Thursday, April 02, 2009

Devolution of an Offense | by Michael

Eons ago -- well, way back at the end of the 2006 season -- when Charlie Weis had an impressive 19-6 record and two BCS berths already under his belt, we examined how his offense had evolved in that short two-year span, and how it might continue to progress given the new talent recruited on offense.

Fast-forward to the present: the offense seems to have lost its way, and has yet to recover its imaginative spirit from those first two years. In fact, it's gradually devolved into a primitive, generic version of that efficient, high-powered unit that shredded defenses in 2005 and 2006.

Four years and change have passed since Weis was first introduced as the head coach. Supposedly nothing has changed -- philosophically speaking. But the declining performance on the field tells a different story.

(This post might require some background reading. If you're hiding in the back of the class because you need a refresher on Charlie's 'multiple' offensive philosophy, check out our remedial course from May 2005, or peruse this excellent 400-level course offered by the must-read Smartfootball.)

An Offense Badly Out of Balance


A quick glance at the distribution of packages utilized tells the story of the offense's devolution. Packages are listed from heaviest (more tight ends) to lightest (more wide receivers).



The 3-WR set Half has been swallowing up the offense's diversity since the end of the 2005 season. Irish fans have noticed, and you've probably seen complaints (or voiced them yourself) that the creativity and imagination of the offense seems to be missing.

There is certainly some truth in that. When an offense uses one personnel grouping 62% of the time,
it's simply less imaginative because so much of it is the same. There really are only a finite number of ways to deploy three wide receivers, a tight end, and a halfback. Furthermore, the offense appears stale because a limited number of plays are run from a core group of formations within that grouping, rather than running a multitude of plays from a few formations.

Let's contrast this situation with 2005. Back then, Half was only used 37% of the time. The Irish threw the ball from Half more often (71% compared to 66% last year), but Weis employed it more as a hurry-up device, both when the Irish were behind or simply to throw the defense a curveball, and in obvious passing situations. It was a situational personnel grouping instead of a staple.

3 receivers, a tight end, and a halfback.
Look familiar?

Last year, however, Half was the base offense. Of course, the reason why it became the base offense was the fact that we had only one viable tight end for much of the year: Mike Ragone went down with a knee injury during fall camp; Will Yeatman was suspended for an alcohol-related offense; Luke Schmidt was sitting out because of multiple concussions; Konrad Reuland had transferred to Stanford; and Paddy Mullen had long ago moved to defensive line to shore up recruiting misses. True freshman Kyle Rudolph was the only tight end in the rotation for much of the year, with fellow freshman Joseph Fauria finally getting in the mix towards the end of the season.

The tight end position is what gives the Weis offense its flexibility
and diversity; with only one, Rudolph, the offense was severely limited. This sameness eventually put the Irish in poor situations on the field in 2008, especially against top-five defenses like USC and Boston College, where the Irish failed to even reach the red zone. Opponents could spend the majority of their time game-planning for Half and Regular formations, and the lack of tight ends prevented Weis from being more creative with the offense.

Back in 2005, the Irish possessed three stellar tight ends in Anthony Fasano, John Carlson, and Marcus Freeman. Not surprisingly, the production out of the multiple tight end packages was solid:
  • Detroit - 5.7 yards per play
  • New York - 4.2 yards per play
  • Two Tites - 5.2 yards per play
Two of those three are levels that haven't been matched since; only the Golden Tate 48-yard touchdown and the 29-yard heave to Rudolph in the Hawaii Bowl pushed the average for Two Tites in 2008 to an incredible 8.6 yards per play. However, it's worth noting that the Irish only called 12 plays all year from Two Tites, compared to 88 in 2005.

Earlier this week assistant coach Bernie Parmalee noted that the lack of tight ends was a serious problem in last year's offense:
"[Last year] we couldn't do what we needed to do, we couldn't use certain packages with multiple tight ends because we only had one. The more guys we have at that position, the better we are. We have three guys right behind Rudolph that really can play some ball and allow us to do more things."
With Rudolph, Ragone, Fauria, transfer Bobby Burger, and incoming freshmen Tyler Eifert and Jake Golic, the tight end depth is beginning to look more promising again. Their development, and in particular, Ragone's health, will play a major role in helping the offense diversify its attack, and will force defenses to spend extra time game-planning for the full variety of looks that Charlie's offense can put forth.

Same Old Song and Dance


Once tight ends became scarce, the diversity of personnel groupings within a game all but vanished, as can be seen in the distribution below.



After Yeatman disappeared, the Irish used Half over half the time in every game until USC.

Tied very tightly to the lack of overall diversity in the offense is the homogeneity of personnel within a single offensive series. Although this tedium was rife through the entirety of many games in 2008, we can see it especially in an examination of several opening drives.

If we look at the first drive of games in 2008, after Yeatman's suspension following the Michigan State loss, the Irish used only one personnel grouping against Purdue, UNC, Wash, Navy, and BC: Half. In the Stanford and Pitt games, the Irish subbed a short yardage grouping on a 4th & 1 and 3rd & 1, respectively. Against Syracuse and USC, the Irish used just two personnel groupings on their two opening drives (both went three-and-out).

However, in the Hawaii Bowl, the Irish used four different personnel groupings on their opening drive. That, of course, is Charlie Weis football.

Again, let's look to 2005 for the platonic ideal. Back then, the Irish mixed and matched liberally: Charlie used five personnel groupings on the opening drive against Tennessee; four each against Pitt, Washington, USC, and Navy; three against Michigan State and Purdue; and, only two against Stanford (although Jeff Samardzija's 80-yard touchdown prevented any further plays on that drive). Weis topped all those efforts in the bowl game against Ohio State, as the first 11 plays featured six different personnel groupings. (Recall that the first drive of the Michigan game and the entire BYU game plan revolved around hurry-up offenses that didn't change personnel.)

The Irish badly need to restore such diversity in each offensive series. Defenses must be pressured to adjust constantly throughout a series to the offense's personnel groupings. Defenses should be shuffling personnel on and off the field, and they need to be forced to consider and think, rather than just react (Who's on the field? What are their tendencies? What did I see in film study?). Last year, Weis and the offense allowed defenses to find their groove, and never forced them out of their comfort zone. That can't happen in 2009.

Ir-Regular Production

The decline of overall offensive production can also be clearly linked to a decrease in production from the Regular personnel grouping, which was the second most popular grouping (21.6%) in 2008. Regular places two receivers, a tight end, halfback, and fullback onto the field. Despite its prevalent use, the Irish failed to meet two basic offensive benchmarks: they were short of four yards per carry and failed to achieve seven yards per pass attempt.

Production from "Regular" (2/1/1/1)
"REGULAR"
Rushing
Passing
Sacks
Breakdown
Year Plays % Used R Yds Ypr PC PA Yds % Ypa No Yds Pass Run Per Play
2005 180 20% 82 372 4.5 61 89 941 69% 10.6 9 23 54% 46% 7.4
2006 106 12% 51 187 3.7 23 42 214 55% 5.1 13 -40 52% 48% 3.4
2008 189 22% 121 453 3.7 26 57 290 46% 5.1 11 -61 36% 64% 3.6

Looking back, it's hard to believe that the Irish averaged over ten yards per pass attempt in 2005, but it's no typo. Fourteen times that year the offense completed a pass out of Regular for 20+ yards, and five of those went for 40+ yards. Secondaries were not nearly as kind to the offense in 2008: the Irish completed only four passes all year that went for more than 15 yards, with three of those coming against Hawaii. Likewise, the poor average in 2006 resulted from the same lack of long passes: only one pass play of 20+ yards all year from Regular. It's worth noting, though, that Regular was used less frequently in 2006 once Asaph Schwapp suffered a season-ending knee injury.

A second issue with Regular was its predictability. Even Weis admitted it to an extent:
"Usually when we go into that personnel group [Regular], James and Asaph are in the game, the teams we're going against knows that there's a good chance, higher chance it's going to be run than pass. Of course, you can never just sell out because it could be a play action shot that you're throwing at the same time." (Oct 28, 2008)
For the season, the run-pass ratio out of Regular was a heavily-tilted 64-36, whereas in prior years the Irish showed a slight preference in passing the ball in 2005 (54-46) and 2006 (51-49). Teams loaded up against the run when they saw Schwapp and Aldridge, and the Irish rarely took any play-action shots until the bowl game against Hawaii. In fact, Clausen entered the bowl game having completed just six of 24 play action passes for a modest 53 yards. The 69-yarder to Tate is a great example of the possibilities of passing out of Regular; the Warriors cheated a safety into the box, Clausen faked the hand-off, and the cornerback, with no deep help, was toast. We need more of that in 2009.

The third issue with Regular was the horrifically inept toss play, which constituted 15% of the Regular rushing attack. The toss play sadly averaged 1.6 yards per carry on 18 chances, and on ten of those carries, the play was stopped for no gain or lost yardage. The best two tosses? Aldridge picked up 17 yards against Purdue, and Jonas Gray had a ten-yard run in garbage time against Washington. The other 16 attempts managed a total of one yard.

Without an effective toss play, and for that matter, without a decent outside zone run (5 attempts) from this grouping, the Irish were very limited in how they could attack a defense on the ground. Schwapp's sluggish pro day results provided further evidence of why the Irish had trouble creating gaps on the perimeter of the defense -- their lead blocker couldn't get out there. Whether the potential dual-back Aldridge, Steve Paskorz, or even Burger (his role has yet to be clarified) can do better remains to be seen.

Perhaps the break after the Southern Cal game provided Weis time to work out the kinks in Regular, with play-action passing and big play capability were evident against Hawaii. (The lone toss play run lost seven yards, so there is still work to be done). With that in mind, given the loss of Schwapp, and the move of Aldridge to fullback, which will leave Hughes or Allen as the starter in Regular, there is reason to be hopeful about Regular production heading into 2009. As Charlie noted this spring:
“The good thing is, now James is repping in there and that will eventually allow you to put in formations where you can take somebody like Armando (Allen) and line Armando out of the backfield; now all of a sudden you are running or passing against base defense because you have a base offensive personnel group out there without putting a lesser skilled player in a position to be able to do that. I think there are some serious options that would open up if this continues to move in the direction that it is going.”
Aldridge should have the foot speed to get to the perimeter, which Schwapp didn't; the question is, can he block? Aldridge is also a much more reliable runner between the tackles, and likely more of a threat to catch a pass. Furthermore, as Weis hinted at, Aldridge at fullback would also allow Weis to use Armando Allen as a receiver out of the backfield. With Hughes or Allen as the tailback, and Aldridge at fullback, the Irish should be a more diversified running team from Regular, and defenses will always have to remain vigilant for a screen pass (or even the double screen pass). Weis called a screen pass for Aldridge less than 1% of the time he was in the game as a Regular halfback; for Hughes, it was 5%; and, for Allen, it was 20% of the time.

In sum, Weis cannot afford to allow Regular to be as one-dimensional as it was last year; he should return to making defenses account for every offensive player in the huddle, and force them to defend every inch of the field.

The Decline of Detroit

As mentioned above, the lack of tight ends clearly limited the offense's creativity and production. Nowhere is this more evident than with Detroit, which utilizes two receivers, two tight ends, and a running back. In Detroit, that extra tight end allows an offense more flexibility than having a fullback or a third wide receiver: he can be lined up in the backfield, flexed out wide, on the line of scrimmage, or on the outside shoulder of the other tight end. Additionally, the match-up of a linebacker on a tight end can often favor the tight end, especially if one has recruited tight ends as well as the Irish have in recent years.

Weis mentor Bill Parcells describes the advantages of Detroit best:
"The nickel player nowadays is...playing about half the time, so he's considered a specialist. He's working on being a nickelback. My contention is the match-up that you get with an additional tight end against a normal safety or normal linebacker is really more advantageous than what you get by deploying your third wide receiver in the game and having defenses put their specialty player in."
As we saw with Regular, production from Detroit has gone down since 2005, although for different reasons. While the issue with Regular has been its predictability, lack of big play capability, and inability to force defenses to account for every player, Detroit has suffered from a shortage of healthy tight ends on the roster.

Production from "Detroit" (2/2/1/0)
"DETROIT" Rushing
Passing
Sacks
Breakdown
Year
Plays
% Used
R
Yards
Ypr
PC
PA
Yds
%
Ypa
No
Yds
Pass
Run
Per Play
2005 179 20% 76 353 4.6 53 92 696 58% 7.6 11 28 58% 42% 6.0
2006 142 17% 57 213 3.7 45 78 478 58% 6.1 7 -31 60% 40% 4.6
2008 54 6% 32 74 2.3 16 22 116 73% 5.3 0 0 41% 59% 3.5

(It's worth mentioning the Detroit's passing numbers took a dip once John Carlson went down with injury in 2006, and when he returned less than 100%. Without Carlson, Brady Quinn completed 3-13 passes out of Detroit for 34 yards. As far as the rushing numbers, the longest run from scrimmage in 2006 was 15 yards; by contrast, Darius Walker had five runs longer than that in 2005, including 37- and 38-yard scampers.)

When we talk about "lack of creativity" in the offense, much of this can be traced back to the disappearance of Detroit over the last two years (as well as other multiple-TE packages). Lack of misdirection plays has been noticeable; but, most, if not all, of the offense's bootlegs (a staple misdirection play) are run with at least two tight ends, something we simply didn't have. Another misdirection is the counter run; these averaged 8.7 yards per carry in 2005 and 5.5 yards per carry in 2006, while all but disappearing in 2008. Another missing piece has been the wham-trap play, which Weis had predominantly used with an extra tight end; again, nearly non-existent in 2008.

The importance of Detroit to Weis is clearly evident in how he tried to bring it back late in the season against USC, albeit with Trevor Robinson as the second tight end, and then against Hawaii with Joseph Fauria. In those games, Weis called a play-action bootleg for the first time since the San Diego State game, a wham-trap play, and play-action off the wham. These were relatively small wrinkles in the offense, but important ones nonetheless.

The decline of Detroit explains the last two-thirds of the season, but it raises a question about the first three games, when Yeatman was still a factor: if the 2-TE packages are the missing link, why wasn't the offense more impressive against SDSU, UM, and MSU?


Let's start with the good news. The passing game from Detroit in the first three games was actually working fine. In the season opener against San Diego State, Clausen was 7-8 passing out of Detroit for 66 yards, including a nifty bootleg play action off the stretch play (an option that disappeared from the playcall sheet after Yeatman's suspension). In the next two games against Michigan and Michigan State, Clausen completed 4-8 passes for 25 yards. While that stat line seems a little meager, the circumstances of the incompletions suggest the passing game wasn't quite that futile and, in fact, just missed out on a few big plays:
  • Inc 1: Rudolph can't protect Clausen's backside, and the hurried quarterback can't improvise and connect with Kamara;
  • Inc 2: A defender tips a play action pass at the line of scrimmage, which results in an interception in the endzone;
  • Inc 3: Play action works against the aggressive Spartan defense, but Clausen slightly overthrows an open Mike Floyd in the post; and,
  • Inc 4: Clausen throws a nice ball down the seam to Rudolph, but the defender makes a terrific play to force an incompletion.
The bad news: there is no glossing over the fact that the run game was awful out of Detroit, no matter what the Irish tried. Robert Hughes was the primary back in this personnel grouping, and Weis tried to pound it with him in the first three games. Twelve times the Irish trotted Detroit personnel onto the field for a first & ten, and the Irish ran all twelve times, picking up just 36 yards. Worse yet, half of those yards came on one misdirection play. That's 11 carries, 18 yards, for those of you scoring at home.

Which brings us to the last point...

That Offensive Offensive Line

Schematics are important, but they're worthless without fundamentals. In every personnel grouping except for Half, rushing numbers have dropped since 2005. And that makes sense-- Dan Stevenson and Mark Levoir, both now in the NFL, were replaced by Bob Morton (not in the NFL) and Sam Young (a true freshman); in 2007, we replaced nearly everyone else with brand-new starters. The importance of OL starting experience was covered by Pat back in August 2008, and it's definitely worth another look. Assuming that Paul Duncan reclaims the left tackle spot, the Irish would have 97 returning starts at offensive line heading into the 2009 season.

For all of the items discussed above, the impact of the offensive line cannot be underestimated. Unfortunately, line play is also the hardest to measure: we simply do not have the exact knowledge of assignments against different defensive alignments, nor do we know the line calls. Whereas formations, receiver routes, running plays, and personnel groupings are easier to discern, OL play is a lot more difficult to assess. But Frank Verducci's arrival in South Bend is confirmation that Weis saw the need for change.

Although we were initially lukewarm about Verducci's hire, early returns seem positive. From the practice videos on Irish Eyes, Irish Illustrated, and UND.com, Verducci seems like more of a detail-oriented teacher than Latina. He gives constant and constructive feedback to his players rather than just barking at them, he stresses the importance of mental reps. Most importantly, as Sam Young explains:
"He's a technician. For me, that's a good thing, because I think I've been lacking in my technique, and it's something I've needed to work on. It's a different perspective. He brings a lot of things to the table that I haven't heard before, that the guys haven't heard before."
While the prospect of a three-year starter not having his technique down is alarming, in a way it's also potentially reassuring. For one, it finally sheds light on what might have vexed the line the last two years. And if there are techniques that the Irish line were not taught, and Verducci is providing that instruction, perhaps the gap between potential and performance can finally be narrowed.

And that's also the outlook for the entire offense. Although there are nine returning starters on offense, questions abound:
  • Who will emerge as the second TE? Who can emerge as a third TE?
  • Can the offense regain its big play capability-- both on the ground and in the air-- from 2005?
  • Can Aldridge help Regular become a more productive and less predictable personnel grouping?
  • Can Verducci coach up the offensive line to play more like it did in 2005, and hopefully transcend even that?
Keep an eye on these things during spring football, but realize that answers are temporary for now. It's September that counts.