Wednesday, September 05, 2007

I come to praise the Spread, not to bury it | by Jay

After the game on Saturday, I walked out of the stadium with a ton of questions flying around in my head: was Tech that good? Were we that bad? And why hadn't I used more sunscreen? Mostly, though, I had questions about the baffling, alien offense we tried to run with Demetrius Jones. What the hell was that? Instead of the rocket-powered ground attack I was expecting, we fizzled like a dud.

With the benefit of a second look at the game (and some pointed press conference explanations) I think the picture of what happened is a little bit clearer. Some of my questions, answered:

First of all, why Demetrius Jones, instead of Sharpley or Clausen?

According to yesterday's presser, Clausen wasn't 100% healthy until last week. Because of that, he wasn't able to get full reps in practice in the couple of weeks leading up to the game. As Charlie said:

COACH Charlie Weis: At the end of the spring, [Clausen] was the leader of the competition...And then, you know, after the spring, he had that spur taken off and then he rehabbed...he was going to be ready to throw full speed by September 1. But he wasn't ready to do everything on a daily basis, you know, until September 1. So we got kind of got in that gray area where we really couldn't consider him as the starter for the opening game.

Q. Would it be fair to say if he was healthy the whole time, would he have started last week then?

COACH Charlie Weis: That's potentially the case...But in training camp, he didn't do enough. He couldn't do enough to really be in contention for that first team.
So that eliminated Jimmy, and the question became, Jones or Sharpley?

Now, the Tech defense is a whirling dervish of aggression. It's especially hellish for young quaterbacks who lack experience in blitz recognition and might be hesitant in decision-making. As SMQ presciently wrote before the game:
Tech also has a great history in its favor against young quarterbacks early in the season, having claimed the hides of first-year starters in upsets of ranked teams in four consecutive Septembers (Auburn in 2003, Clemson in ‘04, Auburn again in ‘05 and Virginia Tech last year). DC Jon Tenuta is aggressive and unorthodox and has his usual pass-rushing raptors in Phillip Wheeler and Michael Johnson, all the better to force whichever freshly scrubbed scholar Weis has selected into the same eye-opening mistakes of his wilting forebears against Tech. The Jackets got after Brady Quinn something fierce in last year’s opener, and it was Quinn’s instincts and experience that got the limping Irish offense out of trouble on more than one occasion. I’m not giving Sharpley or Jones (or hell: Clausen. Why not?) the benefit of the doubt where icy decision-making is concerned until one of them has demonstrated it in live action. There won’t be much scoring, unless some it comes from the defenses.
If a QB is at all tentative against Georgia Tech, forget it. Last year Charlie had a Heisman contender at QB in whom he had full confidence to read defenses, audible out of plays that might get stuffed, and with whom he could talk to on the sideline about adjustments on the fly. Charlie knew he didn't have that this year. And furthermore, I think Charlie knew a hesitant, slow-footed passer making his first start wouldn't stand much of a chance dropping back and setting up in the pocket, especially with such a green offensive line. (The game, of course, bore this out; Sharpley was sacked 7 times. Consider what we saw of Sharpley...tentative, indecisive, holding onto the ball too long. A feast for the raptors.)

In retrospect, facing an aggressive defense, and with a young line bound to make mistakes, Demetrius really was the only sensible choice to start. He had a weapon that could counter the blitz: his speed. And why not pepper in a few plays designed specifically to take advantage of that speed, right?

What did we install, special, just for Demetrius?


Spread formation, first play of the game
The story of Charlie going to visit Rich Rodriguez at West Virginia really intrigued me. I figured he'd steal some plays from the spread option playbook, and come September we'd see a brand new dimension to the Irish offense. Sure enough, on the very first play of the game, Notre Dame came out in the Spread.

(Quickly, what is the Spread? The version we saw: the quarterback is in shotgun, usually with 1 or 2 running backs next to him in the backfield, and with 3 wide receivers on the field. It's what Purdue plays most of the time, and it's what West Virginia plays, among others. For the Irish, we'd never seen it used under Charlie before the first play of the game on Saturday.)

With the new formation, the Irish also ran 5 brand new plays for the spread:
  • QB Sweep
  • QB Draw
  • Inside Zone Read (either a give to the RB, or a QB keeper)
  • Rollout Pass
  • Screen Pass
That's it. Maybe there were more that were practiced, but that's all we saw in action.

Was this too much to install? Too ambitious for the young Irish? I'm not so sure. We weave in new wrinkles and plays each week during the season, and this set isn't terribly exotic. Keep in mind, we installed a completely new offense before the '05 season when Charlie first came on board -- much more than simply one formation and five plays -- and that worked out pretty well. I have to think this wasn't too much to handle -- in theory. In practice-- well, we'll get to that.

How did we utilize the Spread against Georgia Tech?

It's important to note that in the game we didn't use the Spread exclusively. In fact, Demetrius' play calls were split almost evenly between the Spread and the usual Irish set, with the quarterback under center (let's call it Base Charlie).

DJ ran 27 plays over 6 drives. Two were negated with false starts. There were 13 plays from the spread (highlighted in yellow below). 14 were out of Base Charlie (in blue). Each section break in the table indicates a new drive.

Plays with Demetrius Jones at Quarterback
# Down YTG Package Play Player Yds Mental Errors Notes
1 1 10 Regular QB Sweep 3 +7

2 2 3 Two Tites Toss 26 -3
Unblocked WDE caught from behind and 44 missed block badly
3 3 6 Half-5 Inside Zone Read 5 +11

4 1 10 Out People Rollout Pass 33 inc
DE stayed home, applied pressure, low throw
5 2 10 Detroit Penalty
-5 False start, 77
6 2 15 Out People Screen pass 3 -6 Fumble, 3 Heavy blitz, 3 didn't know what to do - screen pass

7 1 10 New York Inside Zone 34 +1
74 badly missed block on 41 - nice hole otherwise?
8 2 9 Half-5 Inside Zone Read 5 +6
74 beaten badly - should have been called for a hold
9 3 3 Half-5 Inside Zone Read 3 -1
Designed to go to 5, but snap is way too high - busted play

10 1 10 Half-26 QB Sweep 3 +5

11 2 5 Half-26 QB Draw 3 +3
Weirdly designed play - no one to block LB, 51 no push
12 3 2 Two Tites FB Dive 44 -1
Toss to HB is a read here for QB; didn't look at 26

13 1 10 Two Tites Scramble (?)
3 +10
GT sends 6; 44 sneaks out to flat, 26 protects, 3 tucks and runs
14 1 10 Detroit Inside Zone 34 +6
Play designed to go left, cuts back right
15 2 4 New York Toss 34 +3
nice lead block by 89
16 3 1 Two Tites Iso 26 +2
Big hole, nice lead block by 44, DB filled nicely
17 1 10 Half-5 Inside Zone Read 5 +8

18 2 2 Half-5 Inside Zone Read 3 +6
QB Keeper
19 1 10 Half-26 QB Sweep 3 +8

20 2 2 Half-26 QB Draw 3 -1 Fumble, 3

21 1 10 Regular QB Sweep 3 0
44's cut block doesn't work, LB forces 3 back inside
22 2 10 Regular Swing pass 11 +4
Toss play or swing pass
23 3 6 Detroit-5 Pass 3 -3 Sack Cut blocks by 72 and 77 are ineffective.

24 1 10 Half-26 Jab 26 -1
89 couldn't seal, 77 whiffed, 26 never got to hole
25 2 11 Half-26 Penalty
-5 False start, 72
26 2 16 Half-26 Screen pass 26 inc
Overthrown, heavy pressure - 74 bad cut block, 51 bad block
27 3 16 Half-26 Scramble 3 0 Sack GT rushes 3 DL and sends one OLB blitz - 26 can't pick him up

Running game
What's really surprising -- and this certainly was NOT my impression upon leaving the stadium -- was that running out of the spread was actually pretty effective.

The best running plays were the QB sweep, and handing off to Allen on the Inside Zone Read. The QB sweep was run four times, for 7, 5, 8, and 0 yards. Whenever Armando was in the game, we almost always ran an Inside Zone Read. It was either handed off to Armando or Jones kept it. Here's our production for that play:
  • Allen: 3 carries, 25 yds
  • Jones: 2 carries, 5 yds
Jones had one carry for 6 yards on a Inside Zone keeper, and on the other, it was the third & two play on the second drive where the snap was so high that the play was disrupted. DJ couldn't hand it off to Armando, which forced him to run wide and lose a yard. If there had been a good snap, we would have easily earned the first down (there was a solid block by Wenger, and a great cut block by Young on the LB). That would have been another Armando carry, and another 5+ yards.

Contrast rushing from the Spread with the rushing production from Base Charlie:
  • Spread: 11 carries, 50 yds
  • Base: 7 carries, 7 yds
Passing game
Overall, Jones had seven passes called for him (out of 25 plays) in both Spread and Base. There were only two pass plays called from the spread. One was the play action roll out pass to Hughes (which Demetrius underthrew), and the other was the screen pass where Jones developed "deer in the headlights" syndrome and got himself sacked.

There were also two scrambles on passes but both came out of Base Charlie. First, there was the play action on the first play of the 4th drive where he tucked and ran for 10 yards. The other was in the last drive on 3rd down. The OLB blitzed from the outside, and Thomas engaged him but didn't stuff him. Jones stepped up in the pocket to throw, and was nailed. It's technically a scramble because it didn't go for negative yardage, but it's really a sack. Charlie gave Demetrius opportunities to throw the ball, but terrible pass blocking, bad throws, and panicky decisions to tuck and run conspired to kill the passing game.

That said, after a second look, the spread still seemed like a good plan considering the circumstances coming into the game. It was even producing some positive yardage before we fell too far behind and Charlie pulled the plug.


First drive, 2nd & 15. Slot receiver points out the blitz


Allen (btm) & Hughes (top) flare out.
Corner & LB blitz.



Allen & Hughes looking for the pass; DJ scrambles


Sacked. Hughes with all kinds of daylight,
and blocking downfield. Drat.

So why didn't it work?

Mistakes. Piss-poor blocking. Two drives ended by fumbles, one on a critical drive where we might have taken the lead, 7-6. Another hindered by a terrible snap, and another drive ended on a sack due to a bad block. Poor decision-making by Demetrius. And don't forget the terrific, aggressive defense that confused the hell out of our linemen and consistently beat our blocks.

On every drive-ending play, there were crucial mistakes. The starkest example (and most heart-breaking, because it could have been a big play) was on the very first drive.

It's 2nd & 15 on our own 38. ND comes out in the spread, with Out People personnel (2 RB, 3 WR), and Hughes and Allen in the backfield on either side of Jones.

The corner on the slot receiver is showing blitz, and the slot receiver alerts DJ before the snap.

At the snap, Hughes and Allen both flare out. The corner blitzes, as noted, along with a linebacker.

Hughes and Allen are both looking for the pass, and the receivers are setting up blocks downfield. But DJ never seems to look right or left, and instead steps up and squirms around, trying to avoid the blitz.

That play's a killer, because all DJ had to do was loft it out to Hughes and it would have gone for 10+ yards, and maybe more. It didn't require a perfect pass, just a recognition that Hughes was going to be open (again, with the corner blitzing) and an easy pop-up that Hughes could have run under and hauled in.

Thus was the potential of the spread, and thus was it frittered away.

Disclaimer. It strikes me that this post probably reads like I'm putting most of the blame on the players' failure to execute an otherwise sensible gameplan. In truth, we didn't look prepared for Tech's blitzes, we didn't execute our cut blocks very well, and the quarterback obviously lost his head under pressure. This goes to coaching and preparation as much as execution. Charlie (as usual) owned up to not having the team ready for the speed of the game or the defensive schemes presented by the Yellow Jackets, and he rightfully deserves his own set of nice, big goat horns.

So is this the end of the spread?

Charlie said something interesting on Sunday's presser about how the spread gameplan "was intended for this opponent," meaning, presumably, the extent of its usefulness was for attacking Georgia Tech, and Georgia Tech alone. But after seeing the potential with the spread, I really hope Charlie doesn't relegate it to the trash bin. It would be easy to junk the whole thing and go back to Base Charlie with a prototypical dropback passer, but elements of the spread could still be quite productive.

Recall how successful the inside zone read play was from the shotgun with Armando Allen. Now consider this. We trot out against Penn State with Clausen at QB. It's first and ten from the 25, first play of the game. Shotgun formation, Half personnel (3 WR), Armando Allen standing next to Clausen.

You're Lions defensive coordinator Tom Bradley...what do you do? Do you go to nickel because of the three wideouts? Or do you stay in your base defense? Base defense...okay, what do you do with that 3rd WR? Do you bring a safety down, and play man to man on the outside? If you use a LB on the 3rd WR, that's a mismatch the Irish can exploit. And you still have to account for Allen on the handoff; with a safety or linebacker committed to outside coverage, there's a lot of room to run in the middle. While I doubt you'd ever see Clausen on the keeper with the inside zone read, you could mix in play action, or the swing pass to soft coverage, or the Hughes/Allen flare pass. There's a world of possibility here.