TENDENCIES: 2023 Full-Season Defense
Blackshirts were sick!!
Glossary of Terms1
After each game, I chart every snap to log various elements and variables of how Nebraska played schematically and how successful each was. I then add up the results in a big spreadsheet that shows me each game from the whole season and a cumulative total, giving me percentages and letting me see trends on how the Huskers are operating and what is working.
College football teams have been doing something like this — though it’s light-years more advanced, digitized, and accessible now than it was even 20 years ago — for decades. Coaches have been looking at and tabulating things like, “How often does a team run to the boundary on first down?” or “What coverages are we going to get on third down?” since the modern era of football began. People talk about football analytics like it’s something new, when teams have been compiling and evaluating data and providing it to help their players for 50 years. They just didn’t call it analytics back then.
This is a pretty rudimentary attempt at compiling the same things Nebraska might be. Below are charts showing various percentages of tendencies on how Nebraska operated in 2023 and how successful those tendencies were by success rate. I’ve broken it up further into the first half of the season (the games against Minnesota, Colorado, Northern Illinois, Louisiana Tech, Michigan, and Illinois) and the second half of the season (the games against Northwestern, Purdue, Michigan State, Maryland, Wisconsin, and Iowa) so that we can compare and contrast how NU’s modes of operation changed as the season progressed and how successful they were. That’s in the first two sets of columns. I’ve also provided a third set of columns with the cumulative, full-season totals. On the left of each set of columns, you’ll see the percentage for how often a particular data point was used by Nebraska, and on the right of each set of columns, you’ll see the success rate for plays with that data point in gray.
That’s a lot of information, but it’s simplified if you think about it this way: The darker red column is how much Nebraska used something in the first six games of the season, the lighter red column is how much Nebraska used something in the final six games of the season, and the black column is its combined total for the entire season — and the gray columns next to each are how often that element worked in that part of the season.
I’ve broken the categories up into General Performance, Personnel, Opponent Operation, Fronts/Box, Pass Rush, Safety Alignment/Rotation, and Coverage. Some of the numbers in the graphics come off kind of small in the thumbnail view, but you should be able to click on the graphic and it will enlarge.
General Performance
The incredible turnaround by the defense from the poor play of 2022 was a nice, feel-good, positive story in the first six games of the year, but the unit became a legitimate, game-wrecking, top-15 force in the second half of the season. In the back six games, NU improved its overall success rate seven percentage points to an absurd 67% (11th nationally over that stretch), allowed just 4.1 yards per play (fourth nationally over that stretch), and allowed just 2.36 points per scoring opportunity (fifth nationally over that stretch), all while cutting the explosive plays it allowed by 25% and boosting its own amount of caused havoc plays. And the NU defense did this over a stretch where it faced only Big Ten teams. The Blackshirts were one of the best defenses in the country in the final six games of the year, full stop.
One reason they were better over the back stretch was an improvement on key downs. NU was at about a 58.7% success rate on combined third- and fourth-down snaps in the first six games, but raised that to nearly 70% in the second half, one of the best rates in the country. It also had a big jump in effectiveness on passing downs (any second downs of 8 yards or longer or third and fourth downs of 5 yards or longer) going from a 63% success rate in the first half on passing downs to a nuts 77.5% of success rate in those situations in the final six games. A consistent theme in the first half of the season was NU forcing opponents into long down but letting them off the hook by allowing long plays or conversions — especially against Minnesota, Colorado, and Michigan, all losses — but the Huskers really cleaned that up in the back half of the year.
The issue in the second half of the year was on shorter downs. Nebraska was poor on short yardage in the final six games of the season, having just a 46% success rate in third- and fourth-and-3 or shorter situations, a fall of 10 percentage points from the first half. NU’s 3-3-5 defense seemed to often be getting overpowered often by heavy personnel groupings in those final six games, especially in situations that kept drives alive against Wisconsin and Iowa.
Personnel
The 3-3-5 defense is what’s known as a “base nickel” scheme, meaning its standard, typical way of operating is to have five defensive backs on the field; in a 3-4 or 4-3 scheme, the base defense would have four defensive backs. Another hallmark of the 3-3-5 is an increased flexibility and hybridity of your personnel that allows you to do more without having to substitute players.
With that flexibility came less of a need to operate in more distinct personnel groupings. In 2023, NU played in its base look of three down linemen, three linebackers, two corners, two safeties, and a Rover safety/linebacker hybrid on over 85% of its snaps. For comparison, Nebraska’s most used personnel last season under Erik Chinander and Bill Busch was at about 70% usage. NU in 2023 packaged personnel less than any Husker team in recent memory.
Coordinator Tony White did have some distinct looks, though. NU played about 7% of its snaps in what I’m calling its version of a “nickel” package, where it either replaced one of its safeties with a true corner or took a linebacker off the field to play with a sixth defensive back on clear passing situations.2 Nebraska also played a handful of snaps in a heavy goalline personnel, with five or six defensive line bodies on the field.
Maybe the most interesting thing here was White’s debut of a pass-rush package after the bye, a look that brought on light, lanky sophomore walk-on James Williams as a designated pass-rusher on long down situations. Nebraska used it at increasing amounts in the second half as it proved more and more effective: It was used on 4.9% of snaps against Northwestern, then went to 16.7% against Purdue, 21.1% against Michigan State, and 21.2% against Maryland, getting 14 total snaps against the Terrapins. Williams was shut down after that game to preserve his redshirt, and the package saw no use over the final two games, but it still finished at about 10.6% usage in the second half of the season and 5.9% of NU’s snaps in 2023. And when it was used, it went absolutely off: The package with Williams had an 82% success rate — because it was almost exclusively used on third downs, this means the plays using Williams were getting the opponents off the field on nearly 80% of the third downs. There’s some slight noise to those numbers — the longer downs when Williams was used are naturally more difficult for an offense to convert — but it was undeniably a wildly successful look for the Huskers.
Opponent Operation
Nebraska was consistently one of the best teams in the country against called runs for the duration of the season, but its improvement in efficiency against the pass in the final six games was what took the unit to a truly elite level. Against called pass plays in the first half of the year, NU had a 57% success rate — still good, but around 30th nationally3:
But over the isolated final six games, that rate shot up 11 percentage points to 68%, 13th nationally for that period:
The improvement against the pass came from down-to-down efficiency in stopping pass plays and not from limiting big plays; NU’s explosiveness is steady around the 1.5 mark on both charts across the whole season, but its *efficiency* improved in the final six, with the success rate moving to the left and passing several teams.
These are raw, non-schedule-adjusted numbers, so the quality of the passing offenses Nebraska played in the back half of the year is certainly a factor in this. NU played two elite passing attacks and quarterbacks in the first half in Colorado and Michigan, but over the back six games got to face Michigan State and Iowa (two of worst passing offenses in the country) and played a game against Northwestern’s backup QB.
But not all of this is schedule-based. Against the bad teams, NU dominated — with success rates of 68.5%, 65.5%, and 73.3% against called pass plays in the games against Northwestern, Michigan State, and Iowa — which is what you want to see a good team do when you face poor competition. And the other three games — vs. Purdue, Maryland, and Wisconsin — were against league-average or better passing attacks with talent and experience at quarterback and receiver. The Husker pass D was arguably *better* in two of those three contests, with a 75.8% success rate against called pass plays against Purdue and a 65.9% rate against Maryland.4 The competition was worse, but the pass defense *did* show tangible improvement in the second half of the year, regardless of competition.
The pass defense’s jump was also evident in that opposing teams in the second half — facing one of the best run defenses in the country — elected to run the ball *more* instead of challenging the Nebraska pass D. Teams called pure runs or run-pass options at around 37% of their snaps in the first six games. Opponents’ in the second half, though, were at 47% — despite Nebraska’s run defense actually slightly improving its play in those games. Opponents in the first half of the year had an easier answer to Nebraska’s tough run defense: Throwing the ball into an good-but-not-great Husker coverage unit. For the final half of the season, with the pass defense also entering the top 15 nationally, that easy answer went away, and teams were forced to contend more with trying to run the ball.
The other interesting piece of data here is the success against run-pass option calls. NU was good against them in the first half of the year, but saw a dip in performance against RPOs in the second half of the season. The drop in numbers was largely do to two consecutive poor performances against these types of plays against Maryland and Wisconsin. NU had just a 42.8% success rate against RPOs against the Terrapins, then was at 56.52% against the Badgers — both key factors in those teams’ ability to have some success on key plays against NU in those wins. But removing those two games, the success rate was 68.2% in the second half. It was really just those two games.
But still, it’s an interesting data point: The 3-3-5 defense is known for generating confusion with post-snap movement and rotations and unorthodox rush patterns. RPOs are plays that allow offenses to change whether they’re running or passing or where the ball is going after the snap. I don’t think it’s a coincidence that the two conference teams NU played that used a lot of these optionality-based plays had more success against the chaotic 3-3-5 style.
Fronts/Box
An under-discussed element of this season was how drastically Nebraska altered its general front-seven plan from the early part of the season to the end.
In the first six games of the year, Nebraska was true to the 3-3-5 as a predominantly three-player front team (on 46.5% of its snaps), while also using the occasional four-player front with the Jack linebacker walked down (26.7%) and five-player fronts against multi-tight-end sets (26.7%).
The structure of the box changed dramatically in the second half of the year, though. NU transitioned to playing with a four-player front on over 60% of its plays, dropping the use of three-player fronts by over 30 percentage points to 15%. That’s an insane drop. The use of five-player fronts was static though both halves of the year.
So much of the talk in the preseason about the 3-3-5 had been the creativity and flexibility and weird rush packages and how they would adapt to the Big Ten, and then … NU ended up playing a pretty traditional 4-2-5 shell for most of Big Ten play. This isn’t to say White’s pressure packages or third-down strategy weren’t unique, because they were, but this ended up being a lot less of an experiment than it looked like in the preseason.
It’s hard to say the move to a quasi 4-2 wasn’t effective. NU’s snaps with a four-player front had a 67.5% success rate in the second half of the year, while the effectiveness of the three-player-front shells fell to 59.2%. Nebraska’s three-player front was dominant in the first half, but either through other teams catching up or because of the Big Ten’s physical, run-the-ball nature, it wasn’t quite the same fastball in the later part of the year.
With more of the four-player fronts, we also saw heavier boxes by Nebraska in the second half of the season. NU played with a heavy box (at least one more defender within five of the tackle box than the offense had blockers) at 26.4% in the final six game, a seven percentage point uptick compared to the first six. In both halves of the season, it was worse out of heavy boxes than it was out of both light and standard boxes, likely a reflection of teams having success out of heavier personnel against the base nickel scheme.
The other big change In the second half was White turning up the misdirection and pre-snap motion and shifts in the front seven. He had some element of a shift on about 20% of Nebraska’s snaps in the first half but would nearly double it to 38% in the second. Plays with a shift in the first six games were much more effective by success rate than those without, so he likely made this adjustment after analyzing the trends of the first half.
Pass Rush
With the increase in four-player fronts came a natural boost in four-player rushes, but there were some other interesting changes to the pass-rush pattern in the second half of the season.
Nebraska also predominantly used four-player rushes in the first half, at 32.2%, frequently bringing an extra rusher when in the three-player front. NU also used four-player rushes the most in the second half, but that number jumped about 13 percentage points in the final six games as NU went to the 4-2 scheme. Most of that increase, as discussed, came from the decrease in the three-player-rush, Drop 8 package White largely stopped using in the second half; White brought three rushers on 29.8% of Nebraska’s snaps in the first half, but only on 6.1% of snaps in the second. That makes sense.
But NU *also* boosted the amount of five-player rushes it was bringing by a huge margin, sending five on 37.4% of its snaps in the final six games, a 12 percentage point jump from the first six games. That’s in addition to a steady rate of six-player rushes on about 10% of snaps throughout the season.
If there was one general trend from the first half of the season to the second, it’s that NU was just bringing way more rushers in aggregate. The Drop 8 light rush plays almost entirely went away, replaced not only with more standard rushes but with additional five- and six-player blitzes.
Overall, White did just decide to bring more heat in the second half of the year. The blitz rate was at an already-high 37% of the plays in the early part of the season, then jumped to 47% in the back stretch. They were bringing heat on very nearly half of their snaps in the back half of the season. The number of “sim pressures” or “creeper pressures” — disguised blitz looks where rushers and coverage players come from unexpected places — also rose by three percentage points in the second half.
These increases probably happened because they were working. Nebraska’s plays with blitzes and without blitzes had a virtually identical success rate in the first half, of around 60%. But in the final six games, NU’s plays without blitzes held steady at around 60%, while the snaps with blitzes skyrocketed in efficiency in the final six games to a nearly 72% success rate. Nebraska’s blitzes were generating a successful play for the defense on nearly a quarter of their snaps in the second half. Insane.
The pattern of where the blitzes came from also changed a bit. Nebraska predominantly blitzed into the interior A gaps (between the offensive center and the offensive guard) in the first half of the year, with about 42.8% of its total extra rushers going through the A gap. NU still blitzed a lot through the A gaps in the back half — 36.5% of the time — but also raised the amount of pressure it was bringing off of the edges. Its blitzes through the C gaps (the gaps outside the offensive tackles) raised eight percentage points to 39.9% in the final six games, the most-blized gap of the second half. It also increased the amount it blitzed through the D and E gaps (the gaps outside one or two tight ends) by a percentage point.
Outside blitzes had been effective in the first half of the year, with a 68% success rate on blitzes through the C gaps and an 83% success rate on blitzes through the D/E gaps in the initial six games. They remained super effective, at 70.4% success and 82% success rates, respectively, in the second half. But the blitzes largely improved so much in the final six games because of how much better interior pressures started working. Pressures through the A gaps in the first half were successful just 54.2% of the time — a significantly worse rate than NU’s non-blitz plays. But their success rate shot up almost 30 percentage points in the back half of the year, to 81%. That’s a gargantuan increase and a super high success rate for any type of play. Nebraska’s interior blitzes were working about half the time in the first half, then started working 8 out of 10 plays in the second. That’s how your pass rush and defense are able to make a jump.
NU’s data on the use of stunts or twists by the defensive line remained largely static over the whole season, with plays that feature stunts performing slightly worst than non-stunt plays, but not by a huge margin. When NU did stunt, it greatly preferred “E-T stunts,” with an interior lineman looping to the outside of the formation, over bringing outside players in.
Safety Alignment/Rotation
The overall distribution of the safeties stayed pretty consistent throughout the season, with an increase in single-high shells and decrease in three-high shells by about eight percentage points each in the second half of the season. But, generally, throughout the year, NU spent about 60% of its snaps in single-high shells, 30% in two-high shells, and a handful of snaps a game with all three safeties deep and/or down in the box.
Nebraska in the second half of the season played better, though, with more safeties deep. Its success rate in both two-high and three-high shells in the final six games of the year was over 70%, which was better than the plays with a single-high shell (65.7% success rate) and zero-high shell (46.7%) success rate. The plays with no safeties deep were generally short-yardage situations or in the red zone, which sort of correlate with the previously discussed struggles in situations against heavy personnel. Nebraska’s three-high alignment, which often rotated into a one- or two-high shell after the snap in an attempt to confuse the quarterback, was dominant all year, with the highest success rate of any shell in both halves of the season.
The rate of rotation by the safeties was also steady throughout the year, staying at around 30% of plays. The plays where NU rotated safeties, though, kind of got them burned. Plays with rotation were consistently worse than plays with static safety alignments in success rate, though not by huge margins. When NU was rotating, it was mostly rolling from two-high to one-high and from three-high to two-high.
Where the individual safeties aligned amplified slightly in the second half of the year but didn’t dramatically flip. The boundary safety was NU’s primary deep, over-the-top cloud safety, aligning deep on about 70% of snaps in both halves of the season, while the field safety was primarily a box player, spending 70% of their time down in the first half of the year and 82% down in the second half of the year. The Rover was about 60-40 deep to down in the first half, then went to about 50-50 in the final six games. All the success rates were consistent throughout the year.
Coverage
Two noticeable trends developed in Nebraska’s coverage in the second half of the season: A major bump in man coverage over zone, and a big increase in press coverage over off.
NU spent 35.8% of its snaps in man coverage in the first six games, a pretty middle-of-the-pack rate nationally, but the rate of man shot up 15 percentage points in the second half of the year to 49%.
Part of the reason NU likely played more man is just because it got better at it. The success rate for man coverage was a pretty mediocre 55% in the first six games — nine percentage points worse than the success rate in zone coverage — but in the final six games, the secondary really improved its ability to cover one-on-one, with a 68% success rate in man. That’s a huge improvement midseason. It’s crazy to flip your overarching coverage philosophy in the middle of the year, but you can do it when your players make jumps like that.
Part of the boost to man coverage was an improvement at press coverage and a subsequent increase in its use. NU’s success rate in press was about 53% in the first six games, but it worked at about a 66% success rate in the second half. NU boosted how much it ran press, accordingly, from on about 26% of snaps to about 40%. White had the DBs playing off a lot in the early games when it did run man coverage, which was letting up a lot of easy completions underneath. That went away a bit, as he more often trusted the corners — especially the returning Tommi Hill — to play closer to the line without getting burned. It doesn’t explain all of the success for man coverage in the second half, but that tweak by White does explain some.
With the increase in man coverage, the rate of Cover 1 naturally shot way up in the second half, reaching 44.5% of NU’s total snaps from 28.7% in the first half. NU still played about the same amount of Cover 3 in the back half of the season — 29.4% compared to 32.8% — so the move toward more man coverage came predominantly at the expense of NU’s two other primary zone coverages, two-high-safety coverages Cover 2 (13.7% to 10.4%) and Cover 4 (11.6% to 4.4%).
The last notable thing, which I’ve already mentioned in other sections, was how little Drop 8 coverages Nebraska ran late in the year. It used these plays — in which three players rush and eight players drop into coverage — on an extremely high percentage of its snaps in the first six game — almost 30% of its snaps — then ran them on only 6.3% in the final six games.
Yards Per Play measures how many non-penalty yards NU allowed on a possession divided by its non-penalty snaps. Success Rate measures how often NU prevented a gain of 50% or more of the yards its opponent needed to convert on a first down, 70% or more of the yards its opponent needed on second down, or 100% or more of the yards its opponent needed on third or fourth down. An Explosive Play is any designed run that gains more than 12 yards and any designed pass that gains more than 16 yards. A Havoc Play Allowed is any tackle for loss, sack, fumble, interception, pass break-up or batted ball.
With six defensive backs on the field, some would call that a “dime” look for a defense, but my tracking didn’t differentiate on the season. And because Nebraska’s defense plays in base with five DBs, I just called it nickel by an “extra pass defender philosophy.”
This chart visualizes both success rate and explosiveness allowed of pass defenses. Explosiveness allowed is on the vertical axis, with success rate on the horizontal axis. This measures against plays where any pass is thrown, whereas my data differentiates between called pass plays and RPOs, so the numbers are a little different.
It was less good but still solid against Wisconsin, at 59.3%.