Nylon Calculus: Discerning decision-making from 3-point shot type

Shooting is not just about form and accuracy, it’s also about the underlying decision-making — knowing when to shoot and when not to shoot. A literal mountain of research over the past few years has shown how important shot selection is to a player’s shooting percentages.

I wanted to analyze the 3-point shot and see what separated good shooters from bad shooters. From there, I was hoping to ascertain a way for coaches to see if any players are hurting from a game plan or seemingly sabotaging the team.

The 16 players I’ve selected for analysis represent relative extrema. Here are the boundaries used for the eight best 3-point shooters, and the eight worst:

For the 2017-18 season.

The difference in 3PT attempt rates is an attempt to account for the fact that good 3-point shooters know they’re good, and will always attempt to shoot more. On the other end, we’d like to hope that bad 3-point shooters understand they need to be shooting less and opting into the 2-point shot more often. All 16 players had attempt rates greater than 0.413.

The gap in 3-point field goal percentage clearly leaves off some of the awesome 3-point shooters of the NBA, but as stated, I wanted to get to the extreme ends on this data.

Lastly, I observed players who feel like they deserve at least four shots per game. On minutes, I selected only players who played more than 15 minutes per game, because it can tell us a lot about the coaching involved to leave a great shooter, or a very bad one, in the game. The following graph denotes 3-pointers made out of all shot attempts.

Looking at Darius Miller, a third of his total shot attempts are made 3s, which is outstanding. This means he’s a shooter who understands his role on the perimeter and performing in that role. But the chart confirms more about how great the good shooters are as opposed to telling us how the poor shooters are ‘being bad.’

Usually, bad shooters are already shooting themselves in the foot. For example, if you’ve ever seen a guy dribble too much and chuck it once or twice per game and utterly brick, that’s a poor shooter because the ball never had a chance at going in from the beginning. Most good shooters, on the other hand, understand that the catch-and-shoot 3-point shot is really valuable and will only stick to that option when they get the ball.

This graph gives us the amount of catch-and-shoot 3-point attempts each player is shooting out of all his 3-point attempts. You can see that while Jae Crowder isn’t the strongest 3-point shooter, most of his looks come from catch-and-shoot opportunities. On occasion, he’ll toss one up where he’s dribbling or if someone’s tight on him, but he’s pretty much doing his job (albeit not efficiently).

As you could probably already guess, Stephen Curry breaks all expectations on what it means to shoot the 3-point shot.

This graph reiterates the fact that you should expect to have a great 3-point shooter if they were to make their catch-and-shoot opportunities. The bottom axis is a percentage of catch-and-shoot 3s made out of all 3-pointers made per game. In the bottom of the graph, you can make out a pentagon of “meh” in regards to 3-point shooters. These players do make a reasonable amount of catch-and-shoot shots, but there’s something else going on so egregious that it hammers their overall 3-point field goal percentage.

We’re onto something! It’s probably a bad thing to have most of your 3-point attempts come after a couple of dribbles (3-6). In this graph, each player is weighted by their true shooting percentage (each respective grey bubble). You can tell Curry is on another level when it comes to the 3-point shot. He takes a decent amount of off-the-dribble-3s and still sinks them.

It doesn’t look like Crowder is cutting it from deep. At first glance, his overall 3-point shooting percentage looks reasonable, but it’s concerning how low it is considering his decent decision making (or that the Utah Jazz system is giving him plenty of opportunities).

It’s understood the players in the lower right are primarily ball-handlers or point guards, but given today’s spacing, as assumption can be made that point guards would be able to get better shots up. You can see Lonzo Ball’s minuscule grey dot, which makes you really wonder whether he’ll be able to improve at all from deep.

Coaches could look at this last graph and determine which guys are taking too many off-the-dribble 3s and whether their talents are better suited for catch-and-shoot. Players in the upper left catch-and-shoot more than others, while players in the lower right take more dribbles per 3-point shot attempt. Each player’s green dot is sized by their catch-and-shoot percentage.

Take a look at both Brooklyn Nets guards. One could argue there are buckets being left on the floor trying to force Russell and Dinwiddie to off-the-dribble 3s instead of promoting driving or having a game plan set up where they can shoot without dribbling. Seeing that this is happening for two players on the same team suggests the coaching needs to get better next year because Russell and Dinwiddie could be shouldering too much either right before the shot clock, or after plays break down. Joe Harris is already shooting just under five 3-point shots per game, but the team could afford to give both Allen Crabbe and Harris a couple more shots from deep per game.

The Nets were second in the league for attempts but finished twentieth in 3-point field goal percentage. Coach Atkinson already has his work cut out for him, but the game plan definitely needs to change. Russell and Dinwiddie’s true shooting percentages don’t look unfixable here.

For players whose 3-point shot seems absolutely broken like Ball’s or Marcus Smart’s, it’s short-sighted to say they need to drastically improve because it might not happen. As players add more games, coaches need to develop game plans minimizing  the negatives from the floor. Instead of demonizing Ball and Smart on this end, their coaches should work to limit their attempts from deep.

Most data for the 2017-18 season from basketball-reference.com. Dribble then shoot data from NBA.com.