Cumulative UNC +/- Stats through 15 Games

With the non-conference portion of the schedule behind us, it seems like an appropriate time for a cumulative +/- report (sorry for my gross negligence in keeping up with the game-by-game reports on here).

Let’s start with an individual player breakdown:

As a reminder:

Off Eff: the points scored per 100 possessions with a given player on the court
Def Eff: the points allowed per 100 possessions with a given player on the court
Net Eff: Off Eff – Def Eff
Off OnC/OffC: offensive on-court/off-court rating– how many points better (+) or worse (-) per 100 possessions the team is offensively with a given player on the court (as compared to the minutes with him on the bench)
Def OnC/OffC: defensive on-court/off-court rating– how many points better (+) or worse (-) per 100 possessions the team is defensively with a given player on the court
Net OnC/OffC: net on-court/off-court rating– Off OnC/OffC + Def OnC/OffC

So, from an on-court/off-court perspective, Marshall has been UNC’s most valuable (read: irreplaceable) offensive player, Henson its most valuable defensive player, and Zeller its most valuable overall player (followed by Henson and Marshall). Barnes and Bullock round out the top 5 for net on-court/off-court. With Strickland on the floor, UNC has been slightly better on the defensive end, but much worse offensively. With Hairston on the court, the Heels have been much better on the offensive end, but significantly poorer while defending. Say what you will about +/- (I’ll say it: it’s noisy as hell), but, to me, that looks like a pretty accurate reflection of reality after 15 games.

Moving on to the most-used backcourt, wing, and frontcourt combinations:

The starting backcourt of Marshall and Strickland has played nearly half of UNC’s 600 minutes so far. While it’s been a solid combination, the backcourt of Marshall-Bullock has been clearly superior (particularly on the offensive end). Strickland and Bullock has actually been UNC’s best defensive backcourt, although the loss of Marshall causes a huge decline in the team’s offensive efficiency. Lest you start clamoring for more Stilman White minutes, remember that nearly all of his minutes are against the soft underbelly of the schedule and/or after the game has been put out of reach. Still, it’s nice to see that the backcourt of White-Bullock is playing fast and effective basketball in true Roy Williams style.

3PtA Rate is the percentage of the team’s total field goals attempts that are taken from behind the arc with each wing combination on the floor. With the starting wings– Strickland and Barnes– on the court, that number is a staggeringly low 10.1. When Bullock and Hairston are manning the wings, it jumps to an uncharacteristically high (for a Roy Williams team) 36.6– although it’s hard to argue with the 3-point effectiveness (39.8% from deep) or offensive efficiency (123.6) with that wing pair on the court. In the minutes with Bullock and Barnes on the floor, the 3PtA Rate falls more in line with an average Roy Williams team (22.6)– right in between the two extremes discussed above. The Bullock-Barnes wing tandem has also clearly been the best defensively– at least from an on-court efficiency perspective.

The final two rows of the table show the team’s efficiency as a function of number of 3-point shooters on the wing. Two-shooter line-ups include any two of Bullock, Barnes, or Hairston at the wing spots. The single-shooter line-ups only include one of these players at a wing position.. Not surprisingly, the two-shooter line-ups attempt over twice as many 3-pointers. The combination of more attempted 3-pointers (while maintaining an efficient percentage from behind the arc) and better floor spacing leads to a significantly higher offensive efficiency with multi-shooter line-ups on the court.

Henson and Zeller play over half of the team’s frontcourt minutes together, and are the most effective pairing from a net efficiency standpoint. With the starting bigs on the floor, UNC has an efficiency margin of +38.0. With one or both on the bench, that margin drops to +18.2 (with a loss of 7.5 points per 100 possessions on the offensive end and 12.3 per 100 on the defensive end). The %Min. vs. Top 50 column shows the percentage of a frontcourt combination’s minutes that are against top 50 opponents (40% of UNC’s total minutes fall into this category. With fewer than 10% of their minutes against top 50, the strong numbers for Watts-McAdoo and McAdoo-Hubert should obviously be taken with a super-sized grain of salt. It’s perhaps interesting to note that McAdoo-Henson has been dominant on the defensive glass, but very poor on the offensive backboards. Conversely, McAdoo-Zeller line-ups have dominated the offensive boards while struggling to corral defensive rebounds. The starting frontcourt has been solid on the defensive glass, and phenomenal on the offensive backboards (grabbing 46% of the team’s missed shots).

By far, Roy Williams’s most-used 5-man line-ups have been the starting 5 and the “Big 4” + Bullock (replacing Strickland with Bullock alongside the other 4 starters). As seen when comparing these two units in the above table, it’s been a big +/- advantage for the quintet including Bullock. Not only has that unit been vastly superior offensively, it’s also played slightly better defense (although both line-ups have been terrific). While some of that difference is likely due to the starting 5 playing more minutes against opposing (and fresh)  starters, it is also important to note that the Bullock unit has played a higher percentage of minutes (63% vs. 41%) against the top 50, and has been much, much more effective in those top 50 minutes (efficiency margin of +44.5 vs. -8.9 for the starting 5). Again, I’m not necessarily calling for Reggie Bullock to start, just imagining a scenario in which the Bullock + Big 4 minutes (currently averaging 3.3 MPG) are a little closer to the starting 5’s minutes (averaging 13.3 MPG). Maybe it’s just me, but I love that Marshall-Bullock-Barnes-Henson-Zeller line-up (with no disrespect to Dexter Strickland, who’s been terrific in his role this season).

Beyond the top 2 line-ups, Roy’s shown a strong preference for the Strickland-Bullock-Hairston-McAdoo-Henson line-up while Marshall is resting (simultaneously resting Marshall, Barnes, and Zeller to maximize their minutes on the floor together). That unit has done very well on both ends– a tribute to Strickland’s improved point guard play and the depth of UNC’s roster. As seen in the above table, only 5 line-ups have played more than 1 MPG (i.e., 15 total minutes on the season)– and one of those is the human-victory-cigar quintet of White-Bullock-Hairston-McAdoo-Hubert (the unit that tends to close out the pre-Blue Steel minutes during blowouts). On the season, the top 3 most-used line-ups have accounted for 46% of the team’s total minutes. The top 10 line-ups have played 62% of all minutes (with the remaining 38% of Carolina’s minutes being covered by an additional 79 line-ups that range in total minutes from 0.1 to 11.3 on the season). With ACC play upon us, it will be interesting to track how Roy Williams’s rotations evolve, tighten, and ultimately finalize in preparation for the postseason.