Does Oklahoma have a problem with run defense under Grinch?

That’s a hot topic of conversation these days, especially in light of Bill Connelly recently identifying the run defense and small size of the Oklahoma defensive line as a primary concern for 2020.

From Connelly:

If … the run defense finds some oomph. After crumbling to 84th in defensive SP+ in 2018, OU’s defense rebounded to 48th in Alex Grinch’s first season as coordinator. But the run defense was all-or-nothing: OU was sixth in stuff rate (returning end Ronnie Perkins, tackle Jalen Redmond and linebacker Nik Bonitto all made 10-plus stuffs) but 80th in rushing success rate allowed and 103rd in rushing marginal explosiveness. Of four returning contributing tackles, none is listed at even 280 pounds. Grinch loves speed, but there’s probably still a minimum girth level you have to clear.

-Bill Connelly

I think there’s a few holes in this reasoning but also some obvious truth that he details in the pure numbers. Connelly noted further on Twitter that OU’s 3-down defensive front would seem to be a risky schematic venture given the size of the players on their depth chart:

Hey OU fans, am I missing someone here, or are you guys gonna try to field a 3-man front with no 300-pounders (and no 280-pounders who played last year)? pic.twitter.com/aVSKgdoHhR

— Bill Connelly (@ESPN_BillC) May 28, 2020

There’s a lot of complexity to examine here because Alex Grinch doesn’t run a standard 3-down defense. This isn’t the two-gapping system of Mike Stoops that had Neville Gallimore bulking up to like 320 pounds before Grinch had him slim back down to the 302 where he ran a 4.79 at the combine. He’s not going to be replaced by a guy on this list though but likely one of their JUCO defensive tackle additions, probably Perrion Winfrey (6-4, 305).

I think Connelly is making some interesting points here, then there are some details being missed, and there are bigger philosophical questions at play.

The Alex Grinch defensive system

There’s a play from the playoff semifinal against LSU that really illustrates the structure of the Oklahoma defense. Much like the Art Briles Baylor offense, last year’s LSU offense was really good at revealing the underlying structure of a defense because it was designed to stretch teams out and hammer at their foundations.

Here’s the play:

Here it is on the chalkboard:

Oklahoma is playing press-quarters. The cornerbacks are in press-man on the outside receivers, free safety Patrick Fields is playing as a robber and late run fitter, then the nickel (Brendan Radley-Hiles) and strong safety (normally Dellarin Turner-Yell, this time Justin Broiles) bracketed the slot.

Radley-Hiles was basically an off-man corner here and the safety would bracket inside routes, like theoretically this glance route, and then fit the run. The late movement by Radley-Hiles to back off and the strong safety to play aggressively, ideally robbing in-breaking routes on his way to the run, is intended to solve “the 4-down RPO math problem.”

You have six defenders up front for six blockers but to aggressively stuff the run you normally want to get one extra guy. Against the I-formation that’d mean an 8th man in the box but in spread world with three receivers, it’s the 7th-man that is your “load the box” solution.

Quarters defensive schemes were designed to allow the offense to get numbers wherever they’d be needed AFTER the snap. In this case, Oklahoma’s press-quarters scheme was designed to have the strong safety hawk down pretty quickly and be the 7th man off the edge. The defensive breakthrough of quarters schemes though lead to the offensive counter, the spread RPO revolution.

With RPOs, offenses can attack defenses when they try to get up to the standard defensive tactic of outnumbering the run. That’s what happened here. Both quarters and RPOs are both much older than this game so at this stage things are much more complicated but that’s the underlying battle that’s taking place.

Grinch’s speed D

Oklahoma’s “speed D” is designed to try and solve issues like the 4-down RPO math problem in the same fashion that defenses have been solving these issues for the last few decades, by getting more speed on the field to reduce the stress of run/pass conflict and spacing.

It’s not just on the back end that he tries to emphasize speed though, but also up front. Here was the front for Oklahoma by the end of the season:

I think it’s fair to guess that some of Oklahoma’s slower (per SPARQ) defenders probably run faster now after some college S&C and with training in how to run the 40. Ronnie Perkins in particular.

However, you can see that the design and bent of the scheme and personnel choices was to get speed up front, even at the expense of playing a pair of swing end/tackle players in Perkins and Jalen Redmond that were recruited as weakside ends and potential “jack linebackers” in the Mike Stoops defense. The entire offseason lead up to 2019 Oklahoma sites were writing about how Redmond’s return from blood clot issues was going to answer the question of who’d offer pass-rush as a jack linebacker.

I was skeptical about his return from taking so long off, but he did get into good shape and play really well, but not at jack linebacker. He’s like 270 now and spends a fair amount of his time as a 3-technique.

That’s normal for Alex Grinch, who’s boundary end/tackle in his breakthrough Washington State defense was the 6-2, 254 pound Hercules Mata’afa.

So what’s the game here? To have lightning quick defensive linemen shooting gaps and causing problems with penetration into the backfield that are quickly supported by a press-quarters scheme that defers stress to the nickel and corners in order to get the safeties involved quickly in the run defense. The goal is to swarm and outnumber spread rushing attacks.

The linebackers win in this set up as well, with both safeties playing as flat-footed robbers rather than deep 1/2 defenders the linebackers can aggressively plug gaps, they can afford to play fast and trust the safeties to make them right.

Well theoretically they could, as you can see in the LSU clip Murray didn’t really play interior gaps very quickly. He was better when Grinch utilized his aggressive bear fronts, playing five on the line of scrimmage before or after the snap. The advantage of that set-up for the Sooners, which they used regularly, was that with five on the line of scrimmage they’d get 1-on-1 matchups for all of their speedy linemen up front.

This is a large reason for why you see a stat like Connelly was noting in which the Sooners were very good at inflicting run stuffs and tackles for loss (also sacks) while struggling in preventing explosive runs or success rate. If opponents broke through, explosive gains were there to be had because so many Sooners were up tight or in the backfield. If opponents lined up intending to power through Oklahoma’s aggressive fronts, then their size could be turned against them in situational play.

Flaws in the speed D philosophy?

It’s hard to find all that many games where run defense actually hurt the Sooners in 2019 and looking ahead to 2020 the more serious concern would appear to be the improving passing attacks around the league.

The teams that did hurt Oklahoma running the ball did so due to slow recognition and fills from Kenneth Murray and spotty play in the secondary from the safeties and nickel in run support. Obviously Radley-Hiles wasn’t a plus run defender but they often played bracket coverages like the one above against LSU designed to use DTY as the 7th man instead.

Last year’s leading tacklers for Oklahoma were Murray with 102, DTY with 75, Fields with 64, and then White with 51. This is your boom/bust dimension that Connelly was referring to. Either the defensive line made a stop behind or at the line or the safeties were often cleaning it up.

DTY and Fields figure to start at safety again unless younger guys like Woodi Washington or Jeremiah Criddell are ready. There’s a lot of talk about moving Radley-Hiles, to corner or to the bench, to get a better run support safety in that spot. The problem with that is that Oklahoma’s press-quarters scheme actually works best when the nickel can play some man coverage and carry a vertical, as do their cover 3 schemes. The Sooners need better recognition from their linebackers and rangier, better play from the safeties. That’d help more than replacing Radley-Hiles.

But the real problem that Connelly highlight is Oklahoma’s situational run defense. What’s the plan on 3rd and 1 or the red zone? Modern defenses can give up gobs of yards between the 20s and be fine, especially when the offense on their side is coached by Lincoln Riley, but they need to be able to match up and make stops in key situations. If you want to stop points you need to be good in those big spots.

Let’s look back at one of Oklahoma’s most glorious defensive performances in 2019, the Red River Shootout. Texas had two 3rd downs in that game of five yards or less and three “goal to go” situations. Here’s all five of those plays, starting with the two 3rd down conversions.

In both instances Texas just hit a timing route outside to Brennan Eagles (6-4, 230) vs one of Oklahoma’s smaller cornerbacks in a 1-on-1. Here’s the three goal-to-go instances that took place in the game.

In all three instances Texas is using formations and plays designed for situational dominance. They have an unbalanced formation in the first example with the left tackle lined up as a tight end and the tight end lined up as a left tackle and just bully over the Sooners. In the second and third instances they run Sam Ehlinger on power and stretch concepts and just march in easily. It’s all 11 personnel so they were able to get into this sets quickly without having to substitute and Oklahoma could barely even line up, much less match up and stop it.

The big success Oklahoma had in that game was avoiding crucial money downs on defense, not winning them. When Texas was in 3rd and manageable and goal-to-go the Longhorns were dominant, they just couldn’t get there very often.

Other teams had some similar successes, the Sooners ranked 17th in 3rd down conversion defense but were 128th!!!! in red zone defense. Opponents took 48 trips to the red zone and scored 32 touchdowns while kicking 13 field goals. That means opponents scored on 45 of 48 trips.

Recall that Oklahoma’s season would have been spoiled had the Cyclones made a better play or gotten a defensive pass interference call on their end of game two point conversion. This was an achilles heel that Big 12 teams struggled to hit. LSU blew away Oklahoma on every down so money downs didn’t even matter.

Disguise, speed, and aggression don’t work as well against tempo and targeted play-calling and that’s a potential issue for the Sooners next season as well.

An equally big problem for this approach that Grinch will need to solve is the effect of his run-stopping, standard down defense on his pass defense. LSU’s RPO torched Oklahoma because the Tigers found the 1-on-1 matchups in coverage that Grinch’s press-quarters scheme relies on and attacked it. Ditto for Texas’ 3rd down conversions with the drop back game.

How well is all this going to work when Oklahoma is facing Tylan Wallace, Xavier Hutchinson, or Brennan Eagles and doing so without Parnell Motley? The Big 12 in 2020 will be much better equipped to invite the swarming, 7-man fronts and attack the 1-on-1s outside where a defensive defeat could mean six points.

The philosophy of swarming and attacking opponents by sending numbers in quarters schemes has been tweaked and modified for years by Grinch and other coaches but is in an overall decline. The increasing efficiency and skills of offenses to attack them outside with the passing game just makes it too risky, you can end up avoiding 3rd and 1 or goal to go because the offense scores on a 50-yard pass.

Grinch’s philosophy is to attack offenses and prevent them from getting into favorable situations. The more prevalent strategies taking hold are designed to stop points by keeping the offense in front of them and then making offenses execute in key spots with points on the line.

Is it better to avoid the red zone or to plan to make your final stand there? The Grinch approach worked in 2019 (until it didn’t) but can that work in 2020?