Finding the Celtics’ best lineups using Grant Williams

Brad Stevens loves to tinker with lineups early in the season. At the same time, he’s known to leave certain groups in for too long while they get shelled by, say, a Buddy Hield-type player. In Sacramento, the Celtics started off slow as usual, but really spun out of control with a Jayson Tatum-plus-bench lineup. Of note, Grant Williams was not in that lineup, nor was he in any others. After the game, coach Stevens said that he was still searching for the team’s best lineups, and that meant different opportunities would come for players at different times.

This makes sense for guys at the end of bench. There’s no better time early regular season games to figure out if Javonte Green or Semi Ojeleye are a better fit for Boston’s bench. Shuffling the centers around early also made it pretty clear that Daniel Theis and Robert Williams are better candidates than Enes Kanter and Vincent Poirier to play relevant minutes as a big. Grant Williams, wise beyond his years, has been overwhelmingly positive as a contributor as well, which is why I couldn’t wrap my head around his DNP in Sacramento. I can’t even think of another time when Brad benched one of Boston’s eight best players with little or no explanation.

What’s more confusing is that it doesn’t fall in line with Brad’s hockey-substitution solution of going deep into the bench when the team is struggling. Carsen Edwards, Javonte Green, Semi Ojeleye, Brad Wanamaker, and Enes Kanter all saw early minutes against the Kings, when the Celtics were rightfully plowed into the dirt by Buddy Hield and Nemanja Bjelica. Not every deep-bench stint is going to break that way, but you know what also might not collapse so quickly? A regular rotation to - at the very least - tread water against Hield, who played the entire first quarter.

Fast forward to the Suns game on Monday, where we saw Tatum subbing out early again (to foul trouble this time) and coming back in with an otherwise all-bench unit (Ojeleye, Rob Williams, Wanamaker, Edwards). Boston took a lead, lost it, took it back, and lost it again. It was a blessing to be in the game after a horrid start, but not encouraging to see history repeat itself. The details I took away from the first quarter were:

-Slow start

-Enes Kanter first off the bench

-Tatum foul trouble

After trading blows in the second quarter, the other starters (Smart, Brown, Walker, Theis) came in as Tatum subbed out for Grant Williams. Immediately, the game was flipped on it’s head.

Rick Scuteri-USA TODAY Sports

Boston trailed 34-36 when Grant subbed in with 6:25 left in the second quarter. At halftime, the Celtics had a 57-42 lead. You could rightfully call it a small sample size, but the pattern is undeniable. On a fundamental level, the Celtics play smart basketball with Grant in the game. And on a more spiritual level, the team is energized with him on the court.

One thing to note is that Grant plays a lot of minutes with starters. Boston’s most-used lineup with Williams groups him with Brown, Edwards, Smart and Tatum, who’ve played 13:03 together with a point differential of +42.9 per 100 possessions per Basketball Reference. This also happens to be Boston’s seventh most-used lineup.

The second-most used lineup with Grant includes Brown, Tatum, Walker, and Hayward. They’re +83 in 6:41 together - by far the team’s most efficient grouping.

The third and final five-man lineup with Grant (that has logged more than five minutes) includes Hayward, Walker, Smart, and Theis, and is +42.1 in 5:45 together. Judging by point differential, three of Boston’s four best lineups include Grant Williams.

So, what do we take away from this? That the starters are playing so well that you can sub any individual player out for another and it’s no problem? Or do we start to look at Grant as a starting-caliber player already? Is the sample size of 13 games too small to mean anything?

The Celtics are very top heavy outside of their seven or eight best players. So, yes, I do think you can add almost any bench player to the starting lineup and they’ll have no issue. It makes sense that Grant, the most NBA-ready rookie on the team, benefits the most in these lineups because of his effectiveness outside of scoring.

This also explains why Kanter, known only as a scorer, was a poor fit to be first off the bench against Phoenix. Those lineups aren’t going to feature Kanter on offense, and he’s not going to anchor them on defense.

Therefore, Celtics’ best groups unsurprisingly consist of whoever can coexist with the team’s core players. This sounds strikingly obvious, but it hasn’t always been the case. Boston’s core was mostly formed by draft picks, but has since been refined by the addition-by-subtraction formula.

Since the traditional center position is weak, that makes both Grant and Robert Williams seamless additions to those groups, which is why some of their numbers jump off the page. Maybe not because they’re that good, but because their skill set gels with Kemba and the Jays, for example, instead of overlapping with or impeding their play style.

I don’t look at Grant as a potential starter on this team, but only in the same way that I never saw Smart as one either. One reason is that glue guys can enter the game at any time and make an impact.

Maybe it’s too small of a sample size to conclude anything. Personally, I hate writing off data as irrelevant due to sample size (even though it’s often a sensible thing to do). I constantly have to remind myself that I called Evan Turner “Evan Turnover” before shamelessly endorsing him as a legitimate sixth man of the year candidate as soon as he turned it around. Maybe, as far fetched as it seems, coach Stevens can work that same magic on Kanter. I mean, I won’t be the one to predict that, but I gave it a moment’s thought regardless.

The Celtics might be too dependent on their starters, but Grant’s development could bring them one step closer to some stronger bench lineups.