There’s Beauty in the Details

By Spencer Pearlman

When my mother was growing up, she trained professionally to join a ballet company. She would spend hours upon hours practicing the tiniest details to make sure every aspect of her performance was perfect — nothing was viewed in isolation.

A tour jeté, though one movement, is composed of multiple building blocks. You start in the correct position, slide backward, kick your leg up high, turn yourself around in the air, and then land on one leg with the other extended in the back.

She would repeat this movement across the studio, breaking down the whole into its component parts. Her teacher would point out areas for improvement in each specific component of the step. For example, the leg could be higher; the turn could be sharper; the arm could be positioned better; or the landing could be softer.

Even if well executed, there was always room for improvement. 

My mom instilled in me that evaluation of the whole is really an analysis of the component parts — to do otherwise betrays the beauty and depth of the whole.

In all aspects of basketball analysis, I keep this wisdom in mind, but it is no more relevant than when examining passing. If the angle of delivery is slightly off, the player receiving the ball might not be able to get a shot off; if the pass is soft rather than zipped through the defense, it might get deflected or picked off; if the wrong pass is used (lob instead of pocket), the ball may never reach the target.

By saying “____ is a great passer” rather than breaking passing down into its sub-skills, we are doing a disservice to the art of passing (and inaccurately assessing the player’s true ability). 

“By saying ____ is a great passer … we are doing a disservice to the art of passing …”

Rather than saying someone is a “good” passer, we must focus on the components that make the passer good — maybe it is the craft, the accuracy, or the variety of passes made. But, if you want to honestly quantify passing with existing metrics, you are at a loss.

Assists neglect perfectly executed passes resulting in missed shots; assists treat both complex and straightforward passes producing makes as equivalent. Meanwhile, turnovers are much more a symptom of usage than a passing metric.

A picked handle or fumbled dribble, though having to do with ball-control, has nothing to do with passing ability. Neither assists nor turnovers, though the closest things to passing metrics the box score offers, provide much insight into passing.

The 2020 Draft presents instructive examples of how the passing whole is distinct from its parts. LaMelo Ball and Tyrese Haliburton were two of the best passers in the draft and Isaac Okoro and Patrick Williams were two of the best among wings. However, the wings had little in common with the ball-handlers in terms of raw box score output. 

2020 Wing AST/TO Ratio

Player AST/TO
Devin Vassell 2.13
Desmond Bane 1.68
Saddiq Bey 1.61
Josh Green 1.59
Aleksej Pokusevski 1.56
Elijah Hughes 1.49
Skylar Mays 1.43
Leandro Bolmaro 1.36
Deni Avdija 1.25
Jahmi'us Ramsey 1.11
Isaac Okoro 1.04
Isaiah Joe 1.00
Robert Woodard .69
Jaden McDaniels .65
Patrick Williams .58
Cassius Stanley .56
Aaron Nesmith .54

Using traditional metrics to draw conclusions on passing ability misses the mark because there is no way to divide passing into its parts. However, by breaking passing down into its sub-skills, we can show that a “good” passer might have holes in his passing repertoire that need to be improved, and a “bad” passer might have legitimate passing acumen. By approaching passing through a more nuanced lens, we better understand each skill in actionable ways. 

Micro-skills of “Good” and “Bad” Bad Box Score Passers

While at Auburn, Isaac Okoro showed legitimate skill as a second-side creator due to his ability to make intelligent passing reads against a moving defense. However, his passing aptitude did not show in the box, as he was barely above a 1:1 AST/TO ratio.

SIS Passing Metrics Ranks

Metric Isaac Okoro Patrick Williams
Complex % 2nd 15th
Pass to Result 1st 16th
Moving Dribble % 7th 4th
Simple Inaccurate % 14th 16th
Complex Inaccurate % 5th 14th
+ Pass Craft 7th 1st

Okoro had real passing acumen in a few areas. He often made complex passes (2nd among wings), quick decisions (3rd among wings), and put zip on the ball to ensure it reached the intended target on time.

However, Okoro also had a genuine shortcoming as a passer: accuracy. Though he ranked 5th in our accuracy metric as a complex passer among wings, he was 14th out of 17 in simple pass accuracy, 10th in defensive manipulation, and 13th in touch passes. He completed complex passes and possessed some passing craft, but he would also whiff on simple ones and lacked craft in other areas. 

A casualty obscured by role, Patrick Williams played almost exclusively off-ball at Florida State and had fewer passing opportunities than Okoro (13th among wings for Williams vs. 6th for Okoro). In the limited chances he was tasked with creating for others — despite a very poor AST/TO ratio — Williams showed valuable passing skills, including those that did not show up in the box score.

Williams was 4th  among wings in our Pass Expected PPP[1], 4th in both Complex Success % and % of Passes Off Dribble, and ranked 1st in our +Pass Craft and Zip metrics.

Williams flashed advanced passing ability but also had legitimate flaws, most notably in Volume of Complex Passes[2] (15th) and overall accuracy (12th, including 16th in Simple and 14th in Complex). From our data, we can conclude that Williams’ main shortcomings were accuracy and pure volume[3]

Williams also had the 2nd lowest Pass to Result rating among wings in our database. This means that his passes often did not play a role in the result of the possession. Despite Williams’s passing craft, he struggled in the only dimension that assists capture: results, the element of passing the passer controls the least. Williams threw valuable passes, but his teammates failed to realize that value in results. 

Even passers who put up gaudy box score numbers are not perfect.

LaMelo Ball’s room for improvements were his accuracy with simple passes (18th of 18 ball-handlers), quick passes (18th of 18), and ability to fit the ball in tight spots (11th of 18).

Tyrese Haliburton’s passing skill gaps were in zip (12th of 18 ball-handlers), accuracy on simple passes (15th of 18), and passes off dribble penetration[4] (15th of 18). Haliburton and Ball routinely threw elite passes and had the box score numbers to support the narrative that they were top-level passers. However, as good as they were, they still had areas for improvement.

Accurately Evaluating Passing

Nothing in basketball is black and white.

You would not describe a good pick-and-roll defender who misses rotations and is prone to losing his man off the ball as a great defender.

You would not regard an elite pick-and-roll defender in a drop scheme who is not a good defender in a soft-hedge or switch-heavy scheme as a terrible pick-and-roll defender.

You would not label a great catch-and-shoot shooter who cannot fire off movement or off the bounce a great shooter overall, but great in a single subset of the entire skill.

Passing is no different.

By recognizing the nuance of the sub-skills within passing, evaluators can provide more accurate representations of the entire skill. In doing so, we can project and develop players precisely. The great catch-and-shoot shooter who cannot shoot off movement is not drafted to sprint off screens or told to work on “shooting,” but is selected to hoist spot-ups and told to work on movement shooting.

Projecting and developing passing must go beyond the rigid world of “good” and “bad.” Players must be considered in contexts and roles as passers. Perhaps a ball-handler is elite at hooking passes to popping big men but cannot execute accurate lobs; the pick-and-roll partner he will be paired with will prove crucial to his projection.

Likewise, a player development coach would be remiss to work on pick-and-roll passing with that ball-handler, instead needing to focus on particular deliveries. Through all phases of player evaluation, projection, and development, we owe it to players to be precise. Only then can we put them in the best positions to succeed. 

While speaking with my mom to get the proper ballet terms needed for this piece[5], she said movement in ballet is never perfect and the ballet dancer is always on that quest for perfection, constantly breaking down each piece of the movement to improve the whole. It’s time for passing evaluation to match that same energy[6].

[1] This indicates that he set his team up with good shots that may not have been capitalized upon by his teammates

[2] Williams showed the ability to complete complex passes, but the volume was not high

[3] Volume can at least be partially attributed to his role at Florida State

[4] Passes from the paint that came off dribble penetration

[5] I know absolutely nothing about ballet except for what a tour jeté is (thanks mom for laying it out in layman’s terms!)

[6] If it doesn’t, I won’t hear the end of it from my mom