Thanks as well to ATQ’s Tristan Holmes for some of the preliminary charting work on OSU.
Nota bene: I recommend reading ATQ’s Oregon State previews from the 2021 season and 2022 season by way of understanding the structure of Coach Smith’s offense and the evolution of DC Bray’s defense out of former DC Tibesar’s. There are now years-long trends regarding critical personnel pieces to those structures which are starkly clear when comparing the video documentation across seasons (I’ve made sure the links to the clip compilations are working). I think the outline of the two phases of Smith’s offense — and why tight ends are necessary to it — in the 2021 article is the most essential.
Oregon State’s offense is predicated on its efficiency run game, which they use to stay ahead of the chains and set up their passing game. The Beavs are rushing at a 56.5% success rate outside of garbage time (150 successes vs 115 failures, given the down & distance), averaging 5.9 adjusted YPC with 19.6% of designed rushes gaining 10+ yards.
Here’s a representative sample of successful rushes:
(Reminder – you can use the button in the lower right corner to control playback speed)
- :00 – The short-yardage wildcat runs that had been done in previous years by Jack Colletto have been replaced by starter #5 QB Uiagalelei doing it himself (including once against Colorado when they swapped him in for #0 QB Chiles during his scheduled once-per-game drive on the 3rd possession); he has a 78% success rate on these which is better than the RBs in the same down & distance.
- :07 – This run is typical of #6 RB Martinez’s style to press in one direction from an under center snap, get the defense slanting that way, then cut the other way outside. The end has a shot on it by beating their pass-catching TE with poor overall blocking grades, but doesn’t keep his hips square and loses the edge.
- :22 – Stanford has backed out their fieldside OLB into coverage but this is a mistake, only two of the players on the field ever receive any targets (#7 WR Bolden and #88 TE Velling) and this formation is virtually always rush to the wide side for backup #1 RB Fenwick, now with no edge containment.
- :30 – OSU uses this technically 11-pers formation to get defenses into their lighter alignment since UW was defeating their run game with a three-down bear front against 12-pers, but it’s a predictable run since only #2 WR Gould gets any catches while the TE and two other WRs are only ever blockers. This run is typical of their bias to running right, somebody getting grabby, and a missed 2nd level block made up for by the RB breaking tackles.
Compared to previous seasons with the same offense, the Beavs are about three percentage points more explosive than they were in both 2021 and 2022, reflecting the maturation of Martinez in his second year in finding ways to break already successful runs bigger by juking defenders and running through contact. But despite that phenomenon, YPC has remained static over the last three years, and in 2023 the efficiency rate has fallen three and a half points from its peak in 2021 – in other words, the rushing offense is a bit more boom and bust than in previous iterations.
That effect shows up on a game-by-game basis as well: in half of the Beavs’ ten FBS games their rushing efficiency is at or below 50% (SDSU, Utah, UCLA, CU, and UW), while in the other half they operated at a championship caliber north of 60% (SJSU, WSU, Cal, Arizona, and Stanford)
The cause of the dip in efficiency, in my opinion, is that blocking is somewhat less consistent than in the past. There are two reasons: first, this year’s offensive line, while quite possibly the best developed in the nation by OL coach Michalczik, simply doesn’t grade out as well on my tally sheet as it has earlier (2021 looks like the peak with a cumulative 12.64% per-play error rate run blocking; they’ve since lost a couple guys off that line and today the number stands at 16.85%). They’ve also faced a few injury issues that seem to have been harder to recover from, as recounted on the podcast, and Jake told us that there’s a strong possibility that an unrated Juco with very little playing time will be making his first start in Autzen tonight at left guard.
The second issue, which I’ve been tracking for several years now, is that the Beavs’ performance falters when the o-line isn’t supplemented by tight end blocking. This year the effect is especially noticeable on the left side of the line, which needs the help more than the right does, and is why it was so curious that during the offseason the staff was so complacent about not adding more help to the TE and many other units.
Here’s a representative sample of failed run plays:
- :00 – Utah’s making a pretty safe bet by blitzing the corner that this won’t be a pass, because they know the odds of it in this formation are so slim. OSU’s line was pretty consistently losing to Utah’s like this for the first three quarters of the game until fatigue set in for the latter due to their inability to rotate.
- :14 – This is a weakside run to the left with no tight end blocking help, at which OSU has only a 36% success rate. The defense easily achieves a leverage advantage across the board, and the blocking WR whiffs on the CB, so it’s a TFL.
- :22 – On the next drive after the above clip, the same DE has corrected his mistake when facing the same play (mirrored to the right this time). He’s defeated the TE’s block but has stayed wise to the cutback and maintains the edge, getting the stuff.
- :34 – UW’s #91 DT Letuligasenoa was on a pitch count last Saturday but had an astronomical 94% block destruction grade on my tally sheet during the roughly third of all snaps he was in for. Here he’s blowing up the replacement rookie LG for what’s arguably the game-winning play, though it’s not like the RG or TE are doing much better.
The inefficient passing game and the absence of a reliable quarterback to extend drives when they got behind the chains was a widely observed problem in OSU’s offense the last several years, and Coach Smith got his highest profile transfer of his career in Uiagalelei to address that issue. His arm strength is obvious on tape and when he connects it’s certainly for longer passes than anybody the Beavs have had since Jake Luton. They’re up to 8.6 adjusted YPA and 20.8% of passes gaining 15+ yards, an improvement of a full yard and about five percentage points compared to 2021 and 2022.
Here’s a representative sample of successful passing plays:
- :00 – OSU’s redzone specialist is #88 TE Velling, who has one of the highest rates of TD receptions among any TEs in the country. His hands are pretty good and Coach Smith has always had excellent redzone TD schemes; Velling’s numbers are are an artifact of not having any tall WRs appropriate to catching fades and not having any other TEs who can catch passes. Cal seems not to know that they only have one guy to defend, and doesn’t defend him.
- :09 – The vast majority of OSU’s big passing plays look just like this – an intermediate throw over the middle that’s an easy read and throw for Uiagalelei for one of the short, quick slot guys (Bolden, here) who takes it big. This one comes from a coverage failure by the linebacker chasing someone out of his zone, worse, it’s one of the decoy WRs.
- :31 – Here’s Gould, on effectively the same thing but lined up wide. The CB gets flipped around while worried about outside leverage, but he shouldn’t be given his size advantage if this goes to the sideline, the inside break is the only real threat.
- :40 – Half of all plays of 15+ yards that OSU got last Saturday were screen passes (all but one of the rest were runs). Defeating them takes a lot more discipline than this – the DL needs to be more aware of when the OL is letting them go on these inside RB screens, the CB shouldn’t overpursue out of his zone, and the inside LB shouldn’t overrun the play since Martinez excels at cutting it back.
However, passing efficiency has actually fallen, to 45% (108 successes vs 132 failures) which is a decline of three to four percentage points from the last couple years. In my opinion, Uiagalelei has many of the same accuracy and inconsistency issues as previous OSU QBs and which OC & QB Lindgren hasn’t fixed since the transfer’s arrival from Clemson.
But as Jake fairly pointed out on the podcast, there’s an apples-to-oranges problem when evaluating Uiagalelei, because he just doesn’t have the same caliber of pass catchers in 2023 as there have been at OSU in the past. Jake and I discussed several examples of reliable go-to outlets like Tre’Shaun Harrison, Isaiah Hodgins, Tyjon Lindsey, Luke Musgrave, Teegan Quitoriano, and Noah Togiai who QBs in 2021 and 2022 had to throw to but aren’t around anymore.
That’s what made it so strange that Coach Smith pursued no transfers in terms of pass-catching TEs or any kind of WRs, and has been left to move his only experienced receivers — who are 5’8” slot guys — to the outside with predictably poor results. Also, the slip in offensive line blocking plays something of a role here as well – the 2021 line graded out at an elite cumulative 8.7% error rate in pass-pro on my tally sheet, whereas the 2023 line comes in about four percentage points worse (although still well above average for the Pac-12) at 12.6%.
At any rate, the core issue with OSU’s offense remains the same: it’s deeply predictable by down & distance and formation (the 79% run rate when under center and 29% run rate when in the shotgun in 2023 are both more extreme than last year), and they have one of the lowest 3rd & long conversion rates in the country which is essentially unchanged over the last three years (29% in 2021, 32% in 2022, and 31% in 2023).
What that adds up to is that, despite the bigger arm of the QB, the same well noted problem still obtains this year: once the Beavs fall behind the chains, they’re unlikely to jump ahead of them again. And since they’re slightly less efficient on other downs as well this year, OSU is if anything even more likely to be knocked off schedule and have to cross their fingers for a big play.
Here’s a representative sample of failed passing plays:
- :00 – Arizona used a lot of its now-famous 7-DB defense on obvious passing downs like this against OSU; they’re still getting pressure while rushing three and can easily cover the diminished receiving options. The receiving target isn’t big enough for this sideline catch and the QB has misplaced the throw, which needs to be further outside.
- :09 – Colorado went the other way, consistently bringing pressure at a pretty high success rate. They evidently don’t notice the obvious creeper here because the LT is uselessly doubling the end with the TE while the blitzing DB gets a free path, and the C just gets run over.
- :28 – Here’s Chiles in on his designated 3rd possession. This sideline route is clearly defeated as the bigger CB has worked Gould out of bounds (note the official’s thrown hat); Chiles should have checked this down to the back given how easily Stanford covered his actual receiving options but he tends to force the ball downfield regardless.
- :36 – OSU has only two viable receivers in, Gould and Velling, and UW knows it so they blitz. Uiagalelei delivers a wild ball which is probably a throwaway on one of only two targets for Velling on the night; his single catch for 1 yard last Saturday was one of six games this year in which he has under 10 yards per reception, underlining his role as a redzone specialist.
The throughline for every team that had a strong performance against OSU — either a win (WSU, Arizona, UW) or a closer than expected loss (CU) — was not being intimidated by the reputation of the Beavs’ offensive line, and instead going right at them to disrupt the offense rather than contain it. Attacking one’s opponent where they’re strongest is an old strategy in the history of games, and it appears to be an effective one against this year’s offensive line which, in my opinion, struggles more with a stiff challenge than previous iterations. Some representative examples from those games:
- :00 – This is typical pathing from the back as the defense slants, but Wazzu’s having none of the cutback as they just beat the backside blocks and blow up the back – this is one of the downsides to using a small WR as a quasi-TE.
- :11 – A blitz in the redzone plus disciplined coverage works well for Arizona – they’ve got the real threat in Velling double covered the whole way, stay on top of the WR screen and RB wheel to the field, and Gould’s drag is too long developing as the additional pressure gets through.
- :19 – Colorado did much better against OSU’s run game than either team’s preceding numbers would have suggested, the Buffs’ secret was simply attacking the OL with speed that they couldn’t handle and getting multiple defenders in the backfield on over half of all plays.
- :29 – This is a pretty basic twist while rushing four and backing out of the blitz, it runs right over the reconfigured line and gets to the QB, who lands on his throwing arm uncomfortably.
Last year’s surge into a top-20 defense in F+ advanced statistics for the Beavs was built on NFL-caliber cornerback play in the pass defense, as well as several other long-tenured super seniors throughout the squad. Those have all departed in 2023, and predictably enough they’ve fallen into the 30s.
On my tally sheet, OSU’s defense has a 53.5% success rate against called passing plays in 2023 (168 vs 144), allowing 7.1 adjusted YPA and 15% of opponents’ passes to gain 15+ yards (their success rate dips slightly underwater if a rather horrific outing by Utah’s third and fourth string QBs are excluded from the sample). That’s a bit of a tumble compared to 2022’s numbers, which were 58% success, 6.4 YPA and 12.6% explosives allowed.
As Jake and I discussed on the podcast, the one unit basically unaffected by all the personnel losses from last year was the edges in what is effectively a 2-4-5 defense (the podcast also discussed the evolution of the structure and why the school officially referring to it as a 3-3-5 is so misleading). They have a decent if unspectacular 17% sack, scramble, or throwaway per dropback rate on my tally sheet. There’s also not a huge split between 3rd & short vs 3rd & long pass defense success rates, indicating that the pass rush is not essential to their effectiveness.
What has been effective have been their trio of starting safeties, including the nickel – all of them veterans of the last couple seasons who had a good distribution of time with the only meaningful loss from 2022, Jaydon Grant who wound up with the Raiders. They had also been getting, until recently, fairly good play from the corners who replaced Alex Austin and Rejzohn Wright when they left for the NFL, although injuries have been severely curtailing that unit lately.
Here’s a representative sample of successful pass defenses over the season:
- :00 – Good recovery by starter (and according to Jake, the only uninjured member of the original secondary) #28 DB Oladapo to cut off the slant, and true freshman #27 CB A. Jordan in one of his few reps helps with the PBU.
- :12 – Man coverage can handle all these hitch routes, and the QB panics against a simple four-man rush.
- :25 – Bray has introduced a new wrinkle this year that Jack and I discussed on the podcast, which is pulling the nickel and putting in a third backer against heavy offensive formations on standard downs (instead of a third down lineman, as his predecessor would). Good coverage of that rollout flood concept here, including by that third non-starting backer #44 ILB M. Jordan.
- :34 – This RPO throw is misplaced, it should be on the receiver’s upfield not downfield shoulder (also the line is getting away with a hold and an IDP because of course they are). Nonetheless, great PBU by true freshman #23 CB McCoy, who’s had to become the new starter the last several weeks due to an injury.
Unfortunately, Jake told us about quite a stunning list of injuries to the back end of OSU’s defense, many of which I was unaware of until we spoke:
- Starter #0 DB Arnold had a boot on his foot at the end of last Saturday’s game
- Backups (or potential starters) #7 DB Julian, #30 DB J. Johnson, and #17 DB S. Thomas have season-ending injuries
- Starter #1 DB Cooper at nickel has been playing on a nagging leg injury for weeks
- Starter #4 CB Robinson didn’t play the second half of last Saturday’s game with an injury
- Initial starter #25 CB Ivy and backup #21 CB N. Thomas have season-ending injuries
- Backups (or potential starters) #20 ILB J. Miller, #14 ILB Shaw, #3 ILB Tongue, and #42 ILB Tufaga have season-ending injuries
Combined with several other guys in those units being healthy but not playable or hanging onto their redshirts according to Jake (a full recounting of the situation is on the podcast), OSU is in a position where they are unable to rotate at any of the back seven in the defense – inside linebackers, safeties, or corners. It would appear, unless OSU gets some very positive news from the trainers, that they’ll be playing two true freshmen at cornerback and a walk-on at safety, with no options for midgame relief.
It’s difficult to say what effects this may have going forward, exactly. Many of these injuries and personnel constraints had already been part of OSU’s season, while others were new last week. It’s made selecting representative clips of the pass defense somewhat challenging, to show the players in previous games who Jake has told me should be available:
- :00 – I think OSU’s DBs are well coached but there’s a talent ceiling, and it’s most evident in deep passing coverage where they can get cooked, even with big cushions. Their solution is to play zone when there’s a lot of grass to cover and it just takes a moment’s hesitation to put them in an unrecoverable situation against speedier receivers.
- :16 – OSU uses this inside blitz with the OLBs bailing into coverage against about 17% of dropbacks, it leaves them no effective underneath coverage or over-the-top help against in-breaking routes and was a poor choice against Arizona, especially with a true freshman hoping to cover #4 WR McMillan.
- :32 – Midfield coverage has been an issue with OSU, I don’t have particularly high grades for the starting (and essentially only) ILBs and the DBs are stretched thin by their many responsibilities, particularly on a blitz like this one.
- :47 – McCoy’s beat off the break here and then overruns the play, giving up an extra five yards and a first down – effectively tackling and preventing YAC is the freshman aspect of his game and where he gets the lowest grades on my tally sheet.
Rush defense has been operating nominally for the Beavs for the past three seasons: decent at preventing explosive plays, but liable to give up efficiency runs. They’re underwater, by pretty much exactly the same amount as they were in 2021 and 2022, in per-play success rate against the run at 45.5% (91 vs 109), but limit opponents to an okay 4.8 adjusted YPC and a perfectly average 15% of designed runs gaining 10+ yards.
Here’s a representative sample of successful rush defenses:
- :00 – This is what the lion’s share of rush defenses look like for OSU – the front four making a mess of the middle against a bad offensive line, which are replete in this league, and then the rest of the defense playing heads-up football to come get the back while he’s dancing around trying to figure something out.
- :08 – Since Bray embraced the switch to a 2-down front, they’ve really emphasized reading and slanting along the line, which gives their all-veteran rotation an advantage on outside plays like this by winning at the point of attack and just maintaining the pursuit.
- :14 – One trick that Bray kept from Tibesar is doubling up on zone read runs – the OLB goes for the back to induce the keep, but the DB is right on his heels for the QB, and neither have to worry about clouding the read. It nearly 100% kills this play as designed, without a TE slice to kickout the DB or an RPO tag to exploit it.
- :21 – The backers’ best attribute is that once they’ve decided to attack the run, they’re missiles, and can blow up much bigger (albeit technically unsound) o-linemen.
What continues to strike me as odd about the Beavs’ Pac-12 opponents is that, despite the fact they’ve had this vulnerability to efficiency rushing for three years now, opposing offenses choose to pass against them more than 60% of the time. In fact, OSU’s short-yardage rush defense rates, which were under 30% in previous seasons but have collapsed to 26% on 2nd & short and an appalling 7% on 3rd & short in 2023, would seem to invite automatic conversions in those situations. And yet opponents still pass the ball half the time (and fail on half of those passes) in both 2nd and 3rd & short. I reached my wit’s end watching UCLA throw away a sure victory running the ball with effortless success against OSU, and came close to concluding Coach Kelly was deliberately trying to get fired by botching his QB and redzone management.
Here’s a representative sample of OSU’s failed rush defenses:
- :00 – I don’t have any obvious deficiencies for the defensive line, but I don’t have a lot of big plus talents or size either, and if the OL is playing properly they can be handled without much issue. The flipside for the ILBs is that when they’re not attacking aggressively, they tend to stand pat and just wait to get hit rather than flowing to the play.
- :06 – The wide trips have one of the starting ILBs out of the box and the backup ILB Jordan in over the center, and he’s showing some inexperience misplaying this – he needs to leave the boundary to the safety and get in that A-gap.
- :22 – The little hesitation is essential to this play, because while the front does get some lateral movement, they don’t have the power to break through well developed outside blocks or the speed to catch a back once he’s gotten going to the outside. UCLA averaged over nine yards a carry running to the wide side prior to garbage time on plays like this.
- :31 – Another baffler to me was how rarely Pac-12 offenses called gap schemes like this counter play, since OSU has only a 31% success rate at defending them despite it being in every Pop Warner team’s playbook. Again the backers are not squeezing down as they need to be, just waiting for contact.