What lineups are working for the Celtics?

NBA coaches don’t have much time to reflect on the past for the majority of the NBA regular season. The quick turnaround between games, injuries, travel and the vast array of styles between opponents leaves little time for some internal evaluation that could to be valuable for a team when it comes to maximizing their parts and production.

“I think it’s a challenge just generally with our group,” Brad Stevens said of assessing this group before the All-Star Break. “I don’t think we’ve had enough time to truly evaluate the group as a whole. But we feel good about our team when we haven’t had everybody and we feel better about our team when we have. We look forward to that day, we hope that it’s sooner rather than later.”

The All-Star Break provides a valuable opportunity on this front for the Celtics coaching staff and the timing couldn’t be much better with this group expecting to receive a clean bill of health across the roster for the first time in over three months (assuming Rob Williams’ has no setbacks in the next week). A full roster will lead to tough decisions for the coaching staff as he attempts to maximize the performance of his core with supporting pieces that either lack experience (I.e. rookies) or have glaring holes in a part of their games (ex: Enes Kanter - defense, Rob Williams - shooting)

“You’re always taking an assessment of your team – where you think you are and where you think you can go. But we’ll go deeper dive into where we stand after this game tonight,” Stevens said last week before the win over the Clippers win.

With so little data for the team’s five-man lineups due to injuries (only three have played over 100 minutes), we decided to scale back a bit to find some of the three-man combinations that have and have not been working over the first 54 games of the season. What kind of things did Stevens see in the first four months of the season? What combinations might we see more and less of as each game starts to matter a little bit more in a tight East race? A closer look at some key data that tells us what the best version of this Celtics team is.

Rather than going through some of the obvious choices (The Celtics are playing well when Kemba/Tatum/Hayward are on the floor together), we will be looking closely to find success (and failure) among some secondary units that have seen plenty of action all year long.

WINNING THREE-MAN COMBINATIONS

Lockdown D
Hayward/Smart/Tatum: Best defensive rating (94.1) among lineups that have played 100+ minutes
Analysis: These are arguably the best three-wing defenders on the Celtics and they are the most successful trio when it comes to locking down opposing offenses consistently. Smart’s presence is no surprise here and Hayward’s always been a heady/solid team defender. Tatum’s presence on this list, however, is a sign of how big of a leap he’s made defensively this year. His length/versatility allows a good Celtics defense to become great when he’s on the floor.

Grant Williams looks like the ideal role player
Walker/Tatum/G Williams: 125.4 offensive rating in 192 minutes (Best among lineups with 100+ minutes)
Tatum/G Williams/Kanter: 27.5 net rating in 108 minutes (Best net rating on team for lineups with 100+ minutes)
Analysis: We all know how good Tatum has been all year long so it’s no surprise to see him pop up on these lineups but it’s been jarring to see how many lineups with positive success have had Williams in them as a valuable fourth or fifth player in those groups. It’s no coincidence the C’s success has increased in the past month (11-2 in last 13) as he has seen a bigger chunk of consistent minutes. He’s thriving more as a power forward than center but his minutes should feel safe as a good complementary piece for lineups with Tatum (screen setter/not looking for own shot).

Rob Williams intrigue
Hayward/Smart/Rob Williams: 43.1 net rating in 31 minutes (72.2 defensive rating)
Analysis: This is the definition of small sample size theatre but it still explains why the Celtics braintrust is so eager to see the second-year center back in action. Williams was an overall flawed player in his limited action this year but when he’s at his best, he provides a higher ceiling for this group than perhaps anyone else at center with his rim running, shot-blocking and athleticism. His minutes will be limited as he gets back into shape but putting him on the floor with strong heady defenders like Hayward/Smart that allow him to simply patrol the paint could produce more great results.

3-man lineups that have worked well since C’s turnaround (last 13 games)

Wanamaker/Brown/Tatum: 41.4 net rating in 49 minutes
Analysis: A heady point guard in Wanamaker who can get the ball moving on offense and put these scorers in good spots.

Smart/Tatum/Theis: 78.6 defensive rating in 78 minutes
Analysis: Hard to find a stingier defensive trio than this one at all three positions.

Smart/Hayward/G Williams: 29.2 net rating in 64 minutes
Analysis: Three of the smartest defenders on the roster.

THREE-MAN/TWO-MAN LINEUPS THAT HAVE STRUGGLED

Most trios with Semi Ojeleye
Analysis: It’s been a solid season for the third-year forward with career-best shooting from 3-point range but that hasn’t stopped the vast majority of Boston’s lineups to flatline offensively when he’s on the floor. That also explains why his spot in the rotation has been so inconsistent all year long when Stevens has a healthy roster. The data says the Celtics are better when Ojeleye is off the floor by and large. Outside of a few exceptions (defending Giannis or other strong bigs), his minutes could diminish more with everyone healthy.

Hayward/Kanter: 3.7 net rating
Analysis: This isn’t a terrible mark but its still far below the C’s season average. Among key rotation players, the chemistry between the duo just has not worked much yet, particularly in the pick-and-roll where Hayward/Theis is a much more compatible duo. I’d expect Stevens to try to keep these two separated a lot with their two units.

Walker/Smart/G Williams: -1.1 net rating
Analysis: Just doesn’t appear to be much chemistry in this group, especially defensively with Walker on the floor Surprisingly, this is the worst three-man combo among lineups that have played over 150 minutes and they’ve been getting killed on the glass with their lack of size.

All bench lineups
Analysis: This is an obvious one but it’s worth noting that these should disappear for good despite Brad Stevens’ propensity over the years to play four or sometimes five-man bench units together. There just isn’t enough talent on either end of the floor to roll out Smart/Wanamaker/Wing X/G Williams/Kanter groups out there together and the data reflects as such. There should be two starters on the floor at all times for this team in any meaningful game moving forward and that should become far easier with this group getting healthier.