Fee Fi Foe Film: Purdue Offense

Submitted by Ace on September 21st, 2017 at 2:09 PM

Previously: Purdue Defense

yooooooooooooooooooo

So, uh, Purdue is crazy fun now. Jeff Brohm's offense put a serious scare into Louisville before easily handling Ohio and Mizzou. The Boilermakers haven't just been far more productive than the moribund Darrell Hazell squads, they've also racked up major style points. Brohm's spread features a dizzying array of run-pass options, motion, misdirection, and straight-up gadget plays. They've managed to be good despite fielding two unreliable quarterbacks throwing to wideouts who, with one notable exception, can't seem to hang onto the ball.

If Brohm sticks around, this offense is going to be terrifying in a couple of years. For now, they're merely frightening.

Personnel. Seth's diagram [click to embiggen]:

Running back Tario Fuller has turned Purdue's backfield from a committee into a bell-cow situation with a great start to the season. There's a fair amount of rotation elsewhere. We're very likely to see both David Blough and Elijah Sindelar at quarterback, and #2 tight end Brycen Hopkins is currently third on the team in receptions.

Spread, Pro-Style, or Hybrid? Spread to the max. Purdue went under center for only four snaps this game, with one of those occurring on the goal line.

Basketball on Grass or MANBALL? Zone blocking, though sometimes with a twist; Purdue's RPOs often feature a pulling guard with the rest of the line blocking outside zone. More on that in the play breakdown section.

Hurry it up or grind it out? Purdue rarely huddles but they also take their time at the line of scrimmage. This isn't like a Kevin Wilson IU offense.

[Hit THE JUMP for the rest of the breakdown.]

Quarterback Dilithium Level (Scale: 1 [Navarre] to 10 [Denard]): Of the two quarterbacks, David Blough is easily the most mobile. He can keep for a modest gain on a read option, but he's more likely to throw when he breaks the pocket, especially since most of Purdue's option plays also have a passing element. He did convert a third-and-two which I'm including just to show the brilliantly designed fake jet sneak, which I'd never seen before:

Blough's not a big-play threat on the ground unless the defense really screws up and a lot of his attempts to escape the pocket end in sacks. He gets a four. Elijah Sindelar is a statue; let's give him a two.

Dangerman: You may remember Danny Anthrop, the obnoxiously productive Purdue slot receiver wearing a running back's number a few years ago. Purdue claims their current starting slot receiver is "Jackson Anthrop," who's definitely not just his older brother with a fresh shave. "Jackson" caught two touchdowns in this game, the first a toe-tapper in the corner of the end zone off a Purdue coverage bust, the second a tough snag between two defenders (both of those clips are later in the post). He also had two fourth-down conversions on PU's ultimately failed last-gasp drive, one when Sindelar made him fight for it with a bigger defensive back:

Anthrop isn't big, but he runs sharp routes and displays really strong hands. He's also used on jet sweeps, though he hasn't broken any for more than a modest gain.

Anthrop may not even be the most dangerous pass-catcher on the team. Cole Herdman is a flex tight end who runs routes like an oversized wideout, and he's reliable both as an underneath receiver and big-play threat; he's caught seven of ten targets for 200 yards(!) and a touchdown on the double-flea-flicker against Ohio. He makes tough catches look easy:

Herdman and second TE Brycen Hopkins are both major threats as receivers, with Herdman boasting slightly better hands and athleticism. Neither player possesses much power as a blocker and can be overwhelmed when they have a hand in the dirt, but that doesn't happen often in this offense.

Finally, running back Tario Fuller had a couple impressive runs on limited chances in this game, getting to the hole quickly and finding ways to twist for extra yards after contact. He blew up in the Ohio game the following week and now has an apparent stranglehold on the job.

Zook Factor: There's no Zook in Brohm, who's already had Purdue attempt seven fourth-down conversions, only three of which came in desperation mode against Louisville. They've converted five.

HenneChart: Purdue has rotated Blough and Sindelar in each of the first three games, though Blough has started to seize control of the job after very productive—and, crucially for him, interception-free—games against Ohio and Mizzou. I can understand why Brohm wanted to give Sindelar every chance to win the job, however. Blough's arm isn't nearly as strong and he compounds that issue by forcing throws to the outside, often a beat late. He threw 21 picks a year ago and had two against Louisville, including an awful pick-six:

He'll also force the ball to his first read. His other interception was a throw over the middle into double-coverage off a flea-flicker when he had a much safer throw to a single-covered receiver on the outside, and he nearly threw a couple more pick-sixes with soft tosses on deep outs/comebacks. He's more accurate than Sindelar but he doesn't adjust nearly enough for his lack of arm strength. He either looks really good, mostly on throws that require touch, or really, really bad. The HenneChart reflects this:

Opponent DO CA MA IN BR TA BA PR SCR DSR
Louisville (Blough) 3+ 10+ (5) 1 3x (1) 4xxx 1 -- 2 1 64%

Blough could've had a better stat line, as a couple open deep shots were flat-out dropped by Purdue receivers. It also could've been much worse; I counted four very interceptable passes, three of which were potential pick-sixes.

Sindelar, meanwhile, has a cannon that isn't entirely calibrated. His best throw of the day was also one of his riskiest and required a great catch by Anthrop, but this is a tiny-ass window and he nails it:

Sindelar isn't often that pinpoint. In fact, I'd call him scattershot. He gets the ball out quicker than Blough but misses a lot of the short-to-intermediate throws that Blough consistently hits; he even took a running back to the ground with a wayward throw on a slip screen. 

Opponent DO CA MA IN BR TA BA PR SCR DSR
Louisville (Sindelar) 2+ 11 (2) 1 7x (1) 4xx 3 (1) 3 -- 48%

Brohm went with Sindelar to try and lead a late comeback after Blough threw interceptions on back-to-back possessions. He reponded with drives ending in a near-pick on a third down throw short of the sticks, a pick on a throw into two defenders, and a one-yard flare to the running back on fourth-and-four. When he's making those kind of mistakes, there's not much reason to play him over Blough.

OVERVIEW

Did I mention it's a spread?

Formations Run Pass PA
Gun 12 33 24
I-Form -- -- --
Ace 2 -- 1
Pistol 2 -- 6
Heavy -- -- 1

The run/pass/PA designations don't tell the whole story here, as a ton of those PAs were actually RPOs, but Louisville wanted to force the Boilermakers to go to the air. All three plays from ace sets with either jet sweeps or fake jet sweeps.

Down Run Pass PA
1st 7 12 19
2nd 7 6 12
3rd 2 15 1

Purdue has been more balanced over the last couple games; they were playing from behind for a lot of the second half against Louisville, but they also didn't make much effort to establish the run early. The emergence of Fuller is likely to change that approach.

Before proceeding further, I must recommend you check out Ian Boyd's post on the Purdue offense over at SBNation. He provides the requisite background; Brohm comes from the Bobby Petrino coaching tree, which means damn near every play features some sort of misdirection or option. Brohm has also incorporated some NFL-style route tweaks into his scheme, and those have already produced some explosive plays. Boyd highlighted a modified four verticals play that got Herdman open to set up a first and goal:

The idea with four verts is that you create maximal stress by flooding the deep field with targets. Purdue has a few other wrinkles mixed in here, though, to ensure the secondary isn’t in position to handle it all.

  • On the twin receiver side, the wideouts switch, with the slot (H) running an out-and-up and the outside receiver running a sort of sluggo (slant-and-go) to try and create confusion or even a rub on the free safety and cornerback.
  • To the flex TE side, the TE (Y, Herdman) uses an outside fake to cross the strong-side linebacker’s face and get inside leverage on the safety ($) who’s responsible for matching him deep.
  • The RB runs a quick stab route at the middle linebacker, pulling him out of the passing window to the TE or becoming a checkdown option, if Louisville’s handled the TE’s route fake.
  • So Blough checks the weak side to see how Louisville handles the switch verticals before coming back and throwing inside to the TE.

Louisville managed to cover the switch verticals on this play. That may be because they had already busted to allow a touchdown in the first quarter against a similar concept, this one with three vertical routes (and again the switch on the twin receiver side) and a couple checkdown options:

Purdue will constantly screw with your assignments via pre-snap motion, RPOs, and unusual route combinations. The offense is already springing guys wide open for huge plays on a regular basis.

The problem, for now, is the personnel. I've gone over the unreliable quarterbacks. The wide receivers other than Anthrop are equally erratic. The Ohio game started off with a comical series of dropped passes, and that carried over from the Louisville game, which most notably featured senior WR Greg Phillips dropping a dead-on would-be touchdown:

Phillips looked like just a guy, as did the other starting outside receiver, Anthony Mahoungou, who boasts excellent size but has a tough time creating separation—he also had a bad drop early against Ohio. They don't get much going to these guys in the passing game that isn't entirely created by scheme. It's Anthrop and the TEs who have to come up with the tough plays; it's no surprise the QBs usually lock onto them on third downs.

The offensive line benefits from a lot of quick passes and looks shaky when they have to hold up for long. Redshirt freshman left tackle Grant Hermanns biffed a couple assignments to allow heavy pressure, while right tackle David Steinmetz was beaten clean twice by speed rushes. I do like their center, Kirk Barron, who had a couple impressive blocks in space. I'm still not sure their interior line can hold up against Michigan's front, and Don Brown should be able to dial up some easy sacks—Purdue had some trouble with blitz pickups, and Blough tends to hold onto the ball too long if his first read isn't there.

The running game served mostly as a changeup against Louisville to keep them honest. Purdue runs a lot of outside zone with some quick tosses, inside zone, and trap/wham frippery mixed in. I'm not too impressed with their committee of backs beyond Fuller. DJ Knox, the unusual 5'7", 210-pound inside runner, is the next-best option. Brian Lankford-Johnson is a meh change-of-pace guy and Richie Worship is a plodding short-yardage back.

PLAY BREAKDOWN

Here's one more example of Brohm scheming up some really tough stuff to defend. The play-action rollout with a throwback to the running back on a wheel route isn't an uncommon big-play shot in football. Brohm added a couple evil wrinkles, however. On this play, the line will block for an outside zone save for the pulling left guard, who leads out in front of Blough after the mesh point. Louisville manages to pursue, dissuading Blough from running himself, and also check the wheel route out of the backfield. Play blown up, right?

Not so much. Herdman, who went in motion before the snap, briefly engaged in a block, then slipped into a drag route to the near side of the field. This provided Blough with a wide open checkdown and netted a first down. Seth diagrammed this beauty [click to embiggen]:

It's really dang difficult to come up with a scheme that accounts for all of that. Thankfully, Michigan has Don Brown.

Comments

Maizen

September 21st, 2017 at 2:26 PM ^

Louisville has the #104 defense in the country. They have been promptly torched by Houston (36 points), Kentucky (41 points), North Carolina (35 points), and Clemson (47 points) in addition to Purdue in their last 5 games dating back to 2016. They are a disaster.

Brohm may be a good offensive mind, but I really don't see how this team is any different than the Rich Rod era Michigan teams.

TrueBlue2003

September 21st, 2017 at 4:43 PM ^

That's a three game yards per game average. It's not tempo-adjusted or opponent adjusted. This is a team that has played three P5 teams (all of which have pretty good offenses, most would agree).

For comparison, Florida in these rankings has the 100th ranked defense.  They've played Michigan and an OT game against Tenn.  Does anyone think they're the 100th best defense?

They had the 19th ranked S&P+ and FEI defense last year, and those are tempo and opponent adjusted measures.  I think they did lose a lot and they may have a bad defense, but it's still fairly unclear just how bad it is.

TrueBlue2003

September 21st, 2017 at 4:50 PM ^

and they were good.  Won the big ten for the only time since 1967 (!!!) so you're probably correct that that was their best team in a long time. I don't even think it was a huge upset, really.

But a brutal game to attend.  We blew an 18 point (I think) halftime lead and lost on a FG at the buzzer.  I assume Lloyd and the gang just ran it into the line the the entire second half but I don't remember those details because I had a lot of fun at what they call "breakfast club" in West Lafeyette.

SoDakforUM

September 21st, 2017 at 2:36 PM ^

I know this isn't a direct comparison, but last year -

Alabama 38

WKU 10

 

It looks like superior athletes on the defensive side of the ball can shut this offense down, especially if Purdue has shaky QB play. I don't imagine that Purdue has a significant advantage in personnel compared to the 2016 WKU team.

Jay Blue

September 21st, 2017 at 3:27 PM ^

So what I took away from this is that their coach also moonlights for them at QB!?!

"On this play, the line will block for an outside zone save for the pulling left guard, who leads out in front of Brohm after the mesh point. Louisville manages to pursue, dissuading Brohm from running himself, and also check the wheel route out of the backfield."

 

bronxblue

September 21st, 2017 at 4:10 PM ^

The more I watch Purdue, the more I see a team that can absolutely beat the crap out of mediocre teams but I just don't see their scheme, with the talent they have on hand, holding up against better defenses.  It's a lot of slow-developing plays that, sure, work if your playmakers can beat the coverage and your line holds up.  Neither of those apply against Michigan, and when they don't work it becomes that Louisville Pick-6, where he stares down the receiver, then throws it behind him, and it's game over.  

They'll put up some points.  But this feels more and more like Air Force, where they get Michigan on skates for a couple of plays but then adjustments are made and it's a lot of panic throws and 3-yard runs.

DualThreat

September 21st, 2017 at 4:11 PM ^

So we get Purdue on our schedule this year, which is just now surging.

We avoid Nebraska, which is in complete disarray.

We get Wisconsin, which is primed for another BIG West crown.

We avoid Illinois, which is one of the only BIG teams Rutgers might be able to beat.

 

Sigh.

 

And Ohio State gets to play both our avoids, and avoids both our plays.

Sigh again.

 

... and while I'm at it, why do FOUR (!) teams we play this year get at least 2 weeks to prepare for Michigan?:

- Florida

- Air Force

- Indiana

- Penn State

Triple Sigh.

factorialite

September 21st, 2017 at 4:34 PM ^

the fact that the only real way for Purdue to beat Michigan is to out-scheme them. Talent-wise, there is a significant gap between Purdue's players and Michigan's. If any coach in the Big Ten can cover that gap with exotic plays and misdirections, it's Jeff Freakin' Brohm. I understand the folk legend of Don Brown around these parts, but football only requires one broken coverage on a series to result in a touchdown.

Let me share some probability theory. When two teams match, higher variance ALWAYS favors the worse team. If you score 30 a game every time, and your opponent scores 25 every time, you win every time. However, if you score 20 half the time and 40 half the time, and your opponent scores 15 half the time and 35 half the time, your opponent wins 25% of those games. Brohm's trick plays serve to raise the variability of potential outcomes - I'd wager a large sum that Brohm shows many, many boutique plays in the hopes that Michigan defenders can't immediately come up with an answer.

I wouldn't be surprised if Purdue completes a few trick plays for huge yardage, and I would be surprised if Purdue DOESN'T have one or two trick plays that fall apart and end up with a boneheaded Blough throw into double coverage or a 12-yard sack. That's the path to victory, however narrow the path may be.

TrueBlue2003

September 22nd, 2017 at 1:46 AM ^

but he clearly has obtained incredible folk legend status amongst the Purdue fanbase rather quickly, considering just a 2-1 start without a win over even a decent team, although blowouts against bad teams do mean something.

I do agree with your assessment though. If you're going to have a chance with less talent, might as well up the chaos factor and hope to get lucky.

They certainly have a chance, as indicated by a fairly modest 10 point spread.  For as god-like as Harbaugh is around here, he's just been ok on the road in two seasons (1-3 against teams that finished with a winning record, with the only win @ a mediocre PSU in 2015, and very, very fortunate escapes @Minn and @IU in 2015).