What do we really know about the science and development of jump-shooting?
At its core, basketball is a game of points. And in the current era of the NBA, no strain of scoring is more in-style than shooting. As the league is increasingly defined by pace and space, there’s never been more of a premium on shooting the basketball and all of its complexities. Many will argue it’s the most important trait for a modern basketball player to possess.
Teams have never invested more into filling rosters with a versatile arsenal of 3-point snipers and gunners. Therefore, finding good, great and elite shooters through the draft is of the utmost importance.
Shooting and shooting projection is a hotly disputed topic among draft enthusiasts and actual scouting departments, I’m sure. Though all facets of prospect evaluation present challenges, some skills and traits are easier to determine. Some are mostly static — height, length, run-jump athleticism — and some are easily identifiable and projectable statistically — an elite decision-maker is an elite decision-maker, prospects who rack up steals tend to do the same in the NBA.
The same can’t be said for shooting. Highly efficient and impactful college 3-point shooters often can’t reach the same level of impact at the next level. Shooting is inherently variable, especially in the minute samples of many high-level prospects. We know 3-point shooting takes an eternity to stabilize (around 750 attempts), rendering raw 3-point percentage an insufficient indicator. A prospect can strut his elite handle in one play he is but may take multiple seasons to prove what level of shooter.
While evaluation is a key component of draft analysis, the more meaningful piece is player development, which is where difficulties arise. It’s far more challenging to predict how good of a shooter a prospect will be in five years than it is to evaluate the level of shooter he is today.
That’s where this piece comes in — it’s my attempt at breaking down my process in projecting shooting for the NBA draft. It’s an admittedly daunting and multi-faceted task with layers upon layers. It becomes increasingly difficult with highly-touted one and done type prospect, whose statistical sample makes projecting shooting tougher. Almost no prospect will accrue a stable sample of 3-pointers over even four-year careers and underclassmen won’t even come close.
To combat this, we have a swath of historical and statistical data on past prospects, none of which is perfect, but leads us to a group of indicators for shooting projection. This series is not intended to be a rigorous statistical analysis. It’s an overview of these indicators, which ones to weight most and how to identify and use them to predict the development of good and bad shooters.
No two shooting profiles are alike, yet the analysis of historical shot profiles is essential in the projection of future shooters. Understanding why past prospects failed or succeeded is key to future success. To simplify and better understand these indicators, we can study these shooting indicators through the lens of one shooting profile that I find particularly informative: Justise Winslow.
During his lone season at Duke, Winslow’s shooting profile looked like this:
3PT%: 41.8 percent (46-of-110, .309 3-point attempt rate)
2PT%: 51.6 percent (127-of-246)
FT%: 64.1 percent (100-of-156)
Long 2PT%: 17.6 percent (15-of-85)
Close 2PT%: 68.7 percent (112-of-163)
Catch-and-shoot: 62.2 eFG% (39-of-94), 1.245 PPP, 87th percentile
Off-dribble: 13.2 eFG% (2-of-19), 0.263 PPP, 2nd percentile
Off-screens: 20.8 eFG% (2-of-12), 0.538 PPP, 11th percentile
Spot-up no-dribble jumper: 65.4 eFG% (34-of-78), 1.308 PPP, 87th percentile
Runners: 19 eFG% (4-of-21), 0.318 PPP, 10th percentile
Five seasons into his NBA career, he’s shooting 33.7 percent on 3-pointers (190-of-563) and 33.3 percent on long 2-pointers.
At the time of the 2015 NBA Draft, some raved about Winslow’s shooting, likely due to his near 42 percent 3-point clip, his jumper described as a “promising three-point stroke could blossom into all-around shooting mastery” by one evaluator.
Through five seasons, Justise Winslow hasn’t been able to match his college 3-point efficiency. By raw 3-point percentage, though, Winslow has made significant progress over his third and fourth season, stalling out in an injury-riddled fifth. By studying not only Winslow’s college shooting, but his NBA shooting and developmental context, we can learn a lot about the shooting indicators that predicted (or didn’t predict) how Winslow’s shooting would develop and how we can apply those indicators to current evaluations.
Let’s start with the most important indicator of them all: volume.
To begin, here are a few crude yet illuminating statistical queries. Since 2008, here is the list of drafted prospects with over 200 3-point attempts in a season and here is the list with less than 50.
The first list predictably is littered with good to great to elite NBA shooters. Of the prospects whose shooting didn’t translate, most are either first or second-year players, who are often bad at many things, shooting included, and players who were not good enough at the ancillary parts of basketball to make up for their shooting prowess, which is a crucial topic beyond the scope of this piece.
This underscores a simple truth: great shooters shoot.
The next list features mostly non-shooters or, more importantly, non-meaningful shooters. The notable outliers are as follows: Karl-Anthony Towns, Victor Oladipo, Pascal Siakam, Jimmy Butler and the Morris twins.
Before moving onto the discussion of meaningful shooting, I want to reiterate the inherent fickleness of projecting shooting and projecting young players in any way. Prospects, especially young ones, develop in unexpected ways. Though many of these outliers can be explained away by outlier work ethic or some other indicator, the extent to which many players develop as shooters cannot be predicted. Even taking into account Towns’ positive indicators and understanding the limitations of the Kentucky context, predicting his evolution into perhaps the best big man shooter ever wasn’t feasible at the time of his draft.
From a broad perspective, shooting is the easiest ‘skill’ to develop. This statement seems to be agreed on by most. While improving ball-handling or feel seems to be a formidable task, turning a non-shooter into a solid or good one is feasible. Aron Baynes attempted 10 triples between his four college seasons and first five NBA seasons before blossoming into a respectable and then a good shooter in year seven. A strong shooting coach can teach almost anyone to make spot-up t3s.
But this is where the important distinction of meaningful shooting and volume applies. Predicting that a non-shooter in college will develop into a catch-and-shoot specialist is far less valuable than predicting which prospects will become genuinely impactful shooters, the kind that bends defenses.
Putting shooting projection aside, volume is equally as important, if not more important than raw efficiency. Defenses won’t stress about unwilling shooters, even if they drain their limited attempts at an efficient clip. A primary example of this is Jerami Grant, a non-shooter in college who has become a 40 percent 3-point shooter in his sixth NBA season. Despite this looking like a case of unexpected shooting development, understanding Grant’s true impact which is still tied to his low volume, is vital.
Lacking a quick trigger, defenders won’t close out hard to Grant and he’s not willing to pull difficult attempts, lessening his impact as a closeout attacker and an off-ball spacer:
Compare that to the type of gravity an elite shooter can command, with defenders following Towns to open the layup:
This tendency of Grant’s, despite his efficiency, makes him a far less impactful shooter than players who sport far lower percentages. A strong example of this phenomenon is Luka Doncic, who shot an ostensibly frigid 31.8 percent from deep this season, well below the league’s average of 35.5 percent. Despite this number, Doncic’s sheer volume — 13.2 3s per 100 possessions — commands defensive respect, opening up his driving and passing games.
Doncic’s volume in combination with the diversity and difficulty of his shots creates the impactful shooter and offensive engine he is, despite low percentages. Doncic thrives off of difficult stepback jumpers, which are higher value shots than open spot-ups because of the defensive attention they command. Digging deeper, Doncic’s ranked among the best pull-up shooters in the league this past season, placing in the 85th percentile (a 51 effective field goal percentage):
Despite this, his overall efficiency remained low, dragged down some by poor spot-up shooting but by the inherent inefficiency of pull-up shooting. Here’s a basic example: this past season, Jae Crowder placed in the 50th percentile among off-dribble shooters and Aron Baynes placed in the 50th percentile among spot-up shooters. Yet, Crowder’s effective field goal percentage and points per possession (42.7 percent, 0.867) were both far lower than Baynes’ (52.2 percent, 1.043) despite ranking in the same percentile. Simply, pull-up shots are more difficult to make. Looking beyond raw 3-point percentage and even volume is key to determining how impactful of a shooter a player truly is.
This brings us to the second indicator of the piece, shooting versatility, defined simply as all of the different types of shots in his toolbag: spot-ups, off-movement, pull-up, step-backs and each of these can be broken into subcategories. As a general rule of thumb, more versatility is better.
The best NBA shooters boast elite volume and versatility, especially in college (or wherever they played). J.J. Redick added to his otherworldly shooting profile with a bevy of tough attempts:
Returning to Justise Winslow, this is a major area where he failed. His volume (116 attempts) was passable, not high enough to predict elite shooting upside, but not low enough to predict failure. But the value of his 41.8 percent clip on fair volume diminishes when inspecting his shot versatility.
The majority of Winslow’s 3-point attempts and 3-point makes looked were the wide-open catch-and-shoot variety:
Winslow shot 0-of-4 on off-screen 3-point attempts and 1-5 on pull-up 3s during his freshman year at Duke. This stark lack of shot diversity not only limited his future shooting prospects but made him less threatening of an offensive player in college. Defenders routinely slid under picks against Winslow, relieved at the sight of his pull-up:
Through Winslow’s first two NBA seasons, his distance shooting volume, versatility and efficiency was predictably low. But in his third season, his 3-point percentage spiked by 18 percentage points to 38 percent, a strong number. Yet, the same volume and versatility issues made him a shooter opponents didn’t stress about defending, as he took only 1.9 3s per game, with just three total attempts coming off of screens (3-of-3) and six coming off of the dribble with no makes all season.
Winslow saw his shooting efficiency sustain and his volume nearly double in his fourth season, which is undoubtedly a positive point for him. Yet his true shooting impact remained limited. Despite strong catch-and-shoot numbers, Winslow still struggled mightily as a pull-up shooter with increased volume from mid-range and 3, shooting 5-of-32 on off-dribble 3s.
So, despite a positive development arc, Winslow has a long way to go before he’s a truly impactful shooter. Thrust into a primary ball-handler role in his fourth year, Winslow needed to be a more effective shooter than he was to succeed. He looked similar to college Winslow as a spacer, living off of easier spot-ups and receiving the same old under treatment:
To further emphasize the importance of versatility, let’s compare the shooting profiles of two 2019 prospects: Terence Davis and Matisse Thybulle.
From a basic statistical standpoint, there’s an argument for both prospects as better shooting bets. Davis’ volume is superior, but Thybulle wins in career percentage and free-throw percentage with solid volume in his own right. Through one NBA season, the better shooter seems clear:
Based on their college shooting profiles, the gap between Davis’ and Thybulle’s shooting success might seem surprising. But the key differences, volume and, more importantly here, versatility, made the difference.
As a senior in college, Thybulle shot 3-of-15 on off-screen attempts and 7-of-26 off of the dribble (5-of-15 on off-dribble 3s), the vast majority of both his misses and makes coming off of spot-ups, most of his attempts on a low degree of difficulty.
During his senior season, Davis shot 15-of-42 on off-screen attempts and 39-of-105 off of the dribble (18-of-45 on off-dribble 3s), the vast majority of his attempts coming on a high degree of difficulty, spraying NBA-level pull-up and off-movement shots:
Those shooting trends held up in their first NBA seasons, with Thybulle bombing mostly spot-up 3s with little volume or success off-movement (4-of-15) or off of the dribble (1-of-17). This lack of shooting diversity, despite solid percentages, hindered his value as a spacer, especially against great teams who ignored his pull-ups and didn’t stick to him off-ball. Watch Fred VanVleet in the first clip, who sags way off of Thybulle to roam, unafraid of his shooting:
Davis, on the other hand, emerged as a key cog for Toronto because of his shooting, though it’s possible his near 40 percent 3-point shooting on high volume regresses. Davis’ versatility expectedly decreased significantly from his senior season as he shifted into a complementary role (9-of-26 off screens, 11-of-51 off the dribble), his volume and intelligent spacing opened up Toronto’s offense, along with the occasional flash of shot versatility which could expand as he takes on more usage:
It’s important to note that Davis and Thybulle’s shooting could drastically change over the course of their careers. Even so, their comparison is an illustrative one on the importance of shooting versatility. It’s clear 3-point volume (most vitally) and shot versatility in prospects are clear indicators of NBA shooting success.
What does this tell us about the 2020 NBA Draft class?
The logical starting point is LaMelo Ball, the top prospect in the class for many, and one who is often maligned for his NBL shooting — 27.9 percent from 3 this season. There are reasons to doubt LaMelo and his shot, but this is a lazy one. Ball shot 8.5 triples per 40 minutes, most of which were difficult pull-up 3s:
Similar to initiators like Luka Doncic, Ball won’t need to hit an astronomically high clip — just a respectable one if he’s taking tough pull-ups on high volume — to accrue significant gravity. LaMelo’s shooting deserves an entire piece on its own. This one is not that.
Anthony Edwards’ shooting projection is interesting, as he passes the volume test with flying colors (237 attempts, 9.3 3PTA per 40), living off of highly difficult pull-up and step-back jumpers. Prospects with this extreme of volume rarely fail as shooters, but Edwards’ raw 3-point percentage is the lowest of any since 2008 with his volume at 29.1 percent. Without outstanding indicators in any other area, Edwards’ shooting development will be fascinating to track. Given his age and raw volume, I tend to lean positive on his projection, his volume likely scaled-down massively early in his NBA career.
Aleksej Pokusevski is fascinating for many reasons, his shooting projection chief among them. Despite underwhelming efficiency, his volume and versatility are uniquely high, especially for a prospect nearing 7-feet, launching 8.3 triples per 40 minutes. That’s more than most of the best shooting prospects in this class, only training snipers like Isaiah Joe and Aaron Nesmith. He’ll bomb from anywhere on the court with a lightning-quick trigger, firing difficult off-movement and pull-up shots:
For a player of his height and coordination, his gumption is rare. I buy Pokusevski’s shooting upside to a fairly high degree, cementing him as a top tier prospect in this class.
Tyrese Haliburton’s efficiency is off of the charts. He’s likely going to be a very good spot-up shooter, but Haliburton’s hesitancy to take difficult 3s is worrisome. He lacks a quick trigger off of screens (8-of-19 last season) and especially off of the dribble (16-57, 34.2 effective field goal percentage), where he often probes to his demise, passing up open jumpers:
His shooting profile is far more equipped for a complementary role, where Haliburton can hit spot-ups off of others’ gravity. If he’s asked to play as a primary, his lack of volume and versatility will become problematic.
Desmond Bane is the best shooter in the 2020 NBA Draft class. Bane’s toolbox is filled with every shot type imaginable: off-screen left, off-screen right, forward momentum pull-up, negative momentum, step-back, from way deep range and on high volume:
One last word on low-volume bigs, even those with strong indicators in other areas: overvaluing indicators such as touch, free-throw percentage or limited versatility can lead to overrating a big’s shooting potential. Again, I wouldn’t rule out any big becoming a passable spot-up shooter, but it’s vital to consider the level of impact a big man needs to reach as a shooter to have real value. Big men like Tyler Zeller, Dwight Powell and Larry Nance Jr. all had strong indicators but lacked the critical component of volume. I will exercise caution when projecting shooting for bigs like Vernon Carey, Xavier Tillman, Isaiah Stewart and even Tyler Bey, who flashes the occasional off-screen 3:
The next installation of this series will cover free-throws, long 2-pointers, runners and touch and their value as indicators of shooting development. Stay tuned.