Bill O’Brien’s Newfound Diversity and Deshaun Watson’s Initial Impression

Bill O’Brien has ridden the merry-go-round of misery at the quarterback position for a while now. Since becoming the Houston Texans head coach in 2014, O’Brien has guided Ryan Fitzpatrick, Case Keenum, Ryan Mallett, Brian Hoyer, T.J. Yates, Brandon Weeden, Brock Osweiler and Tom Savage during regular season starts. The names and faces changed, and the contract in Osweiler’s case, but the standard of play was largely the same.

While the quality of O’Brien’s quarterbacks has generally been the talking point used to define his career, how he set those quarterbacks up was just as big of an issue.

The former New England Patriots offensive coordinator was renowned as an offensive guru, largely because he got to be one of the people who stood next to Tom Brady for a while. O’Brien’s offense in Houston was the blandest in the NFL. He disregarded hard play fakes, screens, clearly-defined route combinations or specific play designs to take advantage of his opponent’s weaknesses. You were more likely to see genuine contrition from Bob McNair than a read-option run, a run-pass option or a packaged play.

O’Brien put his quarterbacks in conditions where they had to make difficult downfield reads while dropping back in the pocket and mitigating pressure. The pass rush wasn’t made to hesitate, the coverage could be proactive rather than reactive. The quarterbacks had to throw the ball on time and often into tight windows.

It was the inverse of Kyle Shanahan’s offense that creates easy big plays for its quarterback by relying heavily on hard play fakes and adjusting play calls to attack specific weaknesses of opposing defenses.

The BD era of O’Brien’s career suggested that he not only had no idea how to evaluate quarterbacks (Osweiler was the guy to lead them where they wanted to go) but he also didn’t understand how to ease the pressure on his quarterbacks. He didn’t adjust to his quarterbacks’ skill sets. He didn’t call plays that simplified the quarterback’s reads or put a greater emphasis on the supporting cast rather than individual passer (Hello Andy Reid, Doug Pederson).

One of the most prevalent knocks on Deshaun Watson coming out of college was the simple offense he ran at Clemson. Docking quarterbacks for the offenses they play in is generally misguided. Quarterbacks can adapt to different offenses and learn the footwork/reads that comes with playing under center opposed to playing in shotgun. The scheme complaint was less concerning than Watson’s performance in that scheme.

A huge part of Clemson’s success was using Watson on designed runs. NFL teams can use quarterbacks on designed runs. The Chiefs and Eagles have two of the best offenses in the league with schemes that incorporate the quarterback’s run threat into the design of its plays.

NFL teams typically don’t run their quarterbacks between the tackles by design though. Only Cam Newton has sustained success doing that and Watson isn’t Cam Newton.

Taking that element out of Watson’s play was going to create a big hurdle for him in the NFL. Putting him in the same situation the quarterbacks from the Before Deshaun era would emphasize his flaws the same way it did the merry-go-round of misery all stars. Watson didn’t go in the top 10 for a variety of reasons. The quality of Clemson’s overall roster allowed Watson to be successful while still showing off consistent flaws.

He had three primary issues:

  • While playing behind an excellent pass-blocking line, Watson regularly ran out of clean pockets into pressure or sacks. It wasn’t about seeing open receivers, it was about the willingness to stay in the pocket and look for an open receiver when he had time.
  • When Watson did attempt to pick apart the defense, he too often led underneath defenders to the ball. His 17 interceptions during his final season in college weren’t a fluke. He was repeatedly making bad coverage reads and throwing the ball to defenders in optimal conditions. This wasn’t a case of him being forced to try and carry a supporting cast or him being unlucky with defenders catching a high percentage of his bad passes, quite the opposite on both accounts.
  • With Mike Williams and a group of tall, athletic receivers who could get open downfield, Watson constantly missed with his deep ball. He hit enough big plays to win a National Championship, but he wasn’t efficient. He wasn’t even close to efficient.

O’Brien didn’t start Watson over Tom Savage initially and while Savage offered no actual upside, it was hard to argue Watson was ready based on what he did during the preseason.

Ironically, O’Brien’s tendency to mismanage his quarterbacks pushed him into making the right decision. For the second time in three seasons, the Texans head coach benched his starting quarterback two quarters into the season. Watson came in for the second half against the Jacksonville Jaguars and started a few days later against the Cincinnati Bengals on Thursday Night Football. He severely struggled.

A long touchdown run against the Bengals on Thursday offered a highlight in a victory to mask the overall impotence of the passing game and ineffectiveness of Watson’s performances. As a rookie thrown into arguably the toughest scheme in the NFL, none of that was surprising.

Even at that point O’Brien had shown some slight alteration to his approach. For the full season last year, Brock Osweiler used play action on 11.48 percent of his attempts. That ranked 22nd in the league. Osweiler used a play fake that sent the quarterback out of the pocket by design only 11 times. In Watson’s first two games he used play action on at least 20 percent of his dropbacks and had six total play action passes that sent him outside of the pocket by design. It was a small change that didn’t appear to be indicative of any major alteration. O’Brien was just using his “mobile” quarterback on bootleg designs.

Watson only threw one screen in the first two games also. In 2016, Osweiler used a screen or play action on 16.18 percent of his passes. In Watson’s first two games, he used a screen or play action on 23.4 percent of his dropbacks. A significant jump.

It was the Week 3 game against the New England Patriots where O’Brien went against everything his career had been defined by. Watson opened the game with four “traditional” dropbacks before using play action on eight straight plays. In previous seasons O’Brien’s play fakes would have been subtle movements, the quarterback would hold the ball in front of his running back during a straight dropback without a lineman or tight end pulling to sell the run design. It would essentially be a traditional dropback. That wasn’t the case in this game. Something was different.

42.4 percent of Watson’s dropbacks came with a play fake. 16.67 percent of Watson’s passes were thrown further than 20 yards downfield. Suddenly O’Brien was incorporating misdirection and aggressively searching out big plays with complementary route combinations. Watson missed a number of key throws that ultimately allowed Tom Brady to steal the game away, but the positives from his performance were enough to satiate a thirsty fan base. He had created big plays both inside and outside of structure.

The Tennessee Titans couldn’t handle the misdirection and different points of attack that the Texans offense offered up in Week 4. Watson threw for four touchdowns while the Titans repeatedly blew coverage after coverage.

26.47 percent of Watson’s dropbacks featured a play fake in that Titans game. For context, Matt Ryan was the only quarterback who used play action on more than 20 percent of his attempts last year, he was at 21.99 percent. After the Titans game, Watson used play action 32 percent of the time against the Chiefs, 34 percent of the time against the Browns and 43 percent of the time against the Seahawks.

A newfound affection for play action wasn’t the only thing O’Brien was offering. He also emphasized screen plays in a way that he previously hadn’t. After Week 2, Watson used a screen or play action on a dropback 42.68 percent of the time. That’s a essentially a different sport to what Osweiler was doing with 16.18 percent last season.

For the whole season, Watson used play action 32.35 percent of the time. That put him more than 10 percent above Matt Ryan from last year and almost 20 percent above last year’s 13.05 percent average. Combine that with 17.46 percent of his passes travelling further than 20 yards past the line of scrimmage, Ben Roethlisberger led the league by a distance last year with 16.5 percent, and O’Brien’s offense went from bland, lazy and extremely taxing for the quarterback to a complete outlier.

The Texans played football like no other team in the league when Watson was on the field. It wasn’t just the commitment to play action and the aggressive approach to seeking out big plays, they did those things while showing off diversity and deception. Opposing defenses were constantly put on the backfoot the way teams that play the Eagles and Chiefs are. No longer could opposing pass rushers explode off the edge knowing what spot they needed to get to on the field. No longer could defensive backs solely concentrate on their assignment and who they were coverage. O’Brien switched his offense from reactive to proactive.

While his rookie quarterback has a different skill set to those quarterbacks who preceded him, O’Brien could have easily done these things with his other quarterbacks. The system he ran previously wouldn’t have worked with anyone who wasn’t an elite passer. The Aaron Rodgers, Tom Brady, Philip Rivers and Drew Brees of the world. It was almost as if O’Brien expected those other players to function in that offense but didn’t think that Watson could. Again, falling into the right decision rather than leading himself there.

But what did all that mean for Watson’s individual performance?

His raw numbers look great, especially for a rookie. His charting numbers less so.

This chart shows off Watson’s accuracy for the 2017 season. His inaccuracy was consistent to every level of the field. Breaking up the passes by depth is important to do with Watson because of how many deep passes he threw. He’s so far removed from the league average that his 63.49 overall accuracy percentage that would have ranked last by a distance in 2016 is largely meaningless.

Watson hasn’t solved his deep ball issues that were apparent in college. He did hit a couple of perfect passes, notably to Will Fuller against the Chiefs and Seahawks, but his positive raw numbers are primarily reflective of how often he threw downfield, O’Brien’s newfound ability to scheme receivers open and DeAndre Hopkins’ gravitational pull.

Against the Chiefs and the Seahawks, Watson threw the ball deep 28 percent of the time.

O’Brien’s play fakes evolved as Watson’s season unfolded. They peaked in creativity against the Seahawks. In the above play Watson hits Will Fuller with his deepest pass of the season. Arm strength question marks aren’t answered when a quarterback throws the ball far. Arm strength question marks are about velocity and trajectory more often than the ability to throw the ball far. This play is a great example of that.

Watson doesn’t really have to hit a window on this play. He has to lay the ball out behind the defense and time the throw well so that Fuller doesn’t have to wait on the ball. He does those things well, but it’s a high trajectory pass (a floating pass) rather than a dart.

Limited arm strength from a quarterback can often lead to more overthrows than underthrows. Quarterbacks overcompensate and strain to get the ball downfield when they release the ball. This is something that has plagued Jameis Winston. Watson really steps into the throw as he releases it to get the most out of his arm and he controls the ball perfectly to give Fuller a chance to run beneath it.

Fuller has been a perfect complement to Watson. Fuller’s value is in this type of offense because he not only gets open deep he can run away from defenders the longer the play develops. He can catch up to overthrown passes if they have a high trajectory like the above throw. He’s not on DeSean Jackson’s level because Jackson is a more well-rounded receiver, but specifically on deep throws he helps a quarterback in the same way that Jackson does.

The only knock on Fuller coming out of college was drops and he didn’t have drop issues while Watson was on the field.

You’ll notice in the first play that Watson used a play fake with a seven-man protection and had a crossing route cutting beneath Fuller’s deep route. It’s a simple read for the quarterback. This play from later in the game is one step further down the path of evolution. This time eight players are used to sell the play fake, that draws the linebackers forward and leaves Hopkins wide open on the crossing route.

Watson again makes the right read from a clean pocket.

On these play designs, the quarterback is being given clean pockets with simplified reads. The quality of his receivers mean that he’s regularly got open targets to get the ball out of his hands in a timely manner. There is no full-field read, there is minimal movement required to mitigate pressure in the pocket. He doesn’t have multiple pressure points to work away from in collapsing pockets and he doesn’t have to absorb a hit as he releases the ball.

They are chunk plays where the quarterback doesn’t have to do anything particularly difficult. These are actions you would expect an NFL starter to execute consistently. That’s not to detract from Watson but it’s important to contextualize the production that comes in this offense to create a more accurate assessment of the individual.

For Fuller’s second touchdown, O’Brien showed the Seahawks something his offense has never used before. Pre-snap motion isn’t something O’Brien has emphasized. On this play Bruce Ellington not only motions into the backfield at the snap he reverses his direction while behind the quarterback to act as a second play fake. It’s not a standard triple option look, these are two very clear, aggressive play fakes.

Those play fakes pull the pass rush all over the place. The coverage isn’t impacted but Richard Sherman sees DeAndre Hopkins coming on the crossing route so he anticipates that Fuller is going to continue up the seam. The same route combination the Texans have used repeatedly.

Sherman jumps inside and Fuller uses that to break back towards the pylon behind him. Fuller actually slows down in his route, not expecting to get the ball, but re-accelerates to catch a slightly high pass. The ball is slightly high but not enough to be considered inaccurate.

Again, Watson throws the ball with timing and makes a good decision, but it’s a clean pocket to a wide open receiver after a hard play fake. The degree of difficulty of this play isn’t high unless your standard for good quarterback play is Tom Savage or Brock Osweiler.

Watson had very clear positives and negatives against the Seahawks. He was 58 percent accurate with two interceptable passes and one non-quarterback interception (the hail mary late). His first interception mirrored an issue he had in college where he led Earl Thomas to the ball over the middle of the field while staring down DeAndre Hopkins. His second interception came when he panicked after a bootleg play action that the Seahawks anticipated. He forced the ball into a crossing route where Richard Sherman was waiting to pick it off.

Hopkins’ return to relevance after the Osweiler era was the best part of Watson’s rookie year. Hopkins has a gravitational pull. You only need to put the ball near him for him to have a chance at creating a completion. Somehow, Osweiler turned him into a black hole. Only Osweiler has managed to shut down Hopkins so it shouldn’t be any real surprise that Watson could make it work after doing something similar with Mike Williams at the college level.

Agains the Seahawks Hopkins created one reception when he bullied Richard Sherman on a backshoulder throw that Watson threw too far infield. That was an impressive play but not in the most impressive ones he’s made this season.

Each of the best receivers in the league offer their respective quarterbacks an abnormally large margin for error. None of those receivers are as quarterback-proof as Hopkins. Overthrows to normal receivers are just adjustments for Watkins. He high points passes and pulls balls up off the ground, he can make toe-tapping receptions while falling out of bounds or combine different types of spectacular to make plays that superlatives aren’t worthy of.

That’s without even considering the 75-yard screen touchdown he ran in against the Seahawks.

Hopkins is a complete receiver whose career has been hindered by subpar quarterback play. The greatest argument for Watson’s quality is that he was getting more from Hopkins than those who had come before him. The counter argument would be that the scheme change helped, Watkins himself has been more consistent than in seasons past and those quarterbacks shouldn’t be the measurement for quality.

Watson has seen nine of his inaccurate passes caught for 190 yards and two touchdowns this season. 190 yards would have ranked 15th in the league for the full season last year. That number doesn’t include the 42-yard hail mary catch by Stephen Anderson against the Chiefs.

That Chiefs game proved to be an ideal case study in the difference between performance and production for a quarterback.

Whenever you throw for five touchdowns the result of the game or even your actual performance won’t be able to hold back the swell of excitement and praise. The hype machine that surrounds NFL quarterbacks, especially young quarterbacks and especially during the season, is a predictable one. It’s also an irrational one.

Watson had one truly great play in that game. He evaded a rusher in the pocket and hit Will Fuller with a perfect deep ball for a long touchdown. His first touchdown came towards the end of the second quarter when Hopkins was open on a slant behind a play fake in the redzone, a play the Texans used regularly to great success with Watson. The second was a redzone pass to Fuller on the backside of a play that was designed to go left when Watson read the defense and escaped the pocket.

Three touchdowns would have been easily brushed aside because of Watson’s overall performance. But the narrative swung on two fourth quarter touchdowns. One can be seen above as Hopkins leaped above two defenders to high point a heave into the endzone. The second went to Hopkins after Stephen Anderson caught a hail mary to set the offense up at the goal line.

If you are able to look past the five touchdown plays, you’ll see that Watson was accurate on 14 of 28 attempts, a horrible 50 accuracy percentage, with three interceptable passes.

He threw those three interceptable passes in the first half and was only accurate on seven of 15 attempts threw the first three quarters. The Chiefs defense played their assignments and weren’t thrown off by hard play fakes. Without the schemed open chunk plays, the offense stalled and Watson wasn’t able to move it downfield. That first half touchdown drive? It started with a ball thrown straight to a linebacker, that play was followed by six running plays (one 13-yard gaine from Watson). He capped off the drive but tried to end it before relying on his supporting cast to set it up.

Using that game to kill a rookie would be as irrational as using it to prop him up as a star. The five touchdowns didn’t tell us as much about his quality as the overall performance did. The overall performance wasn’t good, which isn’t ideal but not the end of the world against a tough defense. Watson did have an interceptable pass rate of 5.88 percent so he has ball security issues that will need to be addressed, especially considering the quarterback friendly scheme he executed.

Those same issues existed in college. All of Watson’s issues from college have carried over to the NFL. He is still running into pressure and sacks from clean pockets. He still throws the ball to defenders too much and his deep ball needs to become a lot more consistent. His positives have been prominent also. He has made some very impressive touch passes into tight windows, his elusiveness has created big plays and his threat as a runner has helped to open up the running game for Lamar Miller.

He’s made enough positive plays to be optimistic about his long-term outlook. He’s made enough negative ones to be skeptical too.

Watson faces two challenges moving forward. Firstly he’ll have to return from injury and prove his health and trust in that knee. It’s something all quarterbacks deal with to different degrees. Then he’ll have to prove how effective he can be outside the comfort of a play-action, deep-shot reliant offense. Colin Kaepernick was never given the time to transition from Jim Harbaugh’s play-action reliant offense after his initial success in the NFL. It can be hard to show patience with someone once the hype reaches a crescendo early in their careers.


The 2018 Quarterback Catalogue is available to preorder now

Leave a Reply

Optimization WordPress Plugins & Solutions by W3 EDGE
%d bloggers like this: