Data Says Offensive Breakthrough Possible

Matt Steel takes us through some eye-popping analytics that should help the Steelers complete their offensive revival.

Yes, I'm going back to the well one more time.

It's been my contention for a good number of years now that a quarterback's yards-per-pass attempt is far more valuable than passing yards to his team's success. Despite the NFL receiving the passing-league label (even though teams still run the ball more than 40% of time), I've contended through several years of the naked eye test that the amount of yards accumulated per pass attempt was far more valuable than the yards accumulated through volume. No teams taught me that lesson better than the 2004 and 2005 Pittsburgh Steelers.

Those teams' commitment to the run opened up huge chunks off the play-action passes. Ben Roethlisberger had his two highest yards- per-attempts outputs in his career in those two seasons. In both 2004 and 2005, Ben averaged 8.9 yards per pass attempt. The Steelers averaged only 4.0 yards per carry in those seasons, but their commitment to run limited turnovers and set up more consistently short third downs that allowed the Steelers to rank near the top of the league in third-down conversions, despite having a quarterback who was learning on the fly. Most important, the run commitment set up huge chunk plays out of play-action for their big agile quarterback with great touch on his passes inside or outside the pocket. It was a system tailor-made for Roethlisberger's talents. His ability to make big plays improvising out of the horseshoe protection allowed the Steelers to slay teams early and often. It was a match made in heaven.

But as the football savants known as Tom Brady and Peyton Manning started ringing up wins and points with their pass heavy offenses, most of the league went into panic mode and felt they needed to keep pace. So it's turned into a run-and-shoot league, the only differences being there are more passes near the line of scrimmage and the quarterback is lined up about four yards further from his center.

Yet, while fans, media, former players, and even current coaches are enamored with yardage totals, I've been consistently looking at the most important stat those 2004 and 2005 Steelers taught me: a quarterback's yards per attempt. Those seasons taught me to look at the game in a deeper way. And like I've learned in life, what is immediately in front of your face usually isn't the best answer. The best solution is the one behind it or what you can't immediately see. See the forest before the trees.

Rather than just rest on my yearly eye test, I decided to pull the data from the 2014 season. This yards-per-attempt data also includes sacks and sack yardage lost in the attempts and yardage.

Over 256 regular season games in the 2014 season:

* Teams that won the YPA went 204-52, winning percentage 79.7.

* Teams that won the YPA and didn't lose the turnover battle by two or more went 204-29, a winning percentage of 87.6.

* Teams that won the YPA and either tied or won the turnover battle went 204-20, a winning percentage of 91.1.

* Teams that won both the YPA and the turnover battle went 204-7, a winning percentage of 96.7.

There was a study done by ProFootballFocus.com after the 2012 season that charted all of the play-action passing across the NFL. The entire league averaged an extra 1.8 yards per attempts off the play-action pass. So I factored that in with teams which won both the YPA and turnovers battle:

* Teams that won both the YPA and turnover battle and beat its opponent in YPA by 1.8 had a record of 204-4, a winning percentage of 98.1. The Steelers were on the right side of being one of the fortunate four, but they needed a pick-6 by William Gay and Le'Veon Bell to rush for over 200 yards to overcome that statistical anomaly. In fact, most of the games in which the team lost or tied the turnover battle and YPA had to do something drastically to overcome those stats. Here are some examples:

* The Eagles needed a 100-yard kickoff return to help them beat the Redskins 37-34.

* The Eagles scored a defensive touchdown in beating the Rams 34-28.

* The Jaguars outrushed their opponent 177-95 to win 21-13

The Giants outrushed the Eagles 164-76 to win 34-26.

And these are games in which both teams were even in the turnover battle.

The Steelers lost a game in this category, though we can all agree that Roethlisberger's fumble inside his own 10-yard line had more impact on the game than Cortez Allen's interception on a deep ball in the game against the Buccaneers.

There are similar anomalies in which the team that lost won both the YPA and turnover battle. For example, the Raiders beat the Bills 26-24 by outrushing them 140-13.

To me the most interesting of these games was the Giants beating the Falcons 30-20. Most would wonder how the Falcons could lose by winning the turnover battle, YPA, and having Matt Ryan outgain Eli Manning, 316 to 200. But the Giants ran the ball 32 times and passed it 30. Ryan threw the ball 45 times. The Giants converted 9 of 11 third downs compared to the Falcons 2 of 13. This game is the carbon copy on the difference between true yardage and fraudulent yardage.

The data tells me that two of the most important stats are YPA and turnovers. If a team wins those two stats and runs the ball well, they're nearly unbeatable. For all the talk about how the league has changed, it's still about turnovers, big plays, and running the football.

So how does this relate to the Steelers? Here are Roethlisberger's passing numbers over the course of the season:

* Shotgun on 1st and 2nd down: 198-316, 10 TDs, 7 INTs, 6.6 YPA

* Shotgun on 3rd down: 92-154, 6 TDs, 3 INTs, 7.4 YPA.

* Shotgun on all downs: 290-470, 16 TDs, 10 INTs, 6.86 YPA, 3 lost fumbles, 35 sacks.

* Shotgun with play-action: 22-38, 259 yards, 1 TD, 0 INT, 6.81 YPA.

* Pistol: 4 for 4, 30 yards, 7.5 YPA.

* Pistol with play-action: 11-19, 147 yards, 2 TDs 0 INTs, 7.73 YPA.

* Under center: 31-49, 301 yards 4 TDs, 0 INTs, 6.14 YPA, one fumbled snap. (It should be noted that about half of those passes were plays in which Roethlisberger checked out of a run play and threw quickly. The majority of those plays gained 2 to 6 yards. Very rarely did they look to throw the ball down the field under center on a straight drop.)

* Under center with play-action: 56 for 80, 823 yards, 10.29 YPA, 6 TDs, 1 INT, 1 sack.

As you can see in these stats, 35 of Roethlisberger's 38 sacks were taken in the shotgun, and 13 of his 15 turnovers were made out of the shotgun. It doesn't mean I think all sacks and turnovers would be eliminated if Ben were under center more often, but I do think with 6 or 7 protections in under-center passes would allow the line to better form the horseshoe that provides the quarterback who likes to hold the ball more consistent escape lanes rather than the 5-man protections that collapse around him, and so I think his turnovers would be cut in half and sacks would probably total in the low 20s. I think moving the delivery spot with 3 and 5-step drops and rollouts allows for more success in preventing turnovers and sacks than lining him up in the same target area approximately 80 percent of the time.

Eighty play-action passes from under center is not enough for a quarterback who's been so good at it throughout his career.

A few years ago, I charted Roethlisberger's play-action passes and in that season he had one interception, no sacks and averaged over 11 yards per attempt. Maybe the Steelers' new data team can pull up his under-center play-action stats over the course of his career. He's been the best in the league in those situations. It's where he shines brightest in making his improvisational plays with his size, quickness, and instincts.

It's really been criminal that the Steelers have ranked in the bottom third of the league in play-action attempts the last handful of seasons. Running play-action 4 or 5 times a game on average isn't nearly enough for Big Ben. The average should be between 8 to 10. Heck, I remember in the 2005 divisional playoff in Indianapolis, the Steelers jumped out to a 14-0 lead on the strength of 6 first-quarter play-action passes. But it's hard to run a lot of play-action if they don't commit to running the ball from under center. The commitment and the threat is just as important as the success of the running game. I was backed on that notion by the ProFootballFocus.com play-action study: "It's said that it takes a good running game for the play-action pass to be effective, but this isn't always true."

Now to be fair, I did cheat a little bit on the Steelers' shotgun stats. I took out the last two meaningless drives against the New Orleans Saints and Martavis Bryant's 80-yard TD when those games were decided and the opponent was in its lazy prevent defenses. If I use those stats, Roethlisberger was 304-488 for 3,483 yards, 20 TDs and 10 INTs for a 7.14 YPA. But I did give him the meaningless last-play 26-yard TD pass to Lance Moore in Cleveland. Take those freebees out and that's only 16 TD passes on about 450 passes.

That pass attempt to TD ratio out of the shotgun should give you an idea as to why they continue to struggle in the red zone. Going to the shotgun on a narrow field tightens all the windows. Windows only get tighter when the QB holds the ball past the initial timing of the play. The Steelers (other than on the 1-yard line) used very little play-action in the red zone. They didn't commit to running the football to allow them to attack the seams with the play pass. A more consistent and committed running attack from under center and under-center play-action will solve many more red-zone woes than would a tall receiver.

Finally, I'm not suggesting they completely abandon the shotgun on first and second down. But it should be reduced to between 5-10 combined running and passing plays through the course of the game. I like that it gives the opponent something extra for which to prepare. But drastically limiting it would also allows the Steelers to save their best shotgun plays for third-down conversions and crucial two-minute drills at the ends of halves and games. They have so many well-designed plays from the bubble screen to Bryant to screens to Bell to the quick play-action pass to Heath Miller that seem to work consistently. Just use them when you truly need them.

And use more vertical routes out of the shotgun. A vertical receiver in single coverage is always open unless he's double-teamed by the sideline. I love the quick fade passes down the field the Seattle Seahawks used on their last drive of the Super Bowl. The Dallas Cowboys would squeeze that play in a few times a game while they committed to a season of run to pass balance. Jason Witten made a killing in one-on-one as coverage down the middle on the all-vertical quick pass play. It's a play that allows the QB to get rid of the ball quickly while still looking to get those big, game-winning chunks. It would also set up all of those well-designed short passes the Steelers have out of the shotgun to pop more consistently for big gains.

If they build the offense around the best running back in the NFL next season, they will indirectly build the offense around the strengths of their franchise QB. That means they should complete the rebuilding process by putting together a line that can execute the stretch zone for the patient Bell. Combine that with a commitment from under center with a near 50-50 run-pass ratio and a pass ratio (when factoring third downs and two-minute drills) that is also balanced between plays from shotgun and under center. Otherwise, they won't ever come close to reaching their potential.

For all the talk over the years about how the game has changed, it still comes down to turnovers and big plays. And those plays are best set up by consistently running the football.

The more things change, the more they stay the same.


Steel City Insider Top Stories