How and Why USF’s Offensive Game Plan Failed in 2017

This is part two of a two-part series looking at how the Gulf Coast Offense worked so effectively in 2015-2016 and what went wrong for the USF offense in 2017.

As we’ve reached the deepest depths of the college football offseason, USF fans have been ruminating quite a bit about the previous year’s team, as fans are wont to do in the offseason. One of the more pressing questions that’s spurred full-throated opinions from both sides of the debate is the following: Was the USF offensive game plan truly as misguided as it was accused of being in the 2017 season?

The short answer? By and large, yes. The long answer, of course, is a lot more complicated. Consider the following a comprehensive summation of the shortcomings (and successes, where applicable!) of the 2017 USF offense, as well as how and why they occurred.

The raw stats lie - they don’t account for tempo.

One of the main arguments in favor of the 2017 USF offense is that the raw numbers didn’t dip much from the 2016 season, and any such dip can be accounted for by the simple facts that a) it’s challenging to learn and implement a new scheme, and b) the Bulls had to replace two of the three keystones of Willie Taggart’s Gulf Coast Offense, as well as pretty much an entire coaching staff.

This makes sense! It would have been perfectly reasonable for the 2017 offense to be a bit worse than the 2016 version. Thing is, it wasn’t just a little bit worse.

To wit: when we look at the raw offensive numbers, the drop doesn’t seem that jarring. It’s easy to look at the chart below, call it a day, and chalk up any differences between the two as negligible. In fact, USF’s relative ranking compared to the rest of the nation actually improved in 2017.

But, as they often do, the raw stats lie. We need to consider the context behind these numbers, and account for the biggest inflator of raw yardage - pace of play. In Bill Connelly’s adjusted pace metric, which measures how quickly an offense runs plays relative to the type of plays that they call, 2016’s Gulf Coast Offense ranked a relatively calm 53rd, while 2017’s veer-and-shoot was first in the nation - not a single offense moved faster than Sterlin Gilbert’s last season.

As you might expect, between the frenetic pace and a defense that did a much, much better job forcing opponents’ offenses off the field, the USF offense ran far more plays per game in 2017 (83.4) than in 2016 (71.4). That’s going to lead to some massively inflated yardage numbers.

To get a better sense of how the two offenses fared, let’s look at yards per play, which, without diving into analytics, is really the best metric we have for evaluating offenses.

All of a sudden, things come into focus a little bit. The offense was a full yard per play worse in 2017 than it was in 2016, against a much softer schedule. 35th in the nation still probably qualifies as a “good” offense, but it’s a whole lot less impressive when you inherit the sixth best offense in the country and return the reigning AAC Offensive Player of the Year at quarterback, two running backs that got invites to NFL camps, a fifth-round draft pick at receiver, three of five starters on the offensive line, and a Tyre McCants. Throw in a schedule that featured only three teams that finished with winning records, and 35th in the nation isn’t quite good enough.

Let’s talk more about that returning personnel, though.

The rushing numbers are boosted by Quinton Flowers’ scrambling ability.

Once again, if we look at yards per game, USF’s rushing offense appears to have been one of the nation’s best. The Bulls ranked eighth in the country last year with a remarkable 264 rushing yards per game, a number not far from their school-record 285 yards per game in 2016.

Of course, the numbers get uglier when you look at the per-play averages. USF averaged 5.0 yards per carry in 2017 compared to a remarkable 6.5 the season before. But let’s try to evaluate the 2017 offense on its own merits here—5.0 yards per carry is nothing to sneeze at; it ranked 30th among all teams in the nation. How much of that is attributable to the scheme, and how much is attributable to having the second-best rushing quarterback in college football freelancing from the backfield?

Not counting sacks, Quinton Flowers rushed for 1,181 yards on 179 carries last season, good for 6.6 yards per carry. That’s a tremendous boost for a running game, to be sure, but it’s not quite fair to evaluate the Bulls’ running game by removing Flowers’ numbers from the mix - after all, a good chunk of Flowers’ yardage came on designed runs.

What if we isolated the yardage that Flowers gained on scrambles and took that out of the mix instead? In essence, we’re trying to see how effective the USF offense was when the coaching staff called running plays. If we take a look at the final two games of the season (both of which were against solid opponents, and late enough in the year that we can safely assume that the whole playbook was in use) and omit any yards gained on scrambles, it paints a much bleaker picture of the running game. Note that in the interest of fairness, I also excluded sacks from the below numbers, as those aren’t designed running plays either.

In two of the biggest games on USF’s schedule, the Bulls’ running game outside of Flowers’ improvisation was nonexistent and entirely average, respectively. They weren’t going up against all-world defenses, either - UCF’s defense ranked 74th in the nation, per S&P+, and Texas Tech’s was all the way down at 88th.

Zoom in on the running backs and things look even worse. The backs managed just 48 yards on 19 carries (2.5 YPC) against UCF, and 116 yards on 32 carries (3.6 YPC) against Tech. USF’s rushing attack may have been solid in 2017, but when it mattered most, Flowers was the only one who could consistently move the ball, and even he struggled at times on designed runs. And all of this, of course, occurred with two of the most productive running backs in program history in the backfield.

There are obvious caveats here. Quarterback scrambles are a big part of the college offense, and the Bulls’ running game performed much better against the underbelly of their schedule. Still, whoever wins the QB job ahead of this upcoming season won’t have near the running ability that Flowers had, and Sterlin Gilbert and company have yet to show that they can build a consistently effective running game without the help of their now-departed safety valve. They’ll need to get more out of their designed runs - particularly with the halfbacks - if the 2018 season is to be successful.

Is the #HBDive phenomenon real? Yes, and no.

Ask any USF fan what the main problem with the aforementioned running game was, and nine times out of ten, their answer will involve the infamous halfback dive. Sterlin Gilbert’s adoration for running up the middle became a bit of an in-joke amongst the fanbase during the season, to the point where #HBDive became a microcosm of the angst that many fans felt toward what they perceived to be a repetitive and uncreative offense. There is a USF HB Dive Twitter account.

Was all of this vitriol justified? Well... sort of. In an attempt to be as fair towards USF’s offensive staff as possible, it feels necessary to mention that a large chunk of what many fans classified as halfback dives weren’t dive plays at all.

On a dive play, each member of the offensive line is tasked with blocking the player directly ahead of them. Two of them will attempt to shield their assigned defender away from the gap where the play is supposed to go, and the running back runs through said gap. It’s the most basic of man blocking schemes; you’re trusting your linemen to overpower the defensive linemen one-on-one. It’s pure matchup football.

You can see some pure, unadulterated halfback dive action in the below picture. Every USF lineman is assigned to block a specific UCF defensive lineman (Cameron Ruff, the center and William Atterbury, the right guard are both assigned to the same defensive tackle, at least at the snap). Eric Mayes, the left tackle, is shielding his defender away from the “B” gap between the left tackle and left guard. Jeremi Hall, the left guard, is doing the same. That’s where this play is designed to go. There is no hesitation in the handoff - D’Ernest Johnson takes it directly from Quinton Flowers and immediately charges towards the hole.

USF ran this frequently last year - the most they had since 2014 - but it only made up a portion of the inside running plays that Sterlin Gilbert called. Before we continue further into our deep dive (ha!), it bears repeating that USF’s offensive scheme was not quite as basic as just calling halfback dives 50% of the time.

Here are a couple other common inside running plays that are easy to confuse for a dive when you’re watching from afar.

In the above, USF’s running a trap play in which the left side of their line (Mayes and Hall) is pulling right, while the rest of the line (and tight end Elkanah Dillon) is assigned to shield the defender to their left from the “D” gap to the right of the tight end. In theory, Mayes and Hall should clear out any defenders in the gap, providing Johnson with a lead block to follow, but here they get tripped up in the backfield, the hole never gets sealed off and Johnson is forced to dart forward into the line for a minimal gain. Traps are designed to take advantage of an over-aggressive defense by overloading one side of the line and letting the other side shoot through their gaps for naught, but like dives, they still utilize man blocking schemes for the linemen who aren’t pulling.

The zone read didn’t disappear with Taggart, either. USF ran quite a bit of zone read last year, but with one critical difference.

For context: on a zone read, the linemen will be zone blocking rather than man blocking - in simpler terms, blocking a particular area of space rather than an assigned defender. This is particularly ideal for quicker or smaller teams, as it’s more about deception and making the correct read than it is about man-to-man strength. The below is a good example of the style of zone read that USF liked to run last season; on this play, Quinton Flowers is tasked with “reading” the unblocked UCF defensive end. If the end cheats inside and follows the path of the running back, Flowers will pull the ball and run outside to the space where the end used to be. If the end holds his ground - as he’s doing below, likely in a wise effort from the UCF staff to limit Flowers’ running ability - Flowers gives the ball to Johnson, where he follows the blocks of his offensive line.

USF ran a ton of zone read plays similar to this one last season, and most of them had the halfback as the inside option and Flowers as the outside option. Therefore, a lot of them ended up going inside to the halfback, ultimately appearing much like a halfback dive would.

The issue isn’t that USF was calling the same running play and again and again, per se. They were calling a handful of different running plays, but they almost all had the same destination - up the middle.

USF didn’t need to repeat the Gulf Coast Offense. They just needed to play to their strengths.

So how might one go about running zone reads while also not sending their halfbacks plummeting into the line? I’m happy you asked!

Here’s a play from the Bulls’ 2016 game against Florida State. This is not a drastically different call than the play above - both are zone reads, and both have Flowers reading the left side defensive end. There’s a key difference though - on the above zone read, Johnson is the inside option, while Flowers is the outside option. Below, Flowers is the inside option, while Marlon Mack is the outside option. Here, Flowers sees the end cheating inside towards him, so he hands it off to Mack on the sweep and the Bulls pick up an unorthodox first down on 3rd and 7.

See how on the former play, Johnson receives the handoff from behind Flowers, while on the latter play, Mack receives it while he’s already running horizontally to the left, as one might on a jet sweep? This presents the defense with an unenviable choice - they’ve either got to try to defend Mack in open space, or try to stop Flowers on what essentially amounts to a QB draw. There’s no good option there.

When the running back is the inside option on the zone read, particularly when he’s a shifty, east-west runner like Johnson, the choice isn’t anywhere near as frightening. You’ve either got to defend Johnson or Tice plunging into the offensive line, or you can try to defend the best athlete on USF’s offense in open space. Any defense, as we saw above, is going to try to force the play inside.

USF didn’t have to bring back the Gulf Coast Offense to be successful, but they did need to get their athletes in the best possible positions to succeed. They failed to do that in 2017. Check out the chart below, which measures how successful the Bulls were on different types of running plays in the UCF and Texas Tech games (you can read the details on how success rate is calculated here).

As you can see, USF’s offense was significantly more effective when it got its players into open space (and, in turn, had its lineman blocking zonally instead of man-on-man), but still relied on inside running more than twice as much as they did runs outside of the hashes. In doing do, they failed to maximize their biggest talent advantage and forced players into a scheme for which they were ill-suited.

It would have been foolish to expect a new staff to keep the Gulf Coast Offense in place - every coach and coordinator has a system that he prefers to implement, and there will almost always be some degree of wholesale changes when a new coach takes over (for example, cries for T.J. Weist to succeed Taggart as head coach were almost certainly ill-advised; there’s more to coaching than continuity). But while keeping the same scheme would have been unrealistic - and unnecessary - it’s perfectly reasonable to expect a staff to play to the strengths of their players. In 2017, a USF offensive line that excelled at zone blocking was asked to largely switch to man blocking, and their runners and receivers who excelled in open space were kept largely between the hashes (not to mention the elimination of short, easy throws for Quinton Flowers, but that’s a whole different rabbit hole - Flowers was tremendous enough last year that it didn’t hamper the offense that much). And it seems like the players noticed, and weren’t too thrilled about it.

This is a lesser version of what nearly got Willie Taggart fired three years ago. Then he listened to his players and tailored his scheme to their talents, and it got him the FSU job.

The offense was predictable when it mattered most.

Before we conclude, there’s one last note I’d like to add - something that doesn’t seem to come in in popular discourse regarding the offense as much as #HBDives.

There’s nothing inherently wrong with implementing a new scheme. Even the best, most talent-rich programs in college football are consistently changing and evolving (Nick Saban’s Alabama, for one).

Even if you stick to your guns regarding the veer-and-shoot and want to tailor the team to fit the scheme and not vice versa, you still need to call plays that put your players in the best position to succeed within your game plan. Sterlin Gilbert and company didn’t do that. In their final two games of the season - both of which came down to the wire - the offensive play calling was predictable.

If you insist on running power and are having trouble getting it going, a good way to make things easier on your team is to try to catch the defense off guard and run when they’re not expecting the run. The logic is simple: the fewer defenders in the box, the easier it is for the offensive line to dominate the line of scrimmage. Against UCF and Texas Tech, USF didn’t do that - in fact, they did the total opposite.

For context, the national average for percentage of running plays on passing downs (here defined as 2nd down and 8+ yards or 3rd and 4th down and 5+ yards) is 35%. In two pivotal, close games, in situations when defenses would reasonably expect a passing play, USF nearly always passed.

They were just as predictable on the most crucial of downs - 3rd and short, and 4th and short, when defenses are most likely to be stacking the box against the run. Against UCF, the Bulls were faced with two such situations and ran in between the hashes on both, converting one of two. Against Texas Tech, the Bulls faced seven such situations, and ran in between the hashes five times, converting just one.

The other two? They ran outside twice and converted both times.

This doesn’t mean that the USF offense is doomed.

There are plenty of scenarios in which the USF offense can, for lack of a better phrase, become great again. For one, Gilbert and the staff can learn from the Bulls’ struggles in 2017 (which, it should be noted, were mercifully not as bad as they might have been thanks to a handful of tremendous offensive talents and Charlie Strong’s excellent work with the defense), and tweak their scheme accordingly. Willie Taggart managed to do it just a few years ago, to rousing and almost instantaneous success. There are far worse situations from which to restructure a team than the comfort of a 10-2 debut season.

Perhaps a more likely scenario is that the Bulls start bringing in talent that fits the mold of what Gilbert and Strong want to run on offense. The jury is still out on whether a team can successfully run a power offense on second-tier Florida talent (Taggart and Skip Holtz both struggled with this, to a degree), but Strong has already taken steps in that direction on the recruiting trail; early enrollee and true freshman halfback Brian Norris is an absolute bruiser at 6’2”, 230 pounds, and the Bulls loaded up on offensive line talent in the 2018 class. This year’s team may even be better suited to Gilbert’s system - Elijah Mack may yet be the between-the-tackles workhouse that he craves, and any of the three quarterbacks competing for the starting job may prove to have the arm to open up parts of the route tree that Flowers could not. And, of course, a rare offseason of continuity among the staff and a full year of the system under the offense’s belt can’t hurt. USF fans likely won’t resent Strong and Gilbert for an experimental first season if the pieces end up clicking en route to a conference title in 2018 or 2019.

It should be noted, too, that the failures of USF’s offensive game plan likely did not cost the Bulls a chance at a conference title in 2017. Rue the fates for matching up USF’s most complete team of all time with a buzzsaw of a UCF team, if you like, but anything beyond that is guesswork - the most perfect game plan in the world could not have engineered a more dazzling performance from Quinton Flowers against the Knights.

All that said, it is challenging to give the benefit of the doubt to an offensive staff that so flagrantly misused such a talented team last season. Strong and Gilbert may yet iron out the kinks in the USF offense, but they’ll have to prove themselves capable of such a change first. They’ll also have to prove it without the best player in program history and a handful of other key offensive starters. While there are plenty of reasons for optimism regarding the future of the offense, there are just as many for pessimism.