone.

Red zone stalls are often part and parcel with donated plays. Houston’s first field goal drive against the Colts in Week 7 started out with a failed run and a false start. Their second? Failed run on first down. We focus in on Watson’s sacks because they are loud. Comparatively, these little runs for no gain are silent killers that set up the sacks.

With coaching, it’s always the little things. If O’Brien ran the ball four or five times less on first down a game, this wouldn’t be an issue. That’s all it really amounts to. But the more stagnant and easy to read an offense is, the easier they are to defend. Right now, everybody knows what is coming on first down. Even the fans.


Iā€™m happily writing this article free of charge ā€” this is a labor of love as I am between Texans gigs. This is presented to you ad-free and without any hassle. If you enjoy my work and want to encourage me to produce more, please feel free to leave me a PayPal tip.

Four Downs: Texans 28, Patriots 22

The AFC glass ceiling shattered on Sunday night, and Texans fans were free to envision something more for the first time.

Houston bullywhipped the New England Patriots, long-time Texans tormentors, in a way that the final scoreline doesn’t even completely address. Houston achieved a 95% win probability (per ESPN’s in-house metric) towards the end of the third quarter, and turned the tables on the Patriots by decisively winning the battle of the game plans, getting around tight coverage with annoying short completions, and frustrating the opposing quarterback into a series of annoying passes. They Patriots’ed the Patriots. And you can tell this game meant a lot more to the Texans than they’d ever let on this week.

Summing up what this game meant to the collective fanbase is weird in the same way that any thing that destroys self-perception is. This game is the moment you dropped below 200 pounds for the first time in eight years. This game is the first time you asked someone out and they said yes. There are plenty of Toro-colored glasses out in Texans Internet land, and they did go to Arrowhead Stadium earlier this year and win, but the Texans have been Little Brother to the Patriots ever since they became relevant, especially since O’Brien has been in town. Big Brother finally stumbled.

My pragmatic side wants to tell you that this does not necessarily mean much. It wants to tell you that the Patriots are still heavy favorites to host a return meeting between these two teams if it happens. It wants to tell you that Bill O’Brien has often come up with amazing play designs and has called good game plans before, but that it’s usually a tease rather than a trend. It wants to tell you that New England’s passing attack has looked broken since Week 5 and that this played into that. It wants to tell you that the real threat is roughly 400 miles west of Foxboro, where the Texans got spanked 41-7.

But it does matter. It matters because we were able to see it. The vision that Bill O’Brien has sold his bosses as Patriots South has always been a fraudulent-ass one that relied on closing your eyes any time the Texans played a real team. They haven’t closed the deed on this season yet, and they’re still not likely to grab a first-round bye. But when you watch this game, and the Kansas City game, you are able to see it.

1 — Deshaun Watson, Duke Johnson, and short game dominance

The mantra all week from the Texans was about avoiding turnovers and playing mistake-free football. They didn’t always do that as a team — they had penalties that set back the cause, and two of Watson’s three sacks taken were from almost entirely unaccounted for rushes:

But what we’ve seen when the Texans have cruised this season hasn’t been that they need to reinvent the wheel on offense, it’s that they have so much skill position depth that all they need to do is have Watson get the ball out and go on with their day. Watson took three sacks, but only four total quarterback hits on those sacks. He was 14-of-18 for 135 yards and two touchdowns on passes within 10 yards of the line of scrimmage. If you go through the ones that weren’t caught — one was defensive pass interference, one was a dropped Kenny Stills slant — they’re not exactly plays that show poorly on Watson.

It’s me, I’m that idiot who said that Duke Johnson was the important player that needed to be involved. I get to be right about things sometimes.

When Watson is operating the short-area game well, the Texans feel unstoppable and inevitable. They were so inevitable in this game that they donated 10 Carlos Hyde rushes to charity (1.7 yards per carry, long of four yards) and still averaged 7.7 yards per pass.

The play where Watson avoided a Kyle Van Noy sack on first down, throwing the ball away, and then found Jordan Akins on a tackle-breaking run after catch to get out of second-and-10 on the second scoring drive exemplifies what’s going on here. The Texans aren’t necessarily avoiding hurting themselves with penalties or bad plays — because every team does this to some extent — but when you throw for 7.7 yards per attempt it doesn’t really matter if you mix in a negative play or two.

2 — The emergence of Jacob Martin

Not only did the Patriots look limited as a passing offense, they looked limited in a way that relied on Tom Brady buying time. I don’t think any of their older receivers besides Edelman looked 100%, and I don’t think either of their younger receivers showed themselves capable in this game. There is no tight end play.

But the Texans also did this by bringing heat — they sacked Brady three times, but they hit him 12 total times. Brady was constantly being harassed. His average time-to-throw was a startling 3.4 seconds, which points to both the trouble his receivers had getting open and how often he had to reset his throwing point.

Nobody sent Brady fleeing as often as Jacob Martin on the edge — the Patriots simply couldn’t give Marcus Cannon enough help on the outside, and Cannon was watching Martin go by him on nearly a by-drive basis.

Brennan Scarlett and Angelo Blackson were missed in the base run game — more on that in a moment — but those extra snaps that Martin got showed us some flashes of how the Texans might possibly be able to recoup a little value on the Jadeveon Clowney trade. I don’t think Martin is some sort of burgeoning superstar — he’s not leaving guys in the dust snap-after-snap or anything like that — but I do think he has enough speed on the edge to be an effective complementary rusher. Like Whitney Mercilus, he does a lot of his living on the initial get-off. Cannon couldn’t deal.

3 — The defensive game plan that the Patriots couldn’t counter off of

The Texans — Romeo Crennel — came into this game with a game-specific plan that actually worked. They decided to force New England’s non-Julian Edelman and James White receivers to beat them in man coverage, and those receivers simply couldn’t do it.

On 24 targets to non-Edelman and White receivers, Brady completed 10 balls for 122 yards and no scores. A vast majority of those balls came on New England’s final three drives of the game, after they were down 21-3.

The Texans were able to create pressure off of guarding White with a DB, and they were able to halt Edelman’s routes to the inside with doubles. Edelman’s 44-yard catch came on a deep in with a picked-up blitz that happened roughly six seconds into the down. Outside of that catch, he had almost nothing happening deep.

When the stakes were their highest, the Patriots went to Mohamed Sanu to try to convert on fourth-and-short. The Texans smartly(!!!) stacked the line, forcing Brady away from the sneak. It was Sanu on Johnathan Joseph, and Joseph was able to break the ball up even if he couldn’t quite hang with Sanu on the initial play:

The amount of situational things that the Texans accounted for in this game that they normally don’t was staggering to me. Maybe it goes blind and unaccounted for and we see it brightly here because it was such a big game, but I’m positive I’ve seen Brady sneak past the Texans for first downs on a regular basis. It was wonderful to see some actual opponent-based game planning that worked. If that sticks throughout the season, it’s cause to praise the coaching staff.

4 — Bradley Roby’s game-script shattering interception and the ensuing touchdown

So of course, Roby’s interception (and near pick-six) was enormous. The Patriots were up 3-0 at the time, and the Texans pounced on that to turn it into seven points when they hit Duke Johnson on third-and-3.

But even more than that, against a Patriots offense that had major issues but could absolutely run the ball in this game, it forced a negative game script from the very beginning. Look at Joseph in run defense on this play:

The Patriots wound up running for 145 yards in the game even though they were always behind. With Scarlett and Blackson out, and the defense up front stocked with guys like Joel Heath, Barkevious Mingo, and Eddie Vanderdoes who had seen very little in the way of playing time, the Patriots were able to pick and choose their way to success in the run game.

You can easily imagine a scenario where the Texans play field position with the Patriots for another couple of plays, then the Pats hit a big run or two, and go up 10, and time is a big ally for the Patriots. That’s why the Roby interception was so big — it wasn’t just the specific purposes of points off turnovers, winning the turnover battle, the short field — it was that it kept the Patriots from executing from a positive game script. If they could have banged away with Sony Michel all game, we might have been looking at a very different final score.

Roby’s pick was, quite frankly, one of the most important plays of the season. It might wind up being one of the most important plays in Texans history if this season continues an optimistic trek. So much hinged on him reading N’Keal Harry’s route and running it for Harry. Even down starting center Ted Karras, the Pats were able to bang away on the Texans.

But because of Roby, they were never able to use the run game as a true weapon — it was a change of pace for three-fourths of the game.


Iā€™m happily writing this article free of charge ā€” this is a labor of love as I am between Texans gigs. This is presented to you ad-free and without any hassle. If you enjoy my work and want to encourage me to produce more, please feel free to leave me a PayPal tip.