Roster Analysis: The Bad
A few house rules for this article. “Like” and “Dislike” are relative to what I expect the market to think of a particular team. Self-imposed rule for this exercise is that the team needs to be in the current Torvik top 50. I thought about starting with the “Builds I Like” article (I’ll get there eventually), but there’s plenty of content out there at the moment if you want to read about sunshine and rainbows.
Two big recommendations for roster tracking: Jim Root’s roster sheet (which he kindly makes available to everyone for a measly $20) and the BurnerBall depth charts in Trilly Donovan’s Discord.
Builds I Dislike
The Kyan Evans addition has drawn universal acclaim and he’s popping in most analytical models. I don’t see it. There’s a more-than-plausible chance that he’s a Medved Merchant whose numbers were buoyed by the efficiency of Nico’s pristine Motion system. At UNC, he’ll be thrust into a new role as the primary ball handler in a system extremely reliant upon dribble creation out of P & R — something he wasn’t regularly asked to do at CSU — and where most of his shots were created within the natural movement and flow of the offense. In many respects, this is analogous to last year’s Indiana State duo (Julian Larry and Jayson Kent) departing the comfy confines of Schertz’ scheme for a wonky, dribble-reliant system at Texas. There’s very little margin for error here considering his size limitations and brutal TO rate (21% in a Medved offense is hard to do). If his TO rate further spikes — as is often the case when guards transfer up to a power conference — he could become borderline unplayable.
Notably, 58% of Evans’ made FGs last season (an exceptionally high number for a ballhandler) were assisted - a byproduct of the Medved system and playing with a dynamic wing passer in Nique Clifford. He’s not a “get your own” type of guard, which is an issue for a team lacking such a player in the backcourt (barring a late-career breakout from Seth Trimble or Drake Powell withdrawing from the Draft).
For purposes of projecting playing time and role allocation, I always try to identify a “core” 3-4 player group for a team and then fill in the lower-usage role guys around the core. Here, UNC’s four best players will likely be — in some order — Trimble, Caleb Wilson, Jarin Stevenson, and Henri Veesaar, but nothing I’ve read suggests that Wilson can play the 3 at this point in his career, meaning the Heels can’t play their four best players together. Further, an Evans/Trimble backcourt is extremely light on playmaking, as Trimble has never been asked in his career to create for others while playing alongside RJ and Cadeau. It’s too late now, but a tweener type at the 3/4 spots through whom the offense could run (someone like AK Okereke or LeJuan Watts) would have been a savvy add here to alleviate the creation burden on Evans and Trimble.
UNC should be strong on the glass and at the rim defensively with all of this added size (last year’s team was notoriously small), but Hubert (as coaches often do in the Portal Era) overcorrected last year’s problem and now has an odd lineup construction situation on his hands. Such is the predictable outcome when you don’t have a real coaching identity/scheme, as the Portal becomes a random grab bag where you bring in miscellaneous players because you can rather than because you should. Given a do-over and the benefit of hindsight, I wonder if he would punt on Evans in favor of a more dynamic shot creator/passer on the ball better-suited to pressure the rim. The Heels have decent off-ball shooting and should be fine scoring in transition, but there’s a big roster hole in figuring out who will drive the bus offensively in the halfcourt. Two possible wild cards: Powell withdrawing from the Draft, or Hubert making a late (desperate?) Portal plunge for one of the few remaining big fish. We’ll see.
Quick: name last season’s best players nationally in HoopExplorer’s offensive RAPM (the best all-in-one stat on the market).
Ready? Johni Broome, Trey Kaufmann-Renn, Caleb Grill, Walter Clayton, Kon Knueppel, Cooper Flagg, J.T. Toppin…wait. CALEB GRILL? Yes. That Caleb Grill. What a stud.
Grill’s +8.3 offensive RAPM (HoopExplorer) was an extraordinary late-career performance for an erstwhile-mercurial player. More relevant for next season’s purposes was his effect on Anthony Robinson. Yes, Robinson had the higher usage rate and was presumed to be the “play driver” because he so frequently had the ball in his hands, but Robinson’s numbers dropped off a cliff without Grill, plummeting from a stratospheric +10.1 offensive RAPM with Grill to +3.8 without him. Robinson’s assist rate cratered from 30.6% to 19.7% and his turnovers ticked slightly upwards. Yes, Grill likewise greatly benefited from Robinson’s fantastic rim pressure and passing ability, but what becomes of Robinson when he no longer has one of the ten best offensive players in the country playing alongside him?
Fortunately, we might already know the answer. Dennis Gates, who vacillates between moments of Portal genius and stupefying dumbfuckery, opted to punish his own star point guard by saddling him with Sebastian Mack, a poor man’s version of…Anthony Robinson. So now, Mizzou has two bowling ball guards with no catch-and-shoot ability. Yes, both are stellar at drawing fouls (these SEC games are going to take three hours again), but you still need some degree of gravity to keep the floor from being condensed. Perhaps Mizzou could still get away with this alignment by supplementing it with enough shooting from the wings, but…there isn’t. Mark Mitchell is still here, dominating usage in the paint as a non-shooting (but highly-effective) 4 man. But, like Robinson, he was less effective without Grill on the floor (+7.1 ORAPM to +4.5; rim FG% dropped by 10 points).
Jacob Crews is a good 3PT volume/spacer guy but is often unplayable defensively and likely overshot his ceiling going from the OVC to the SEC. He’s unlikely to see much time with a possible Trent Pierce mini-breakout looming, though again, how much can Gates get away with playing the 6-10 Pierce at the 3 defensively in a league like the SEC? There’s a mishmash of cromulent 4/5 types on this roster, but how many of them can play regularly when Mitchell is eating up 70-75% of available minutes at one spot? As is typical with Gates, he’s starting the season with 12-13 players who “could” play, but in this particular build, his “core three” players (Robinson, Mack, Mitchell) look a little bit 2024-25 Kansas-like in that the core three are non-shooters (by volume, not percentage), which severely limits the players you can put around them at the other spots. Kansas fans learned this lesson the hard way last year when they celebrate the arrival of Rylan Griffen as a panacea for the 3PT shooting void, only to later realize that the core three already accounted for approximately 115-of-200 available minutes, leaving the remainder of the rotation fighting for playing time scraps.
Last year’s Mizzou team finished second in FT rate and sixth in Haslametrics’ “quick points off of TOs” metric. This year’s version is going to need to fully replicate that formula, because the halfcourt offense looks a few notches worse.
There are some teams for whom JSU transfer Jaron Pierre would have been a tremendous addition. Unfortunately, SMU is not one of them. Pierre is an uber high-volume shot maker whose arrival further complicates the already-strained Boopie Miller/B.J Edwards backcourt pairing by introducing a third ball-dominant guard into the fray. Per HoopExplorer, both Miller and Edwards were better when the other was off the court — not a great indicator for a starting backcourt! Boopie, in particular, was markedly hurt by Edwards’ presence (+5.1 ORAPM without Edwards; +2.8 with him). These are two downhill, wired-to-score guards who need impeccable spacing and shooting around them to be fully-optimized.
Conversely, Pierre was quite comfortable with his JSU backcourt mate, Quel’Ron House, who fit the mold of the traditional, table-setting point guard (House is now at SIU). To wit, when Pierre and House were on the floor together, Pierre posted a +4.1 offensive RAPM; without him, it dipped to +0.9, his eFG% dropped by nearly 9%, and his TO rate spiked by 5%. In short, House was a security blanket whose low-usage steadiness on the ball allowed Pierre to thrive as an ultra high-volume creator, primarily working in ISO/off the bounce. Do not be misled into thinking that Pierre’s 12.2 3PA/100 possessions qualifies him as a floor spacer. Of the 100+ players nationally with 200+ 3PA, Pierre had the 4th-lowest rate of assisted attempts, trailing only (barely) Barrington Hargress, Braden Smith, and Kam Jones. This is as “ON-BALL” as “ON-BALL” gets.
So, let’s game this out. There are now three ball-dominant guards. A problem, but perhaps able to be mitigated with sufficient shot gravity elsewhere. Do we have said gravity here? No, no we do not. Whereas Matt Cross was an awesome floor-spacer, the new 4-man, Corey Washington (Wichita), arrives from a team where his skills were poorly-utilized because he played with three downhill guards and zero consistent catch-and-shoot threats. Washington is hyper-athletic and a real mismatch for slow-footed 4s, but he’s a bad finisher (75% of his shots were in the paint, but he shot below 52% at the rim, which is horrible for a 4-man) and needs room to operate. His new frontcourt mate, Samet Yigitoglu, is a grown-ass man with legit low-block scoring capability and some skill as a roller, but he’s often going to be forgotten in this my-turn, your-turn offense. Both Washington and Yigitoglu are non-existent passers, so it’s not as if they can run any big-to-big stuff to diversify the offense. Washington was one of only three players nationally with a 25%+ usage rate and sub-5% assist rate. He’s a total void once he gets the ball. Enfield added Alabama/Michigan transfer Sam Walters, a low-usage, stretch 4 type who offers the only hope of adding a semblance of gravity to this offense but at the expense of contributing little else.
There’s some solid-looking true frosh and redshirt guys lurking here, but they’re stuck behind the chuckers in front of them and unlikely to find many available minutes. I really liked last year’s team and thought Enfield did an admirable job constructing a borderline Tourney team from scratch in his first year, but this year’s build is really fraught with danger and I’m hard-pressed to get anywhere near the current #40 Torvik projection.