A Review of the 2020-2021 Celtics Offense Through the Lens of Advanced Analytics

Hello everybody and welcome to AC Playoff Takeaways, a series where I'll be writing daily articles with a focus on coaching/player strategy, and will be making predictions about player stats, game outcomes, and coaching adjustments along the way! The easiest way to stay updated when we release an article is by following our Twitter. Thanks and hope you enjoy the journey :) 

Note: This article can also be viewed HERE on guybostonsports.com!

A Review of the 2020-2021 Celtics Offense Through the Lens of Advanced Analytics

Personally, I didn’t even think a team constructed by Danny Ainge and coached by Brad Stevens could look like the Celtics did this past season. And no, I’m not talking about their record or their lack of playoff success, and I’m not even talking about their lackluster, lazy defense at times; I’m talking about their offense, if that’s even what it should be called. We played as five individuals throughout the season, which is not something Celtics fans are used to, at least as of late; the Isaiah Thomas led teams always moved the ball with unselfishness and force, and not much needs to be said about how the 2008-era Celtics played together on the offensive end. Even the recent Al Horford and Kemba/Kyrie teams have played with the purpose of making their teammates better and getting everyone involved. This year’s team, though, failed to play together, they failed to play to their strengths, and despite their above average individual talent on the offensive end, they played like a below average offense collectively.


I probably yelled MOVE THE BALL at my TV screen an average of 6 times per game this year. When I think back to this year’s Celtics offense, and even more broadly this year’s Celtics team, the first thing that comes to mind is stale offense, and the advanced analytics certainly serve to prove that point. We had the 4th lowest assist percentage in the NBA, which means a bafflingly small percentage of our buckets came from assists; even worse, we ranked 2nd to last in potential assists, which refers to any time that, if a shot were made, there would have been an assist rewarded. In other words, much of the way we scored was from players having to create their own shot for themselves as opposed to it coming within the flow of the offense. We took the 2nd most pull up field goals per game, while taking the 2nd fewest catch and shoot field goal attempts. Think about that for a second; we’re not getting shots from guys driving the lane, kicking out to shooters, and making extra passes on the perimeter. Instead, we’re shooting off the dribble pull-ups, which clearly have a far lower chance of going in than standstill, catch and shoot 3 pointers. If you’ve watched any of the playoffs, especially this last round and the conference finals going on right now, it’s impossible to miss the thing that each team has in common: they all drive and kick, find shooters on the perimeter, and swing the ball around (quickly) until one of them gets an open look. It’s how Terance Mann made 7 threes in a game, and it’s been the calling card of the Jazz all season. 

I heard an awesome quote from Steve Nash recently during a playoff game while he was wired. He said something like (paraphrasing): “Get that first domino to fall, and then the ball movement will continue on from there”. Basically, as Paul George, or Devin Booker, or Trae Young (or virtually any superstar player on a successful team) starts dancing on a defender on the perimeter, the help side is going to shift towards them. Getting the “first domino to fall” means giving the ball up to a teammate, early in the possession and not after 14 useless dribbles, which allows that teammate to drive the closeout as a result of his defender’s defensive positioning (helping on the star player with the ball). Once the second player gets the ball, he’s able to play with a defender flying out to him, granting him a number of appealing options. Any NBA player’s eyes are lighting up in this situation; if the defender doesn’t close out fast enough, let it fly. If he does, it’s probably because he’s flying out there with reckless abandon, in which case you’re able to up-fake and drive by with ease. It’s literally Dominos from that point onward, because once that second player gets by the closing out defender, he necessitates a lot of help in the painted area (because his defender is out of the picture), which then makes the kick-out even more open than it was initially. Drive, kick, drive, kick, extra pass, extra pass, shot. We did none of it this year. All of the good teams do a lot of it. End of story.

Here is a video of the type of offensive movement and passing I’m referencing

Play To Your… Weaknesses?

The other thing that stood out to me from looking at the advanced analytics was the degree to which we failed to play to our strengths on the offensive end. Despite having the 2nd best score frequency on transition in the entire NBA (score frequency takes into account foul calls that result in free throws), we very infrequently go out on the break and played with pace; we ranked  21st in the league in transition frequency. Conversely, we had the 9th highest isolation frequency (which means we did a lot of isolation offense) while having the 3rd lowest score frequency on those plays. To put it plainly, we are really bad at isolation, but we do it often, and we’re really good in transition, but we don’t do it nearly enough. Every team has to have an identity on both ends; what do you do well? What do teams have to do in order to prepare for that thing that you do well? That thing should have been transition; we had a very athletic lineup that should have been able to take advantage of their speed on the fast break at a far higher clip. JT and JB are both lethal on the fast break, and Kemba and Smart aren’t so bad themselves; each of these guys has the ball handling ability to be able to lead the break, but also the shooting ability to space the floor once we get in the half court or semi-transition. Most important though, I would argue, we had an advantage in the front court. For most of the season, we rotated between 3 big men - Theis, Thompson, Williams - and each of them is above average (for a big-man) at running the floor. Furthermore, those guys would each play fewer minutes than other big-men (because there were 3 of them), which allowed them to be fresh and ready to go while they were on the floor. Although it wasn’t talked about much at all, I really liked this big-man rotation PRECISELY BECAUSE they were each able to give everything they had for whatever minutes they got, which can be illustrated best by the fact that we were successful in transition offense all season. Remember, it’s not just about the big-man scoring the ball; running from rim to rim forces the defense to converge on the big-man in the paint (often with multiple guys because his normal defender is lagging behind), which allows the shooters on the perimeter to be more open than they normally would. Instead of sticking with this big-man rotation that was quite effective (in my opinion), we decided to give away Theis, therefore limiting our ability to play 3 solid, consistent big-men (I love you Luke Kornet and Tacko, but ya’ll just aren’t there yet). Classic 2020-2021 Celtics… playing to our weaknesses.

Check out our Playoff Hot Takes and keep track of how we're doing!

Follow our Twitter for updates on when articles are released!