The battle for Total Basketball, Vol. 2: 40 Minutes of Houston

Or how Jurgen Klopp and Kelvin Sampson are kindred spirits

Hi! This is part two of a three-part August series on two of the three best defenses in college basketball. The first one on Iowa State and their Spanish footballing counterparts, Real Sociedad, is here. Read that first so you get a feel for what this is going for. On with the show.

The last three years of college basketball, via Bart Torvik’s website, looks like this. You can first sort it by that most basic of metrics: wins.

Or you can sort by overall efficiency, a similarly simple metric. How good have you been over the last three years?

Perhaps mere wins don’t do it for you. What about the best defenses in college basketball? After all, every analyst loves to note that Defense Wins Championships.

Maybe you want a list of programs with top-10 units on both sides of the ball, which is a two-program list: Houston and UConn. Or you’d like a list of teams with 20 or more Wins Above Bubble (your W-L versus how a bubble team would’ve done) since 2021, which is a list of three: Purdue, Houston, and Kansas. Programs that are in the top 10 of both defensive TO% and OREB%? That list is exclusively made up of Houston. The #1 team in opponent 2PT%? Houston. The #1 team in opponent 3PT%? Houston. The Power Five team with the most wins on the road since 2021? Houston.

Not that you need me to sell you on these guys. Houston has 11 NCAA Tournament wins post-COVID, the most behind Gonzaga and UConn (both 12). They enter this year as Torvik’s #1 team; whenever the various media polls are released this October they’re a pretty safe bet to be top-5 or top-3 in most of them. They’ve posted four straight top-5 KenPom finishes, the first team to do that since 2018-2022 Gonzaga. (Before that, 2015-19 Villanova.) If they do it a fifth time this winter, they’d be the first team to go top-5 five in a row since 1997-2002 Duke. Fairly lofty for a program with a peak finish of one Final Four appearance.

As Eamonn Brennan wrote recently:

The past five years have provided a steady stream of evidence that Sampson’s Houston is now always excellent, always getting better, and always capable of handling whatever challenge their season or their schedule or their personnel presents, above and beyond whatever minor qualms you might have about the roster.

First and foremost is that defense, the single most stifling and nasty unit in the entire sport. No team allowed fewer opponents (8) to go 1 PPP or more than Houston last year, a fairly notable thing when you’re playing against a top-25 strength of schedule. Even in their eventual Sweet Sixteen downfall to Duke, which came after almost half the team got injured, the Cougars still held Duke to their fourth-worst offensive day of the season.

Here, we’ll answer four questions:

1. How does blitzing work in college basketball, and why is Houston the best at it?

2. Why does Houston’s smaller frontcourt actually lend itself to making them the best defense in the nation?

3. Who are the Big 12 opponents that cracked the code even slightly in Houston’s dominant Year One?

4. Why can I not shake the thought that Houston is college basketball’s Liverpool?

These are FREE FREE FREE since this is the offseason, but subscribe below if you’re a new(er) reader.

---

The Houston blitz

In the previous piece, I made a few references to a Jordan Sperber post from last year about what he called the ball screen spectrum. Included in it was a fun graphic:

With the following explanation.

there’s not necessarily a right or wrong way to defend the ball screen. In recent years, defenses like Houston and Virginia have had success being aggressive against the ball screen — attacking the ball handler and getting it out of his hands. On the other hand, defenses like Creighton and Alabama have had success doing the opposite — sagging off of the ball handler and turning him into a scorer.

…

The type of ball screen coverage used on a given play is not currently tracked at the college level. So we can’t explicitly measure how aggressive a defense does (or doesn’t) guard the ball.

However, Synergy Sports does track every ball screen that directly results in a shot or a pass that leads to a shot. Using that data, we can approximate how aggressive each defense plays.

The formula itself is simple: total number of shots originating from ball screens divided by total number of shots by the ball-handler themselves. The lower the ball-handler percentage is, in theory, the more aggressive your screen coverage(s) are. A team with highly aggressive coverage will probably have their average P&R set look like this:

Norfolk State, #1 in Aggression Index, defending Howard in MEAC Tournament

While a more deep-drop defense may look like this.

Indiana State, #363 in Aggression Index, defending Seton Hall in the NIT

There’s no right or wrong way to do it, necessarily. The fourth-best defense in college basketball last year was UConn, who played a drop coverage most of the time in pick-and-roll sets. The best, Iowa State, and third-best, Tennessee, didn’t place in the top 20 of this stat. But the second-best defense of 2023-24, and the best defense of the last three years, has been #1 or #2 in Aggression Index for three seasons running. Houston blitzes the ball-handler harder than anyone else in basketball right now. I’d argue they do it better, too, by finishing top-3 in P&R defense among high-majors for three years running.

Take this possession against Baylor, the best offense in the Big 12 and the second-most efficient P&R offense in the league behind Texas Tech. Baylor averaged 31 points per game from P&R sets last year, per Synergy, and their bread and butter was any set where RayJ Dennis could come off a screen to find either a spot-up shooter or the roll man headed downhill. Theoretically, against a Houston defense that blitzes, you should be able to turn 1-on-2 possessions into 4-on-3 possessions. Theoretically.

This was one of 19 Baylor turnovers on the day, but more importantly, it was one of 8 by Dennis. This was a career-high by a fifth-year senior with roughly 4.5 seasons of starting experience and the second-highest Assist Rate in the Big 12, behind Houston’s own Jamal Shead. If that guy is rattled by this system, imagine how teams with lesser point guards or simply lesser offenses performed.

Houston played 15 of its 37 games against teams outside of Quadrant 1 in the NET, i.e. outside of the top 50 on a neutral floor. Those 17 teams combined to turn the ball over on an astonishing 29.2% of their possessions. The next-best rate, of course, was Iowa State. When this system takes on lesser competition, even if said competition isn’t all that bad, it gets ugly quick. Kansas State, a team with Tylor Freaking Perry as their point guard, looked like UMKC.

The Houston blitz is extremely hard to beat for a multitude of reasons, the first of which can be explored just by looking at the roster. At point, you had Shead, one of the very best defensive guards in modern history. It’s more important who the blitzer is, which is the post/wing coming up. If you have a more lumbering type as your big man, you’re much more likely to leave that guy in the paint and deal with the potential consequences of open jumpers. That big man can’t recover in the same way someone smaller or more agile can, so it’s not worth the effort to send them out for what turns into an offensive power play. Here’s Illinois, one of the least-aggressive defensive coverages in the sport, functionally running at half-speed for comparison.

There, you don’t see even a hint of an aggressive coverage from the big man. It’s a fairly extreme example. Some teams will play a looser drop, which still isn’t a blitz or even a hedge but is sort of like playing a connector piece. That’s more of the ‘level’ style of coverage, if you had to name it. This is an example from Memphis, where you can see the big (Nae’Qwan Tomlin) keep an eye on the guard and a hand on the big as he retreats backwards.

You ask a lot of your bigs here, but I don’t think it asks nearly as much as what Houston does. The three main Houston bigs last year who blitzed ball screens were Ja’Vier Francis, JoJo Tugler, and J’Wan Roberts. All three of these players are 6’7” or 6’8”. All three play either the 4 (Roberts) or the 5 (Francis/Tugler). All three also have wingspans of 7’2” (Roberts), 7’5” (Francis), and 7’6” (Tugler). What you have here are three guys that, by size, are more like large wings. Their feet move like a large wing’s do, speaking generally.

But their arms - their gigantic robot arms - are like forests. Any attempt to beat these guys over the top requires an arc on the ball that is not natural for most point guards. You have to be extremely quick, decisive, and smart with where you want the ball to go. In general, it’s not going to stay in your hands unless you’d like the turnover to go next to your name on the stat sheet. So after you run off the screen, you have what I’d say is a maximum of three seconds in your head to get the ball out of your hands. You are surrounded by the most annoying point guard in America and a guy with a 7’5” wingspan as a bench, any bench, screams at you to do something.

After what feels like forever, but really is only two or three seconds, you make a decision. That’s great. Now, let’s see how your decision looks in real time.

That is attempting to break the Houston defense. That is what basketball-induced stress looks like.

The Houston recovery

The thing about the 4-on-3 power play, or any power play in any sport, is that it does not last forever. A hockey power play lasts for two minutes, generally. Houston’s lasts for three seconds if you’re lucky. The recovery speed of the bigs are what impresses me most, and what caused several of the Big 12’s best this year to struggle to beat this system. Shead is Shead, but it’s hard to imagine this system without elite defensive forwards to pair with him. (If you want to see the Houston system run by Houston without an elite PF/C combination defensively, 2019-20 comes close…and it was still a top-25 unit.)

This year, they’ll be even more important. Shead is gone to the NBA; in his stead will likely be Oklahoma transfer Milos Uzan and returning starter LJ Cryer. Uzan can be a grow-a-Shead if you squint hard enough, with similar steal numbers to Shead through two years. Cryer less so: he’s openly there to bring the boom on offense. The forwards will be more important than ever, which is why the blitz is just one part of the Houston equation.

The scariest thing is how all three of Tugler, Francis, and Roberts have separate defensive identities. Tugler, he of the 7’6” wingspan, is one of eight high-major players in the last 15 years (minimum 12 MPG) to post a 9% Block% and a 3% Steal%. The other guys on that list are players like Paul Reed, Willie Cauley-Stein, Ryan Dunn, etc. Instead of a traditional screen here, Texas Tech’s big slips the screen. Against a less agile, smaller, or less long big man, this pass probably works.

It doesn’t, because Tugler plays this more like a shutdown cornerback than a traditional big man. And then. I’d like for other, smarter folks to name the number of players that can chase this ball into the corner, save it, and turn it into a stolen possession for their team. Look at where this ball lands, then remember that it ends up not being a baseline out of bounds set for Texas Tech.

If there’s a such thing as a traditional center on Houston’s roster, it would be the 6’8” Francis, who merely has a +9” wingspan at 7’5”. He has the highest single-season Block% in modern Houston history at 14.7% in 2022-23 and quality steal numbers to go with it. All of that is fine and good; more important are the on-court stats for Houston when Francis is at center. The Cougars gave up 81.1 points per 100 possessions, adjusted for opponent, with Francis on the court (88.6 per 100 when off). Opponents shot 39.5% from two, turned it over on 25.1% of possessions, and took significantly fewer shots at the rim.

The length of Francis is such that he can actually make a mistake and recover in enough time to pull off what he wanted to, anyway. Here, Jeremy Roach of Duke splits a pick-and-roll perfectly because Francis sets up about a step too wide as the shot clock is winding down. Roach gets by Francis, and in a good number of circumstances, this is either a quality layup attempt or a foul by the big.

Not the case with Francis, who blocks this smoothly without a misplaced action.

So: you have Francis, an elite rim protector of elite rim protectors, and Tugler, an athletic freak built in a lab that plays defense the way a cornerback does. Having two of those guys on the same roster is an embarrassment of riches in itself, but it doesn’t include the single most impactful season-long performer of them all: now super-senior J’Wan Roberts. Roberts’ verticality isn’t nearly the same as Francis, but given four years in the same system, he knows a bit more about where to be and when than a freshman like Tugler may.

Better than anyone else, Roberts plays both the blitz and the hedge well against basically every opponent. Most importantly, he isn’t a foul machine in the way Tugler (7.6 fouls/40) or Francis (5.2/40) are. That allows Roberts to play reasonably big minutes for a team that needs him to do so. In this set, Francis hedges the screen briefly, but recovers quickly because Kyle Filipowski, a very dangerous player, is headed downhill. If Francis stays there, Roberts would have to switch onto Filipowski, which would allow Filipowski to post the smaller Roberts up.

Instead, Francis recovers, which frees Roberts up to feature his best attribute: reading the guard’s eyes. Duke sets a pretty good off-ball screen meant to free Tyrese Proctor up for a shot. Against the average team, this probably works just fine. Here, it gets wrecked by free safety Roberts.

This isn’t uncommon if you watch Houston; he’s just extremely good at sniffing out any open passing lanes. Roberts had 44 steals this past season, and among starting frontcourt players in the entire nation, only Virginia’s Ryan Dunn - a first-round draft pick, by the way - had more. And you could easily make the argument that on a per-minute basis, he was the least of these three monsters.

History rewards those who survive

Essentially, there are three ways to beat Houston’s defense:

1. Similarly to Iowa State, hit open threes where you get them via skip passes.

2. Beat the blitz by utilizing a smart passing big man in the ‘short’ roll, AKA a quick pass out of a P&R set.

3. Play against them on one of their five-times-a-year nights where they shoot 22% from three.

The first one sounds easy enough, and teams with quality passing guards and bigs do experience reasonable success. All five of Houston’s losses came in games where opponents scored assists on 65% or more of their made field goals. Against TCU in a game where Houston shot 22% from three while a fine-not-great TCU offense shot 53%, TCU got a crucial open three thanks to how Houston slants the floor when the ball is driven to one side of the court. Notice here that at the point Avery Anderson releases the ball, all five of Houston’s players are at or on the right half of the court. Jamal Shead is able to recover to the wing in time to force a pass there, but LJ Cryer can’t get to the corner in time to prevent the three.

Against Oklahoma in a pretty rare case of Houston just getting got, the Sooners were able to create a boatload of open looks because of over-aggression by Houston on the perimeter. The blitz works a lot of the time, but when it doesn’t, it can create a sort of over-switching that paralyzes the defense and generates open looks. Milos Uzan, now on Houston (wonder why!), looks like he’s dribbling aimlessly to the center of the court, but what he’s doing is creating functionally a 3-on-2 situation with how heavily weighted Houston gets towards the ball-handler.

The poster child for the second option - the short roll - was Kansas. Of course, the Jayhawks have a pretty rare frontcourt of KJ Adams and Hunter Dickinson, two forwards that can legitimately pass very well for their size. Kansas was the only Big 12 team last year that started a frontcourt where both forwards posted Assist Rates of 15% or higher. The only other team Houston played that had something sort of like it was TCU, whose Emanuel Miller operated in a similar fashion at times.

Adams was able to post 7 assists in a 78-65 home win over the Cougars. Dickinson had four of his own. Together with Miller, they were the only three forwards all season long to post four or more assists in a game against this defense. Exploiting the Houston blitz isn’t easy, but watch how smoothly Adams takes care of it here because of the quick dump-off by freshman Johnny Furphy.

All of this is easier said than done. TCU would meet Houston in the Big 12 Tournament and get dump-trucked en route to a 45-point, 28% 2PT performance. Kansas saw Houston in the season finale and lost by 30, because Kelvin Sampson - arguably the single best program leader in the college game - made the appropriate adjustments. The first possession of that game saw J’Wan Roberts play drop, of all coverages, which neutered Adams because Adams is a non-shooter. On the next P&R, Houston switched Shead onto Dickinson, which would normally result in a post-up but simply resulted in Shead, the most annoying defender in college basketball (complimentary), annoying Dickinson into one of the worst shots of his season.

So, yeah. If you get it once, serious credit to you, because that’s hard enough. Now try and get it a second time. If you don’t have top-tier talent and top-tier coaching (and honestly, a little luck), you likely won’t.

The Texanpresse

Sometimes in life, the comparisons you have to make are extremely easy. Finding a good one for Iowa State was a little difficult; this one I knew from the moment I had the article idea in my head. They even have the same primary red color. Yes, Houston’s European football comparison is obviously Liverpool.

They’re even aware of each other’s alternate universe

More specifically, peak Jurgen Klopp’s Liverpool, which pressed opponents into a frenzy and is the only team to win the English Premier League not named Manchester City since 2017. Like Houston, Liverpool were a dominant figure in the late 1970s and 1980s who began to slip in the 1990s and fully cratered in the early 2010s. From 2009 to 2013, Liverpool went four years without finishing in the top five of England’s top division for the first time in 60 years. During that same time, Houston began their longest streak of sub-.500 play in conference (six seasons) in program history.

Even if the history wasn’t there, the on-field/on-court product comparisons make a lot of sense. Klopp is a guy that oversaw one of the more notable rises in modern European soccer by taking Germany’s Borussia Dortmund to back-to-back Bundesliga titles, as well as a Champions League runner-up bid in 2013. Dortmund have one of the best fanbases in the sport, but in terms of total spend on their roster, they ranked anywhere from 3rd-5th in the Bundesliga depending on the season. Four straight first or second-place finishes changed the course of their history for a long while, as what was previously infrequent bouts of success has now turned into five top-three finishes in the last six years.

Klopp’s Liverpool tenure had the 2019-20 Premier League title, of course, but more notably, a club that finished between 6th and 8th in six of seven seasons prior to his arrival never once finished below fifth in eight full Klopp years. Concluding his run in 2023-24, Liverpool had as many top-two finishes from 2019 to 2024 (three) as they had from 1991 to 2018.

Jimmies and Joes alone, to borrow a basketball term, don’t tell the story. Liverpool’s transfer expenditure during Klopp’s tenure was merely sixth-highest in England. (It was ninth-highest in the world, but they were $1.2 billion off of first-place Chelsea and $600M off Manchester City. It would be like saying LSU ranks 10th in terms of number of five-star basketball recruits in the last decade, which they do.) More than money, it’s about the system: gegenpressing.

The core philosophy of it is simple: when you lose the ball, you don’t retreat into a shell and wait for your turn. You attack and get that ball back by any means necessary.

In the Iowa State article, I referenced PPDA, which is a stat in soccer that functionally measures how high your press is and how hard you’re pressing. Liverpool ranked in the top four of that stat in every single year but one of Klopp’s eight-season tenure. Take that chart I posted last week, showing Real Sociedad #1 across the top five European leagues in it. Liverpool were #2 with a much more powerful offense.

Scour YouTube and you’ll find videos titled things like “two minutes of Klopp’s Liverpool pressing like absolute animals” and you get it pretty quickly. If you’re a fan of the Houston blitz, you’re obviously going to be a fan of what the Gegenpress does.

The basic tenets of the system are reasonably simple to learn. Via Breaking the Lines’s Ten Commandments of Gegenpressing:

Gegenpressing (counter-pressing) is simply winning the ball immediately after losing possession. The opposition’s intention after getting the ball is to start a counter; hence their defensive organisation is broken, leaving them vulnerable because their players are quite apart in the quest to score a goal.

The specifics of gegenpressing require everyone on the field (or yes, pitch) to be on the same page. If Liverpool’s front three can’t get it back, then the midfield’s job is to steal possession. If you get to the final third against Klopp, his fullbacks drive the ball outside, with the intention of running the opposition out of space. But the biggest advantage gegenpressing has, or at least does with teams that run it well, is pure energy. As a fan of Leeds United, only the promoted 2020-21 side contends with peak Liverpool in terms of teams that played with pure energy for 90 minutes. Watch here for what happens when you make even one bad pass:

Klopp describes his style of play beautifully: heavy metal football.

He likes having the ball, playing football, passes. It’s like an orchestra. But it’s a silent song. But I like heavy metal more. I always want it loud.

That’s why I think the comparison makes a lot of sense. Houston plays with relentless energy from start to finish against basically every opponent. Of Houston’s 37 games played last year, an astonishing 23 saw the opponent have a bottom-five day in offensive efficiency across their entire season. That is a lot of good days stacked together for the defense, but more impressive is the variety of opponents and rosters it happened against. All of Kansas, Texas Tech (twice), TCU, Iowa State, Duke, and Dayton had terrible days of their own against this defense.

The bad news for Houston, if it exists, is that we may have a Manchester City figure on a national scale. UConn is so dominant at this time that people are seriously pondering the thought of a three-peat, something that no one has done in over 50 years. In fact, no one has done it in any of the major American sports since the Lakers did it in the early 2000s.

Along with that, the Liverpool run ended in much the same way a lot of gegenpress teams end: they don’t have the horses to do it forever. A 2020-21 title repeat attempt died by way of injuries; a 2021-22 attempt, their last great one, saw them get bested just barely by an even better Man City. Liverpool became a bit too old for their own good, and now after an amazing run, they’re resetting under new manager Arne Slot.

The good news for Houston: your direct comparison stopped the first Manchester City three-peat from happening. Just don’t look at how long it took to complete the 2019-20 season or anything. Please.