The Celtics Need to Go Small Against the Bulls

With Boston coach Brad Stevens considering lineup changes, the numbers say small-ball is the way to go as the Celtics try to save their season.

Like any team down 0-2 in a playoff series, the Boston Celtics are looking for answers to get themselves back in the series.

The key word there is answers as questions abound in the Boston media and locker room.

How can the Celtics keep the Chicago Bulls off the boards? Can anyone prevent Rajon Rondo from controlling the game? Who will step up and score alongside Isaiah Thomas?

Yesterday, Celtics coach Brad Stevens introduced one possible solution -- change the starting lineup. In a media sitting Wednesday, Stevens acknowledged that he may make a change to the team's starting lineup for Game 3 Friday night. Although he didn't specify what that might be, the move they need to make is pretty obvious with one quick look at their lineup production through two games.

The Options

These are the seven lineups that have seen at least five minutes of time together over the course of the Celtics' first two games.

Lineup Minutes Net Rating Off. Rating Def. Rtg
Thomas, Bradley, Crowder, Johnson, Horford 17 2.1 107.7 105.6
Thomas, Smart, Bradley, Olynyk, Horford 13 -5.2 117.8 123.0
Thomas, Smart, Bradley, Crowder, Horford 9 16.2 116.4 100.2
Thomas, Bradley, Crowder, Brown, Horford 7 -74.8 54.9 129.7
Thomas, Bradley, Crowder, Zeller, Horford 7 2.3 159.9 157.7
Thomas, Rozier, Smart, Crowder, Olynyk 6 36.0 113.6 77.6
Thomas, Smart, Bradley, Crowder, Olynyk 5 -74.5 50.5 125.0

Clearly, the two most-used lineups haven't been able to get the job done. The Celtics' primary starters have struggled a little offensively while their second-most-used lineup has experienced issues defending the Bulls' offensive attack.

We can throw out the lineup that would simply replace Amir Johnson with Tyler Zeller because it is only marginally better than the primary one. It shouldn't take me saying this, but we can do the same with the two lineups with a net rating of -74.5 or worse as they aren't the answer.

That leaves us with two small-ball lineups. One made up of Thomas, Marcus Smart, Avery Bradley, Jae Crowder and Al Horford, and the other consisting of Thomas, Terry Rozier, Smart, Crowder and Kelly Olynyk. Essentially, it's Rozier or Bradley, and Horford or Olynyk.

Despite a higher net rating and far better defensive rating, we can't expect the latter of the two to materialize into anything more than a rotational lineup. There are 26.5 million reasons why, which can be found in the hefty contract of Horford. Even if general manager Danny Ainge approved, it's probably unlikely Stevens would let the Celtics' biggest offseason investment sit on the bench to start the game.

In addition, the Bradley-Horford lineup does have a slightly larger sample size and is a more believable (or reasonable) lineup by measure of its production. Not many lineups are consistently anywhere near a 77 defensive rating like the Rozier-Olynyk lineup above.

To summarize, that one's just not happening. So we're left with the first of the two alternatives, which actually makes sense for a few reasons.

The Best Fit

The very first thing that would benefit Boston is this three-guard lineup's pace of play. In their nine minutes together, this group has averaged 108.39 possessions per 48 minutes, according to That's over 7.5 possessions faster than the pace at which their primary starting lineup has played through two games, and it is more than 10 possessions faster than the series-average pace to date (97.25 possessions).

This plays into Boston's hands, because the Bulls are used to playing at the current pace. Their regular season average of 97.72 possessions per game placed them 20th in the NBA. In contrast, the Celtics' average was 99.32 possessions per, ranking 12th in the league.

If that wasn't evidence enough that Boston needs to pick up the pace, take these Celtics' splits into consideration.

Regular SeasonPace
vs CHI95.76
Wins vs CHI97.46
Losses vs CHI94.06

As you can see, there isn't much of a difference in pace between Boston's 53 regular season wins and 29 losses. In their four games against the Bulls, however, their pace dropped an average 3.56 possessions. Further, in their two losses to Chicago, the Celtics' pace dropped to 94.06. In wins over the Bulls, it was a little higher than the average pace of the two teams' current series.

Faster pace? Check.

What about rebounding? That's been arguably the biggest problem for the Celtics.

Through two games, they have secured a playoff-worst 61.4% of available defensive rebounds. If taken over the regular season, that would rank last among all NBA teams by 12.7%.

How a small-ball lineup going to help?

Well, likely to everyone's surprise (including my own), this lineup has grabbed 80% of available defensive rebounds and allowed just 20% of available offensive rebounds to go to the Bulls. That has made for a 50-50 split between the two teams in the nine minutes -- obviously, not a huge sample size -- this lineup has spent on the floor. After being out-sized and so out-rebounded in Games 1 and 2, Stevens will probably take that.

Finally, if there's any question about whether Crowder can do his part to keep Nikola Mirotic from getting hot, we can put that aside. In his three shot attempts against the 6'6" Crowder, the 6'10" Mirotic has come up empty on all three shots, which certainly speaks to Crowder's quickness and ability to closeout to the perimeter, something we've come to expect from him as a very solid defender.

A Much-Needed Change

After saying all that, a desperate Boston team needs to look at the numbers and cut to the chase -- the decision around their new starting lineup is an easy one to make. And if Stevens rolls with it, the complexion of the game (and possibly the series) could change in the Celtics' favor.

Boston needs to do something, because, per our models, the Bulls have a 77.29% chance of winning the series. In fact, we give Chicago a better chance to sweep (31.49%) than we do Boston to come from behind and take the series (22.71%).