Duck Tape: Film Analysis of Oregon State 2022

Nov 25, 2022 - 3:01 PM

Oregon State Beavers at Washington" src="https://cdn.vox-cdn.com/thumbor/RPcgik2zgQ-rfWfT8uNGsH5B8mY=/0x0:4350x2447/1920x1080/cdn.vox-cdn.com/uploads/chorus_image/image/71672792/1244528088.0.jpg" />

Special thanks to Travis Johannes of Building the Dam for talking with me about Oregon State’s roster on the Quack 12 podcast. LISTEN HERE

---

Offense

Although some of the personnel and the situational success rates have changed, the structure and philosophy of head coach Smith’s offense at Oregon State remains the same as last year. The Beavers run the ball on about two-thirds of all downs, including 1^st downs, and use the run game to set up their most effective passing. As with last year, this is best understood as two different offenses welded together: under-center tight-in formations for rushing and play-action bootlegs on standard downs and short-yardage, and a shotgun spread offense to throw the ball when they get behind the chains.

Overall, it’s a fairly efficient rushing offense. In their ten FBS games prior to garbage time they succeeded on 156 designed runs vs 136 failures given the down & distance, or about 53.5% efficiency. OSU gets 5.1 adjusted YPC on designed runs, and about 15.5% of them gain 10+ yards. Those are each a bit above average numbers for a Power-5 team, though they’re down compared to last year by almost six percentage points in efficiency, one point in explosiveness, and about 0.7 adjusted YPC.

Due to some injuries (including to wildcat #12 FB Colletto), the Beavs have been down to just one reliable running back in recent games, but I’d argue he’s the best player on the entire team and one of the best backs in the Pac-12: true freshman #6 RB Martinez. He’s the complete package of speed, power, and decisiveness in making cuts for this offense. In fact, the other two backs in the rotation — #5 RB Fenwick and #8 RB Griffin — are far enough behind Martinez that in the first half of the year, when OSU was playing those two much more often, the rushing offense dipped underwater in efficiency. Perversely, the fact that those two have missed a lot of time recently has bumped up OSU’s run numbers simply because Martinez has been carrying the ball almost exclusively. Here’s a representative sample of those successful runs:

(Reminder - after pressing play, you can use the left button to slow any video to 1⁄4 or 1⁄2 speed)

1. :00 – Simple but effective hat-on-hat blocking, this is what the line does best. The offset fullback predicts the run direction on 91% of I-formation plays.
2. :22 – Less than a quarter of OSU’s rushes come with the QB in the shotgun, and they only win three-eighths of those, but when they do it looks like this – two linemen instantly climbing to the 2^nd level. The center here is walking back an undersized ILB eight yards.
3. :34 – The Beavs actually run more frequently when they’re under center with one TE than with two. It tends to pull one of the ILBs out over the slot receiver, then they start running strongside which gets the other ILB to commit that way, only for the RB to cut back the other way after the line has washed the defense down.
4. :48 – I have a weirdly high rate of plays like these on my tally sheet this year, complete assignment busts where multiple linemen are blocking the wrong guys or with the wrong leverage. But Martinez is so good that he just bulldozes the unblocked defenders.

Although I think there’s a strong case to be made for OL coach Michalczik’s line development as the best in the league, I’ve been less impressed with the blocking this year than previous iterations of this offense. Part of the issue is constant rotation at left guard, where three different guys have played. I also think that a couple of personnel moves in the offseason to deal with departures — last year’s left guard moved to center and last year’s right tackle moved to right guard — haven’t worked out as well as hoped, as both of their grades are down on my tally sheet compared to 2021 when I think they were playing more natural positions to them.

It also seems like OSU’s run game has been hurt by not having access to either of their great tight ends from last year: Teagan Quitoriano was drafted and #88 TE Musgrave was hurt week 2 and hasn’t played since (Travis suggested that Musgrave might be playable at this point but could be sitting out the rest of the year to prepare for the draft himself). I noted in last year’s article that there was a very strong correlation between those TEs blocking in the formation and run game success (and vice versa); that trend continues, and even when they have TEs in I don’t think the 2022 replacements have been as effective.

Here’s a representative sample of failed rushing plays:

1. :00 – The center and right guard have played every meaningful snap this year, yet they have the worst run-blocking grades on my tally sheet. Technique isn’t horrible here, they’re just getting beaten physically by the d-line.
2. :06 – Perimeter blocking has also been a problem, the TEs just aren’t as commanding and the WRs are all undersized. It’s a wide-zone run scheme but far outside running is their least effective playcall.
3. :15 – Here’s under-center with 12-pers, almost always a run to that side. But both TEs and the RG on the backside are getting beat before the RB can even make it to the line.
4. :22 – Same story, included for proportional representation. This type of run and these particular blockers getting beat represents close to half of all rush failures.

Starter #10 QB Nolan was hurt midway through week 5’s game against Utah and he hasn’t taken the field since. His replacement, #17 QB Gulbranson, recently got a vote of confidence from Smith. As Travis and I discussed extensively on the podcast, it’s difficult to assess Gulbranson despite having plenty of film on him, because it’s almost impossible to disentangle the possibility that he’s getting better as he gains confidence and experience, from the fact that the defenses he’s been playing against each week keep getting worse. Bizarrely, his first full game against Stanford in week 6 was both his least efficient and most explosive - that one-point win over a bad team requiring a miracle final play featured eight 15+ yard passes, nearly half of all such plays he’s ever produced.

Excluding the snaps Nolan took, overall the Gulbranson-led passing offense has been a bit above water, with 83 successful designed passing plays vs 79 failures, or about 51% efficiency. He’s produced a little under 7.4 adjusted YPA and a little over 14.5% gain 15+ yards. Those are slightly below average yardage and explosiveness numbers for a Power-5 passing offense, in my experience.

From reviewing his tape I think Gulbranson has a strong enough arm to hit any part of the field, he has decent mobility and pocket awareness so he doesn’t take too many sacks or negative plays, and he’s a good fit for Smith’s offense in terms of management and hitting the open receivers that the system produces. On the other hand, I don’t think he has great deep-ball accuracy, I very rarely see him try to make contested throws (much less win them), and I don’t think he’s an elite athlete.

Here’s a representative sample of successful passing plays:

1. :00 – This is a typical big play for Gulbranson – Smith has schemed open a guy on the outside (or maybe the CB is just out to lunch on his assignment) and the QB has the strength and recognition to get the ball there, but not quite the accuracy to hit him in stride and let him run for more.
2. :17 – I have no idea what the LT is thinking here but simply not blocking the edge is weirdly common. Gulbranson is pretty tough to rattle though and he still gets off the slant pass.
3. :27 – Smith also knows about the soft spot in Cal’s zone.
4. :43 – It’s only about a quarter of their passes, but play-action rollouts from under-center are by far the most effective in terms of yards per play of the entire playbook. The DB is totally absorbed in the run fake and the TE is wide open. Gulbranson doesn’t lose any accuracy throwing on the hoof.

Aside from QB issues, I think there are a few other factors holding back OSU’s passing game. First, the lineman who has the lowest protection grade on my tally sheet is the left tackle, which is pretty dangerous for a right-handed QB, and also pretty strange since he was very good in the same spot last year. Second, the tight end and wide receiver rooms don’t have as many good options, and in particular they lack size and length in the wideouts (an issue discussed at length on the podcast and in this summer’s preview) with the primary target #0 WR Harrison topping out the room at only 6’1” and the each of the other three, #1 WR Lindsey, #2 WR Gould, and #7 WR Bolden, coming in about 5’8” (though they’re each very speedy and have great hands).

There’s also an unfortunate inverse correlation between Harrison’s targets and the pass effectiveness rate – he gets twice as many as anyone else, but plays on which he’s targeted have the lowest success rate, only about 44%. That may indicate counterproductive lock-on to a favorite target by the QB, but it’s also the case that Harrison is the only (relatively) tall receiver and so gets the tougher attempts.

Here’s a representative sample of failed passing plays:

1. :00 – The Z-receiver is wide open for a touchdown here but Gulbranson doesn’t see it because the blitz has affected his progression, and his throw to the X is off-target.
2. :09 – This is a pretty common reason that 3^rd & long passing fails for the Beavs – throwing well short of the sticks. The defense has a numbers advantage on the outside receiver but Gulbranson isn’t moving past his first read.
3. :22 – This one’s got a little bit of everything – poor line pickup of a simulated pressure, the throw is too early and too high, and the tight end isn’t athletic enough to go up and get it.
4. :29 – Great one-handed catch, but the ball placement has led him out of bounds. Receiver size and the QB’s deep accuracy issues are the major limitations on contested catches that might otherwise produce big yards.

The final factor contributing to somewhat middling offensive performance by OSU’s standards is just how predictable this offense has become. Each of the trends I noted last year about the tendencies and ineffectiveness of switching between the shotgun and under-center parts of the offense have become exacerbated. While OSU is extremely eager and effective at running the ball in short-yardage situations, they’re significantly underwater (46% in success rate) when running the ball on 1^st down, which they do 66% of the time. That leaves them behind the chains a lot (besides 1^st & 10, their most frequently encountered situation is 2^nd & long), and their success rate plummets in pretty much any other situation than short yardage or 2^nd & medium.

The Beavs throw 68% of the time when taking snaps in the shotgun, and at a lower success rate than when they’re under center, which isn’t great considering it’s their passing configuration mostly reserved for 2^nd and 3^rd & longs and some 3^rd & mediums.

They run the ball about 77% of the time when they’re under center, and at a very high success rate of 67%. They’re also pretty good passing when under center, but only on standard downs; Travis and I agreed that this is likely because of the surprise factor. If they line up under center on long yardage situations, the defense isn’t fooled - they know it’ll be a bootleg pass and defend it well, with OSU succeeding on under 40% of such plays.

---

Defense

There are three factors contributing to OSU’s overall improved defensive performance compared to last year, in my opinion. The first is simply that this is a very experienced group of starters who’ve almost all been playing together for the last three seasons at least, and I think they’ve just incrementally improved along the way into high quality defenders despite modest talent ratings on paper.

Second, starting in last year’s bowl game, new DC Bray made an interesting change to the defensive structure he inherited from his old boss Tim Tibesar. The previous front was a 3-4 double eagle, with a nose tackle, two more down-linemen, and two OLBs, but they were rarely able to actually field a nose and so mostly played a 2-4-5 they didn’t really have right the bodies for. They still don’t really have a nose, but Bray has begun playing #90 DE Lolohea as a stand-up end – he’s the key to understanding the entire front, as he’s big enough to play like a lineman and so this is maybe better understood as a 3-3-5, but he’s capable of dropping out like an OLB. That also allows them to be choosier about the OLB playing opposite him, since they only have to put one on the field, and there are some very strong differences in grades on my tally sheet within that unit.

The final issue is that opposing offenses have had a very foolish approach to their run-pass balance. The Beavs are nearly 12 percentage points more effective against the pass than against the run, which is a huge disparity, and yet in 10 FBS games prior to garbage time they’ve defended almost twice as many designed passing plays as rushing ones. On the podcast, I lamented OSU not playing UCLA since that offense would certainly have made a more sensible adaptation, and Travis noted that the one run-first team they played, Utah, blew them out.

There are several pieces in the front beyond Lolohea whom I like a lot and have been writing about for years: #52 DL Rawls, #96 DL Sandberg, #56 OLB Sharp, and #1 ILB Speights, plus #99 DT Hodgins who’s mostly reserved for heavy sets. They’re joined by relative newcomers #10 OLB Chatfield, #8 ILB Fisher-Morris, and #55 ILB Mascarenas-Arnold. That’s a pretty good group of run-stoppers when they can play the starters. Some examples:

1. :00 – That’s Rawls and Sandberg each defeating a guard for a TFL.
2. :06 – Lolohea as the stand-up end has the power to cave in the right side of the line, and both starting DLs get in the backfield.
3. :12 – The starters are all getting in the backfield, and there’s certainly plenty of enthusiasm on this play. Not great tackling but when you have this many guys hit the back he’ll eventually go down.
4. :22 – The LG is flagged for holding Rawls but he’s still winning the play by stringing this out and letting Fisher-Morris make the tackle.

However, overall this rush defense is underwater, with 76 successfully defended designed runs vs 97 failures, or only a 44% defensive success rate. They’re allowing 4.9 adjusted YPC and 14.5% of opponents’ designed runs gain 10+ yards, which are mediocre numbers and worse than last year. They’re simply terrible against short-yardage rushing, stopping only about 10% of opponent runs from converting on 2^nd or 3^rd & short (and preventing only 40% of opponent runs from succeeding on 2^nd and medium). That’s what happens when you don’t have any truly big gap-pluggers in the middle.

The two major issues I see are first that they’re usually playing a fairly light box of just six guys none of whom are real big hitters, with safety help being relatively late to come down (that’s structural, they’re more worried about the pass) and not particularly effective at that, and second that the backups in the front are a big step down in effectiveness. Ultimately despite probably a smarter defensive scheme and some quality experienced pieces in the front, OSU’s talent limitations are holding them back here. Some examples:

1. :00 – The Beavs had a serious problem stopping read-option keeps in this game, they gave up an average of 16 yards per play to it. The OLB generally is going to crash on the back, the ILB routinely gets fooled by the run fake, and the DBs meant to get the QB on the outside kept getting run over comically.
2. :19 – Here are the backup linemen, and they’re badly overrunning the play.
3. :33 – This is UW’s jumbo set with a sixth OL in as an eligible TE. The structure calls for the ILBs to get the outside gaps so the backup DL has to take on a combo of the LT and LG, but he’s not occupying it at all and the safety can’t come down to help in time.
4. :45 – OSU typically plays six in the box against 11-pers (meaning the QB doesn’t necessarily have to read an extra defender to eliminate him), but sometimes introduces a surprise DB run blitzer from coverage as opposed to being in the box initially. It opens up the possibility of RPO throws, though most offenses didn’t take advantage of that, instead just getting hat-on-hat blocking and chunk runs.

OSU’s pass defense has improved significantly compared to last year. This year they’ve successfully defended 174 designed passing plays vs 141 failures, better than 55%, while giving up just 6.5 adjusted YPA and with only 14% gaining 15+ yards. That’s about four percentage points better in efficiency and explosiveness compared to 2021, and 0.85 fewer adjusted YPA surrendered.

I think that the pass rush has gotten better with the starters’ experience and the addition of Chatfield as an edge rusher, with a higher pressure rate than in 2021 using the same pattern of frequent blitzing. I also think that starting corners #2 CB R. Wright and #5 CB Austin (assuming he’s back from injury after missing last week) have matured into very good outside defenders with NFL potential. Some examples:

1. :00 – The reverse angle camera must be on a construction crane at the half-stadium, but at least it shows the cover-1 blitz structure. They’re betting they’ll get home before someone can shake the two corners and a safety over trips, and they’re generally right.
2. :15 – Another cover-1 blitz, the half-roll gets away from it and the receiver gets behind #23 DB Cooper, but #3 DB Grant (who missed the last game but may be back on Saturday) comes over from his high safety spot for a great break-up.
3. :40 – This look is pretty common on 3^rd downs in passing situations, they’ll crowd the line then back out an ILB and OLB to get the shorter crossers and bring a DB off the edge, here the corner.
4. :48 – The OLB is taking the back so this isn’t a blitz, but Speights is just running right through the center and forcing a backfoot throw. Nice coverage by Wright, working the receiver out of bounds.

There are some vulnerabilities in pass defense. First, they don’t generate nearly as much pressure when not blitzing, although they do have some interesting four-man rushes with OLBs or Lolohea dropping out and bringing one or both ILBs. Second, Wright is more athletically gifted than Austin but he still has the same significant penalty problem that has dogged his career (he gets “grabby” when he gets his footwork wrong, as Travis put it), and the backup CBs are nowhere near as good as either. Finally, I think the safeties and ILBs in coverage in the middle of the field are mediocre at best, and it’s been odd to me that so few teams have really attacked it (the other team OSU missed was Arizona, who I think would have eagerly done so). Some examples:

1. :00 – I think when OSU goes to man coverage that inside receivers have a much better chance of beating the DB and getting open than outside receivers do, as this play illustrates.
2. :16 – Given how often the Beavs blitz – and how dependent on it they are – it’s imperative for the QB to know his hot routes and fill the void. The upside for offenses is that linebackers not named Speights haven’t been fast enough to get home before the mesh comes open and the DB in man tends to get picked by the ump.
3. :30 – This probably should be a DH flag on Wright, but even without it one of Cal’s big and talented WRs gets away from him for the catch and then plows through a bunch of safeties for extra yardage.
4. :50 – We got to see some of the secondary backups with a couple guys out last week, I didn’t think they were great matches for ASU’s longtime starting tight end. I think this has been the curiosity about where Lolohea vs the OLB line up, they drop out the latter but the TE is on the other side so with the ILB rushing he’s got a big void to run into.