:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/53472697/504209780.0.jpg)
I’m done. This is the last thing I have for this year’s season review. Now, usually for this last article, a Benthic plunge into the quantitative information surrounding the Texans’ season, I would submerge thousands of words on the offense, the defense, the special teams, and those ‘extra’ numbers. But I don’t have much more to say numerically about the team’s performance. I wrote about it all season, especially during the postseason. At this point, it’s all second helpings of cold regurgitations. Life and time is too precious to scour over the same things.
We all know on offense that Brock Osweiler was the worst quarterback in the NFL to start the majority of the season, and that once he stopped throwing the deep ball he became worthless. We know the run game had a lot of yards because they ran a lot but they weren’t efficient or good at it. We know Lamar Miller was wasted by being forced in between the tackles and ran the ball well outside, especially around Duane Brown. We know Jeff Allen was terrible and ruined the right side of the offensive line. We know the offensive line was average at pass blocking. We know DeAndre Hopkins had a sad season. We know C.J. Fiedorowicz having a mediocre year was the only bright spot in the receiving core. We know Houston had one of the worst offenses in the NFL.
Defensively, Jadeveon Clowney was an elite run defender. Teams ran the ball well against Houston as long as it wasn’t at Clowney or up the middle. A.J. Bouye was a revelation and was one of the best corners in the league. Despite having a mediocre pass rush, the Texans still had one of the best pass defenses because of their secondary. That said, covering running backs was a problem. The majority of their pass rush came from Clowney and Whitney Mercilus. They had one of the best defenses against play-action. Benardrick McKinney and Brian Cushing blitzing were the only real sources of interior pass rush.
Oh, and of course the special teams was last in DVOA again. Even as the calendar changes every year, there are some things that will forever stick.
Instead, in this post, I want to look at the peripheral stats, those high variance numbers that teams historically don’t have year to year success at, those numbers that better measure performance than W-L record. At the midway point of the season, I wrote this same post, and pointed out that Houston was winning games in an unsustainable way. It was possible for Houston to endure a stretch of bad luck and start losing the same close games they had been winning; if Tennessee or Indianapolis made a run, the Texans could miss the playoffs entirely.
This kind of happened. The Texans lost three one-possession games in a row to Oakland, Green Bay, and San Diego, dropping their record to 6-6. They then won three one-possession games in a row against Indy, who failed to score on a game-winning drive attempt, then against Jacksonville, which blew a 20-8 lead with 6:50 left in the third quarter, and then against Cincinnati, which missed a game-winning field goal at the foot of Randy Bullock. That missed field goal, coupled with a 38-17 Tennessee loss to Jacksonville, locked up the division in Week 16, and broke my big stupid heart by taking away a Week 17 AFC South Championship Game.
In the second half of the 2016 season, Houston had a bad streak of close losses against good teams that turned into a streak of close wins against bad to average teams. This change in fortune got Houston to the playoffs.
To measure performance and fortune, there are four key things to look at: Plexiglass Principle (teams that see their W-L record greatly swing tend to see their record increase/decrease the next), One Possession Record (lots of factors and small things swing close games; teams don’t tend to win/lose close games year to year), Turnover Differential (plays that rely on the opponent to make or capitalize on a mistake), and the difference between Actual and Pythagorean Record (point differential is a better indicator of a team’s performance than W-L record). Now, I’m not here to argue the merits of these numbers. You do you. You believe in what you want to believe in. What I’m here to do is look at these numbers, and see where Houston finished compared to the rest of the league.
(Here’s a link to these numbers for the entire league in 2016.)
Plexiglass Principle
The Plexiglass Principle is wins subtracted by previous year wins. The big negative changes in record happened to teams that were among the league’s best in 2015. Carolina went from 15-1 to 6-9-1 (-7), Arizona from 13-3 to 7-8-1 (-6), and Cincinnati from 12-4 to 6-9-1 (-6). Carolina was the worst of these teams. Arizona and Cincinnati were average performance wise.
The teams that improved were Dallas at +9, Tennessee at +6, and Oakland and New York (G) at +6. Dallas was a classic example of a team that was likely to improve. In 2015, their win total swung from 12 to 4. They were last in turnover differential and won 1.2 games less than expected. In 2016, they went from having the worst trio of quarterbacks in the league to MVP candidate Dak Prescott, and they had that same monstrous offensive line. Their nine win swing isn’t unbelievable. Two of the other teams were more surprising. Despite both their young quarterbacks making the leap, it wouldn’t be insane for either Oakland or Tennessee to cede some their gains next season. Both teams have the cap space to make their teams better though, something both need on the defensive side of the ball.
Houston was the same. Again. They won nine games. They won nine games the year before. They won nine games the year before that. Houston was the same.
One-Possession Record
Teams just win. I mean, teams just win for that season. Usually the next season, their skill at winning close games changes because of a fault in character or not wanting it enough. In 2015, Denver was 9-3, Indy was 7-4, Carolina was 7-1, and Arizona was 5-1 in one-score games. This past season, Denver was 3-5, Indy was 6-5, Carolina was 2-6, and Arizona was 3-5 in one-score games. If you want to know why Denver, Carolina, and Arizona fell off in 2016, this is a great place to start.
Oakland at 10-2, Miami at 8-2, New York (G) at 8-2, Houston at 8-4, Detroit at 8-5, and Dallas at 7-2 had the best record in close games. Miami saw their record blast off from 2-4 to 10-5 thanks to a streak of winning seven (7!) one-possession games in a row. Houston did the thing where they kick field goals, get to twenty points, let their defense dominate games, and hold on for dear life. You can go through each of these wins and pinpoint breaks that went their way to save the win. They didn’t do anything spectacular and break the system. They played good defense, made a couple of plays, and survived.
The other teams were the one exception to the rule. Teams with great quarterbacks are the ones who can sometimes consistently play well in close games. Derek Carr, Eli Manning, Matthew Stafford, and Dak Prescott all range from above average to great. Even then, I wouldn’t expect for any of these teams to play this well in these situations next season.
The unlucky ones in 2016 were mainly bad teams. Chicago (1-6), Cincinnati (1-6), Cleveland (1-5), Philadelphia (1-6), San Francisco (1-5), and Carolina (2-6) lost even when they hung in there. Cleveland was 31st in DVOA, San Francisco was 28th, Chicago was 25th, and Carolina was 23rd. They had bad records because they were bad teams. The close losses only exacerbated this fact.
Cincinnati was different. They were an average team, but saw their win total plummet by six wins, losing games due to the silliest things, like Pittsburgh making six field goals against them, or missing extra points and game-winning field goals (hi, Randy). Philly was 5th in DVOA thanks to their great defense and having the best special teams in the NFL. If Carson Wentz makes a second year jump and their offense improves, they could be a possible playoff team in 2017.
Turnover Differential
The main reason why this number varies years to year is that it isn’t an independent play. You need your opponent to make a mistake, or you need your opponent to capitalize on one of yours. Additionally, recovering fumbles isn’t a skill unless you are J.J. Watt, and dropped interceptions can dramatically affect turnover numbers.
Four of the top five teams in turnover differential made the playoffs last season. Kansas City (+16), Oakland (+16), New England (+12), and Atlanta (+11) made it, while Minnesota (+11) was similar to the 2015 Falcons, in that they started off 5-0 and then plummeted off the cliff. Kansas City and Oakland were first and second in takeaways. This really saved Oakland’s defense; it couldn’t stop teams from moving the ball otherwise. Oh, and Kansas City recovered seven more fumbles than their opponent; Oakland recovered six more. New England and Atlanta had the least number of giveaways with 11, thanks to Tom Brady, Matt Ryan, and fumble recovery help. New England lost only nine of their 27 fumbles; Atlanta lost four of their eight.
Aside from San Diego, the teams on the opposite end were ones with awful quarterback play—Chicago (-21), New York (J) (-20), Jacksonville (-16), Cleveland (-12), Los Angeles (-11), and Houston and San Diego (-7). These teams fell in the top eight of most interceptions thrown, except for Cleveland, which finished 16th with fourteen thrown. The Jets threw 25 interceptions, San Diego threw 21, LA threw 20, Chicago threw 19, and Houston threw 16.
Houston had a negative turnover differential for two reasons. The big one was Brock Osweiler. He threw 16 interceptions, tied for four-most in the NFL with Blake Bortles and Eli Manning. On defense, the Texans didn’t force many turnovers. They finished 26th in takeaways with seventeen, yet were seventh in defensive DVOA. Despite stopping teams and having one of the best secondaries in the league, the Texans didn’t create extra possessions. They had only fourteen interceptions and recovered just six of their opponents’ fourteen fumbles. Houston’s defense should force more turnovers in 2017, no matter what happens this offseason.
Pythagorean Record
The probability of external factors derailing games diminishes if you win by more points. If you have a higher point differential, you are less likely to play close games and have wonky things happen against you. Good teams win by a lot of points, while bad teams just get by.
The biggest overachievers made the postseason—Oakland +3.3, Houston +2.5, Miami +2.4, New York (G) +2.2, Dallas +2.1, and Kansas City +1.9. In the past, teams that out performed their Pythagorean Record by two to three wins saw their record drop off by an average of 2.5 wins the following season. Teams that outperformed their Pythagorean Record between 1.5 and two wins saw their record drop off by an average of 1.8 wins. For Houston and Miami, this is very bad news for 2017. Houston won nine games and Miami won ten. A 2.5 win drop-off next season would more than likely keep those teams out of the postseason. The rest is just what happens when a team wins a lot of games. You are going to overperform some if you win 13, 12, or even 11 games.
Jacksonville was an underachiever yet again in 2016, winning 2.9 less games than they should have, the most in the league. From there, it is a walk down the stairs: San Diego at 2.7 less wins than expected, Cleveland at 2.5, Arizona at 2.4, Cincinnati at 2.3, San Francisco at 2.2, and Philadelphia at two. It takes some bad luck, a lack of talent, and disaster to win one to three games, which is the case for Cleveland and San Francisco. Philadelphia, Arizona, and Cincinnati played the best of the bunch. They played well and won less games than they might otherwise would in a different reality.
2016 was an unusual season. Usually these numbers, with the combination of performance, are simpler and easier to interpret. Most of the time, there are pristine examples of teams that should be better or worse the following year. Last season, the Super Bowl participants were expected to regress. Both did and didn’t even make the playoffs. Dallas and Baltimore were expected to improve, and they did. Baltimore won three more games this season and lost to Pittsburgh, destroying their playoff chances. Dallas won nine more games than they did in 2015and had the best record in the NFC. Entering this offseason, there are multiple teams coagulated together that could see large swings in their record. At this point, Cincinnati, Philadelphia, Arizona, and San Diego are most likely to improve in 2017. The teams expected to see their record decline are Oakland, New York (G), Dallas, Miami, and Houston.
With regard to the Texans, they won the same number games as they had before. They didn’t have any turnover luck, but they outperformed their expected record, which was a direct result of their record in one-possession games.
Houston going 7-9 or 6-10 next season isn’t a for sure thing, of course. The offseason still matters. They are going to play sixteen games again. Teams can get better and improve. That’s what the key is for the Texans. It’s unlikely they can meet or exceed their record by clawing past 20 points and winning close games in 2017. It’s an unsustainable way to win football games. The only thing I feel strongly about entering this offseason is that Houston should have franchised A.J. Bouye, and they are going to miss the postseason next year if they don’t get better quarterback play.
The AFC South is going to be better in 2017. Last year, it was fun, but it wasn’t good. But the Tennessee Titans played really well in 2016. They discovered an identity, and their offense transformed once Marcus Mariota started to throw downfield. Indianapolis has Andrew Luck, who was great considering his situation; they will compete no matter what because of him. The Jaguars, well, just like last season, should be better. Additionally, all three of these teams have cap space. Each of them have the ability to immediately transform their roster and add talent to weak position groups. Houston currently has less than $26 million in cap space, or about $28 million less than Indy, $36 million less than Tennessee, and $49 million less than Jacksonville.