There are two tools we have to analyze the game of football. We have numbers, and we have video. One isn’t better than the other. They each have their limitations. There isn’t a partisan divide between the two. All you can do is combine them to understand a complicated game filled with 22 bodies colliding into each other. For the 2017 season, this is what those digits said about the Houston Texans.
Regression
Surprise! The Texans regressed to the mean this season. In 2016, they overperformed their expected win total. They won 2.5 more games than expected, thanks to an 8-4 one-possession record, an easy schedule, and a top ten defense. In 2017, they fell all the way to four wins despite having a nearly identical DVOA. In 2016, Houston finished 28th in DVOA at -20.3%; this year they finished 29th with a DVOA of -21.4%.
Houston ‘Lucky’ Stats
Year | Wins | Prior Year Wins | Plexiglass | One Poss Record | Turnover Diff | Expected Win Loss | Actual-Pyth | SoS | DVOA |
---|---|---|---|---|---|---|---|---|---|
Year | Wins | Prior Year Wins | Plexiglass | One Poss Record | Turnover Diff | Expected Win Loss | Actual-Pyth | SoS | DVOA |
2016 | 9 | 9 | 0 | 8-4 | -7 | 6.5 | 2.5 | -5.90% | -21.40% |
2017 | 4 | 9 | -5 | 1-5 | -12 | 5.7 | -1.7 | 1.30% | -20.30% |
What changed? First off, Houston went 1-5 in one score games. Their only win in a one score game came against Cincinnati on Thursday Night Football because of a Deshaun Watson scamper. With Watson, they lost three additional one score games to New England, Kansas City, and Seattle. Without Watson, they only lost one close game...because their per game point differential was -14.7 points without DW4 under center. Watson was able to score points and carry a bad defense, still falling short half of the time. The Tom Savage and T.J. Yates versions shouldn’t have even suited up.
The other change is that Houston played a tougher schedule in 2017 than they did in 2016. In 2016, Houston’s opponents had an average DVOA of -5.9%. This was the 31st toughest, or second easiest, schedule, depending how you want to read it. The AFC South was fun, but it was bad in 2016. The Texans went 5-1 against the AFC South and 4-6 against the rest of the league, and the rest of the NFL they played was the AFC West, New England, and Cincinnati. This year, Houston’s average opponent had a DVOA of 1.3%, and they played the eleventh toughest schedule. Without the one possession fortune and a more arduous set of games, Houston dropped off.
The Texans measuring out in these aspects isn’t surprising at all. It was expected. The good news is things should be better in 2018. Nothing went the Texans’ way in 2017. They dropped off by five wins. They won 1.7 less games than expected. They went 1-5 in one possession games. They were 28th in turnover differential. They played a tougher schedule. Next year they’ll get a comfier slate, and they’ll have more than $63 million to improve the roster this offseason. That, plus what should hopefully be a healthier roster, should lead to the the Texans bouncing back to compete for a playoff spot again in 2018.
Offense
Ten Texans’ games were identical to the slop heaved around in 2016. Lots of inside runs, worthless tosses into the flats, and force-feeding DeAndre Hopkins. There were also six games where the Texans’ offense was a revelation, an offense filled with deep play action passes that utilized a variety of different play fakes set up by a more diverse run game. Not coincidentally, there were ten games where Tom Savage or T.J. Yates started, and there were six games Deshaun Watson started. The difference between the two was the size of an ocean.
Houston’s Offense With Deshaun Watson
W/ & W/O | Points Scored | OFF DVOA | Pass DVOA | Pass Yards | Y/A | Run DVOA | Rush Yards | Y/C |
---|---|---|---|---|---|---|---|---|
W/ & W/O | Points Scored | OFF DVOA | Pass DVOA | Pass Yards | Y/A | Run DVOA | Rush Yards | Y/C |
W/ Watson | 208 (34.6/G) | 13.90% | 44.20% | 1,494 (249/G) | 8.3 | -8.20% | 875 (145.8/G) | 4.38 |
W/O Watson | 130 (13/G) | -23.10% | -29.50% | 1,784 (178.4/G) | 6.04 | -15.06% | 967 (96.7/G) | 3.87 |
With Watson starting, Houston scored 34.6 points a game and had an offensive DVOA of 13.9% (44.2% passing and -8.2% rushing). Without Watson, Houston scored 13 points a game and had an offensive DVOA of -23.1% (-29.5% passing and -15.06% rushing). By comparison, in 2016 with [NAME REDACTED] starting at quarterback, Houston had an offensive DVOA of -21.2% (30th) comprised of-19.5% (30th) passing and 19.1% (27th) rushing. Somehow, some way, the Texans’ Week One offense became an even worse version of 2016’s nausea.
Watson scored more points starting six games then Houston’s offense put up in the ten games he didn’t start. The biggest difference was Watson threw a lot of touchdowns. He threw 19 of them on only 204 attempts. He had a touchdown rate of 9.3% (more on this another time). Since 1990, only Peyton Manning had a higher touchdown rate when throwing at least 200 passes. If you go back to the beginning of professional football, Watson has done something that has only been accomplished 16 times. He joined a group with names likes Sid Luckman, George Blanda, and Ken Stabler.
The other difference was Watson could actually throw the football down the field. He averaged 8.3 yards an attempt. Savage averaged 6.3 yards per attempt and Yates averaged 5.4. Watson completed 22-53 passes classified as deep for 649 yards, seven touchdowns, five interceptions, and averaged 12.2 yards an attempt. It took Yates and Savage 71 attempts to complete as many deep passes. They averaged only 9.02 yards an attempt, with four touchdowns and four interceptions.
Where Watson had the most success was throwing to the deep middle part of the field. He completed five of his eight attempts there for 182 yards and three touchdowns, averaging 22.8 yards an attempt. He was also great to the right sideline, completing 11 of 24 for 327 yards and four touchdowns. The one spot downfield Watson struggled with was to the deep left sideline. There he completed only 6 of 21 passes for 140 yards, 6.7 yards an attempt, and he threw three interceptions against zero touchdowns. Both Yates and Savage had more success feeding Hopkins to this opposite sideline.
As great as he was at throwing deep, Watson was even better throwing short. He completed 104 of his 150 attempts (69.3%) for 1,050 yards, twelve touchdowns, and three interceptions, averaging 7 yards an attempt. DeAndre Hopkins caught 75 passes for 795 yards and five touchdowns on short routes, including four touchdowns on just seven catches in the short middle part of the field.
The two other facts that jump out when it comes to receivers are Stephen Anderson and Will Fuller V. Anderson was the Texans’ best tight end. He caught only three of his twelve short middle targets, but snagged all five targets to the deep right part of the field. Will Fuller V is an efficiency monster. He caught 28 of his 50 targets, picked up 15.1 yards a reception, scored a touchdown on 25% of catches, and had a DVOA of 17.5%. When discussing injuries and quarterback play, Fuller’s name is mostly forgotten. But if he ever gets a full healthy season with Watson, he could put together a 1,000 yard, 10 touchdown, Pro Bowl season.
Aside from their subconsciousness being linked together and walking the Earth itself, playing under pressure is the one thing all Texans quarterbacks shared. All three were playing quarterback with a red cape, elongated hat, and sequin adorned Chaquetilla. Watson was sacked on 8.5% of his drop backs, Savage was sacked on 8.6% of his, and Yates was sacked on 11.8% of his. Watson was able to deal with it a bit. He avoided nine sacks. He could also escape , scamper for yards, dump it off in the flat after breaking a tackle, or switch hands before heaving it deep.
The other two QBs couldn’t. Overall, the Texans allowed 54 sacks, had an adjusted sack rate of 9.2% (30th), and a pressure rate of 37.9%, which ranked last in football.
The offensive line had twelve players take a snap. Breno Giacomini played 1,102 snaps (100%), Xavier Su’a-Filo played 1,032 (98.2%), Nick Martin played 975 (88.5%), Jeff Allen played 732 (66.4%), Greg Mancz played 565 (51.3%), Chris Clark played 549 (49.8%), Juli’en Davenport played 241 (21.9%), Chad Slade played 167 (15.2%), Kendall Lamm played 160 (14.5%), Kyle Fuller played 90 (8.2%), Duane Brown played 68 (6.2%), and David Quessenberry played 28 (2.5%). The Texans also used seven different starting offensive line configurations. Rick Smith had multiple failures last offseason. Going with Tom Savage as the starting quarterback and having Kendall Lamm start at left tackle may have been the dumbest of them all.
At first glance, the Texans were better in the run game. Houston was average at running the football. They were 16th in yards per carry with 4.0 and ranked 21st in run offense DVOA. They were better than the worst in the league, but Watson was a large reason why they had any success. With Watson, the Texans averaged 4.38 yards a carry, compared to 3.87 without him as mentioned earlier. There was also a 6.86% difference in DVOA between the two offenses as well.
The only thing Houston did mediocre was hold the first level of the line of scrimmage. They were stuffed on only 21% of their runs. Overall, they were the same old Texans. They focused on middle runs and didn’t break anything in the second level or open field. This season, Houston ran the ball 258 times up the middle, 122 times over the tackles, and 57 times to the edges. On these runs, Houston picked up 4.23 yards up the middle, 3.57 over the tackles, and 4.17 on the edges. They had only 0.99 adjusted line yards in the second level (25th) and 0.51 adjusted line yards in the open field (26th).
The middle runs were also helped out by Deshaun Watson. That great draw play was the best part of their interior run game. Watson picked up 10.43 yards a carry on his 14 attempts when running up the middle. Lamar Miller gained 4.03 on 148 attempts, D’Onta Foreman gained 3.4 on 41 attempts, and Alfred Blue gained 4.2 on 45 attempts.
The worst aspect of the Texans’ run offense was on the left edge. They ranked last in adjusted line yards with 1.29. They picked up 78 yards on 29 carries, which comes out to 2.69 yards an attempt. Watson picked up 46 of these yards on six attempts while the rest combined to get 1.5 yards a rush.
It wasn’t all on the offensive line. The running backs didn’t do enough on their end. Lamar Miller broke tackles on just 12.4% of his touches. Alfred Blue broke tackles on 12.0% of his. Before he was lost for the season, Foreman broke tackles at an outstanding and surprising 24.6%. This is the part where we complain about Miller. There isn’t a reason to pay a running back $6.75 million for him to get 3.7 yards a carry, plunge his head into the middle of the field over and over again, and be a positive receiving back. Miller doesn’t fit this team unless Bill O’Brien mixes up the core of his rushing attack. Houston would be better off paying anyone else to close his eyes and fall forward. This isn’t to say Miller is a bad running back. He’s just a bad back in this offense.
All of this is a very long winded way to say the Texans’ offense was really good with Watson, really bad without him in every aspect of their game, and the offensive line bound both of these offenses together with abysmal play.
Defense
The Texans’ defensive personnel strategy entering the 2017 season was that Kevin Johnson would be healthy and step up to fill in for A.J. Bouye, a healthy J.J. Watt, Whitney Mercilus, and Jadeveon Clowney would make the pass rush great enough to cover for any secondary hiccups, and Benardrick McKinney and D.J. Reader would be able to stop the inside run.
Only the last thing happened. The secondary was terrible, Watt and Mercilus were injured five weeks into the season, and Houston was good at stopping the run. But even with a healthy Watt and Mercilus, it’s far from certain the Texans’ defense would have been good. In the five games before they got hurt, the Texans had a defensive DVOA of -15.54%, allowed 26 points a game, and had only eight sacks. The DVOA is lifted like holy goodness by an effort against Tennessee where they had a DVOA of -59.3% because Mike Mularkey smashed his mouth and bit down onto a rock wall. The Texans had an average DVOA of -3.68% and an overall DVOA -18.4% if you remove that first Titans game.
Without Watt and Mercilus, this unknown defense was terrible. They had a DVOA of 13.6% and allowed 27.8 points per game. One of the things that was pestilent was the pass rush. They picked up only 24 sacks without Watt and Mercilus. The pass rush wasn’t great with them already. Aside from a Whitney Mercilus strip sack, they really weren’t getting to the quarterback. Without these two, only Jadeveon Clowney was able to get to the quarterback. He had 42 pressures and 9.5 sacks. Watt played in four games and some change, yet he was second on the team in pressures with 15 and never registered a sack. Nobody else had even ten (!) pressures. McKinney and Reader were tied for third with 9. Over the entire season, the Texans had a pressure rate of 27.8% (28th) and an adjusted sack rate of 6.2% (21st).
With this nonexistent pass rush, teams had all the time in the world to pick on the Texans’ secondary. Johnathan Joseph, Kevin Johnson, Kareem Jackson, Andre Hal, and Marcus Gilchrist were the primary starters on an awful pass defense. Houston had a pass defense DVOA of 19.2% (25th), allowed 7.1 yards per pass (30th), 30 touchdowns (29th), and 3,799 yards (24th).
The worst part of the pass defense was stopping teams from throwing downfield. They had a DVOA of 52.4% on deep passes, which ranked 31st. They allowed 49 catches on 103 targets, 1,506 yards, 14.6 yards an attempt, 10 touchdowns, and a quarterback rating of 101.9. The worst of which was the deep middle part of the field. There, Houston allowed 18 catches on 30 targets for 637 yards, seven touchdowns, 21.2 yards an attempt, and had a DVOA of 139.6%.
Against shorter passes, the Texans were better but still gross. They were fully cooked liver in bacon grease instead of slightly pan fried liver—still bloody, and slightly green. They had a pass defense DVOA of 2.6% (24th), allowed 272 catches on 380 attempts (71.6%), 2,482 yards, 20 touchdowns, picked off five passes, allowed 6.5 yards an attempt, and quarterbacks had a similar rating of 101. They struggled most covering short middle and short left passes, but they were pretty good at covering the short right third of the field. Kevin Johnson and Kareem Jackson couldn’t cover short passes out wide; Johnathan Joseph was good at it, but got beat deep too often. Quarterbacks were able to find holes in the short middle zone.
The Texans’ defensive backs performed poorly, from Marcus Burley to Marcus Williams, according to Football Outsiders’ Premium Charting Data. What matters is the Big Three. Joseph was targeted 65 times, allowed 9.6 yards a pass, had a success rate of 60%, allowed 3.2 yards after the catch, and the average attempt against him traveled 13.4 yards through the air. Jackson was targeted 77 times, allowed 8.1 yards a pass, had a success rate of 52%, allowed 1.4 yards after the catch, and the average attempt traveled 11.1 yards through the air. Johnson was targeted 47 times, allowed 9.2 yards a pass, had a success rate of 45%, allowed 3.0 yards after the catch, and the average attempt traveled 9.8 yards through the air. Johnson not only couldn’t cover; he was an atrocious tackler as well. He missed 27.7% of his tackles. Jackson, on the other hand, has kept that part of his game alive and strong. He missed a tackle only 15.8% of the time.
Houston’s plan to maintain a top ten pass defense failed. Watt and Mercilus were injured. They couldn’t even find a secondary pass rusher without them. Johnson wasn’t able to come close to replicating Bouye’s performance, which was required with the front seven injuries; instead, Kevin Johnson was one of the worst defensive backs in football in 2017. Joseph had trouble against downfield passes. Jackson was picked on. Overall, it was a putrid bunch. The run defense, however, was great.
They finished twelfth in run defense DVOA with -9.9%. The middle of their run defense was anchored by Benardrick McKinney and D.J. Reader; once again, they held it down. McKinney had 46 tackles and Reader had 26. Teams ran the ball up the middle 225 times for 710 yards (3.16 yards a carry) against Houston. The Texans ranked seventh in adjusted line yards with 3.8 while stomping out these runs.
They were great at stopping left hand runs, too. Jadeveon Clowney has become the best run defender in football from the defensive end position. He had 17 tackles for zero yards or less this year. Unfortunately, as the right defensive end played these runs, teams picked up 4.29 yards a carry, but ranked sixth in adjusted line yards on the left edge and fifth over the left tackle.
Where the Texans struggled was on the opposite end, especially the right edge. Houston allowed 6.15 yards a carry, 16 first downs, and they were last in adjusted line yards with 6.04. Andre Hal was stuck picking up the pieces and trying not to miss tackles. He had 17 tackles against runs in this direction. Over the right tackle, the Texans allowed 4.63 yards a carry. Unlike the pass defense, there is no question Houston would have had a top ten run defense if Watt and Mercilus played close to 16 games.
Overall, Houston’s plan was a novel concept. It kind of, sort of made sense, but ultimately two-thirds of it failed and injuries derailed the pass defense. Even then, it is far from certain they would have had a strangling pass defense like last year. The secondary performance was that abysmal. What was great for so long became a crater in 2017.
Special Teams
No longer were the Texans in the darkest depths when it came to special teams in 2017. They made the huge leap from 31st to 26th, finishing with a DVOA of -4.5%.
On punts, they were once again miserable. According to Football Outsiders, Houston lost 10.2 expected points in the punting game. Shane Lechler is a large part of that. He had 92 punts and picked up 4,507 yards. His net average was 41.3 yards, which finished ninth in the league. But he still allowed a return on 53.2% of his punts, and only 34.7% of his punts landed inside the 20 yard line. Lechler finished highly in net consumption of the punt numbers because he had 92 of them, but that doesn’t mean he was good. The Texans’ punter and punting unit combined to once again hurt the team.
In the kicking game, the Texans’ finally found someone who could make field goals from farther than 40 yards out. Ka’imi Fairbairn went 9-11 on field goals from 40+ yards. But he missed three 3 extra points and went 11-14 on field goals from less than 40 yards. The best news is he could actually kick the ball out of the end zone, which kept the kicking unit from making mistakes on returns. 51 of Houston opponents’ 76 drives that started after a kickoff began at the 25 yard line. The opponents’ average drive started at their own 25, which is a yard better than last year over the course of the same number of kicks. Field goal numbers bounce around more than you would expect, but kickoff ability is more of a stable indicator of place kicker performance. I’m here for Fairbairn retaining the kicking job in 2018.
Once again, the Texans couldn’t return kicks or punts well. They finished with -5.3 expected points gained on kickoff returns and -1.1 on punts. On kickoffs, they finished 20th with 20.7 yards per return. On punts, they finished 17th with 8.5 yards per return. On kickoffs, Chris Thompson was the primary returner once Tyler Ervin was injured; he snagged 21.7 yards a return, including a 41 yard one. He still would have been better off just kneeling the majority of the time. The Texans had six different players return a punt and were never able to find a primary returner. Fuller V was the best of the bunch and picked up 15 yards a punt return.
To sum it up, as expected, the Texans weren’t as lucky as they were in 2016. Their entire offense performed differently with Watson as the starting quarterback. Savage and Yates were awful. The run game is still ill by the same poison. The pass defense was atrocious and it’s unknown if it would have been good even with Watt and Mercilus healthy. The run defense was great because of McKinney, Reader, and Clowney. The special teams were horrendous but slightly better than 2016.
Now all that’s left to do is take this as you want it and toss it into the future like bread crumbs into the mouths of salty rats as the Texans move headfirst into the construction of their 2018 roster.