Three Kings

If you're a Celtics fan, you've likely spent some time over the past three seasons thinking about three pointers. Certainly, the Celtics under coach Brad Stevens were not shy about shooting three's, and you can go all the way back to the Jim O'Brien teams in the early 2000's as proof that the Celtics have at multiple points throughout history been a team that is more than comfortable with an effusive number of three-point shots. But in the three seasons under coach Joe Mazzulla, it has really kicked up a notch. This season, that effusive three-point usage has been met with consternation and criticism. So I sat down to see if there were any trends, to see if in fact, there was a correct number of three pointers.

The research was focused entirely on the Celtics. I looked at two different stats over the past three years – raw number of three pointers taken per game and the percentage of field-goal attempts that were three pointers. I looked at each of these numbers for each game of the Mazzulla era, including postseason games, and charted whether the Celtics won or lost each game. Let's do the data first, and then we can talk about what, if anything, we can learn from this data.

OK, that's a lot of data. Let's consolidate it a little:

And one more time. Enhance:

The first thing that I take away from this data is that the already prolific three-point shooting has become even more prolific this season. But unfortunately for the detractors, there isn't a ton of proof here that the Celtics should start shooting fewer three pointers. While the sample is smaller on the margins, even after three seasons and nearly 250 games (242 to be precise), the Celtics have a better winning percentage when they bomb away 50+ times per game than they do 34 or fewer times.

The third thing I take away is that the Celtics are just really good, and kind of malleable. A game with 35 three's taken is much different than one with 49 taken, but either way, the C's have won most of those games.

I didn't want to just leave it at the raw totals though, because another subject of discussion has been how many three's they take in relation to their total shots taken. There have been plenty of games where Jayson Tatum or Derrick White or Jaylen Brown have been ice cold from three but were doing their thing inside the arc. Just take it to the hoop more, right? Well, let's see those percentages. For these, I just did the summaries, since there are so many more permutations.

And let's enhance one more time:

If these breakdowns look very similar, well that's because they are. In fact, they are a little more exacerbated. The C's have done even better when 55% or more of their shots are three pointers, and that has been true in all five of the season/postseason buckets. This season, they're getting to those percentages more often – after getting to 55+% 25 times from the 22-23 regular season to last postseason, they've already done so 19 times this season. And in those 19 games, they're 14-5, for a .736 winning percentage that equates to a 60-win season. That's not as good as the .795 winning percentage there as a whole, but either way, it's pretty great, and significantly better than when they take more two's.

Now, is this research infallible? Not necessarily. I could compare the Celtics' opponents to see if there were trends in terms of how many more or less the C's take than their opponents in a given game. I could look at league trends. I could look at the quality of shots or other advanced statistics to see the numbers behind the numbers.

Perhaps doing such research would shed more light on the edges here, identify specific situations where taking an abundance of three's (or not enough three's) is a problem. But based on this initial research, I don't think all that extra work would really be worth the trouble. The Celtics have found a formula that works really well, and they likely have a lot more statistics and information at their disposal that tells them they are doing the right thing. The only danger to me would be if the C's become too predictable, and thus easier for defenses to react to, but as we can see in the data, the C's can win at a high level at a very wide range of three's taken. The other 29 teams would gladly take even the worst winning percentages in the summary tables above. So instead of counting the three's and worrying that they're taking too many, I'm going to do my darndest to just sit back and enjoy the ride, because until further notice, the C's are the three kings.