UNC basketball: Should the small-ball lineup start against Duke?

After back-to-back road losses against Florida State and Virginia in early January, Roy Williams made the first major lineup change of the season: he swapped freshman big man Garrison Brooks for swingman Cameron Johnson, who had recently returned to full health. UNC rattled off four straight wins after that, including a nail-biter at Notre Dame.

But then the wheels came off with consecutive losses at Virginia Tech, in overtime at home against NC State, and at Clemson. UNC righted the ship a bit with a blowout win against ACC-winless Pittsburgh, but the job doesn’t get any easier with Duke visiting the Dean Dome on Thursday coming off a harsh loss to unranked St. John’s.

With Duke starting Marvin Bagley III and Wendell Carter Jr., both talented players close to seven feet tall, it’s worth asking if it’s time to return to the original starting lineup. Thankfully, although the sample sizes aren’t tremendous, we have plenty of data to draw from: 15 games with a Berry-Williams-Pinson-Maye-Brooks lineup, and 8 with Berry-Williams-Johnson-Pinson-Maye. UNC’s season opener against Northern Iowa, where Jalek Felton started for an injured Joel Berry II, is not included in this data set.

Before we get started: starters are, of course, not everything. Despite starting half as many games as Brooks, Johnson has played more than ten minutes more per game. But starting lineups still tell us a lot about a coach’s game plan, and in this specific case it provides a convenient midpoint in order to analyze. Now, what does this data tell us? I spy five main takeaways.

UNC is taking more threes but not making more threes

UNC’s problems from the perimeter this year have been well-documented. But they’ve been particularly pronounced with the small-ball lineup. After the change, UNC went from 8.1 makes and 20.7 attempts from deep (38.9%), to 8.8 makes and 25.1 attempts (34.8%). So yes, it’s a bit misleading to say that they’re not making more threes, although seven-tenths of a free is a bit abstract. But they’re making far less. Defensively, it’s not much better either. Opponents went from shooting 36.8% from deep to 39.8% after the lineup change.

UNC is playing...slower?

You would think that switching to a smaller lineup would result in UNC pushing the pace a bit more, but that simply hasn’t been the case. UNC’s pace factor, which estimates a team’s possessions per 40 minutes based on their speed of play, dropped from 76.1 to 70.8 after the lineup change. In the first 16 games of the season, UNC had a pace lower than 70.8 only once, when they registered their lowest of the season against Virginia with 58.8.

This hasn’t necessarily translated to worse play. Despite playing slower, UNC actually has a slightly higher points per game average (83.3 to 82.4), largely because their overall shooting percentage from the field has improved slightly as well (47.4% to 45.6%).

UNC is getting to the line way less, but is shooting better from it

After averaging 21.5 free throw attempts per game before the lineup change, UNC has dropped to 16.3 attempts per game since. The lower volume has translated, however, to slightly better results. UNC shot 71.1% from the line pre-change and has shot 76.9% post-change. If more advanced stats are your thing, UNC’s free throw rate (free throw attempts per field goal attempt) dropped from 34% to 25.7%.

UNC is moving the ball much better

Before the change, 57.18% of UNC’s possessions ended with a made shot off of an assist, and 15.63% ended in a turnover. Those numbers changed to 63.53% and 13.64% after Cam Johnson entered the starting lineup. That makes sense—Johnson has the second-lowest turnover percentage on the team with 10% (Luke Maye is at 9.3%), and Garrison Brooks has a turnover percentage of 21.6%. That’s another reason why the offensive output is better even though the pace is slower.

UNC doesn’t need bigs to start in order to rebound well

Offensive rebounding, the bread and butter of UNC’s style of play, has actually improved since the lineup change. Although UNC’s total rebounding numbers have dropped from 41.1 to 38.6 per game, the team is actually pulling down 12.1 offensive rebounds per game as opposed to 11.7 before. Not a big shift, sure, and in fact opposing teams are rebounding their own misses at a rate of 10.3 per game now when it was just 7.1 before.

Those numbers don’t tell the entire story, though. Offensive rebounding percentage, which measures not how many boards you pull in but the percentage of available boards you pull in, has increased more significantly than four-tenths of a rebound. It’s jumped from 35.48% to 38.91%, which is remarkable considering the team went smaller and not bigger.

So who should start?

As tempting as it would be to throw Coach K a curve-ball and switch it up, an old Southern adage rings true especially here: if it ain’t broke, don’t fix it. UNC showed that they weren’t broke by comfortably handing Pitt. That three-game skid was more unlucky than bad. In most cases, the team is playing better with the small-ball lineup than it did with Brooks starting. And with the Blue Devils coming off of a loss, it’s especially important to take them seriously. Go Heels!