Screen assists credit the play, not the screener
The screen assist was invented to solve a real injustice. For a century the box score ignored the player who freed a shooter with a hard pick, crediting only the passer and the scorer while the screener — usually a big man doing thankless, physical work — got nothing. So the league began tracking screen assists: when a player sets a screen that directly leads to a made basket, he is credited, and the points that follow are logged as his screen-assist points. It feels like justice, a number that finally sees the dirty work. But look closely at the trigger and the injustice quietly reappears in a new form. A screen assist only exists when the shot goes in, and whether the shot goes in has far more to do with the shooter and the play design than with the quality of the pick. The stat was built to credit the screener. In practice it credits the make.
What a screen assist actually records
A screen assist is awarded when a player sets a screen for a teammate who then scores, either immediately off the screen or after a defined number of dribbles. The record is a made basket that followed a screen — nothing more. It does not know whether the screen was the reason the shot was open or merely incidental to it. It does not know whether the screener made hard, legal contact that genuinely dislodged a defender, or simply stood near one while a great shooter rose up over the top. And crucially, it never fires when the same screen produces the same open look and the shooter misses. The screen can be identical on two possessions; only the one that ends in a made shot lands on the screener’s ledger. The stat measures the outcome of the possession, not the contribution of the pick.
The shooter controls the trigger
Because a screen assist requires a make, the screener’s total is hostage to who he screens for. Set picks all night for an elite shooter who converts open looks at a high clip and the screen-assist points pile up; set the exact same picks for a streaky shooter and the number craters, even though the screener did his job identically. This is the tell of a stat that has misassigned its credit: the value on the screener’s line moves with someone else’s shooting percentage. A big man on a team full of knockdown snipers will look like an elite screen-setter by the numbers, while an equally good screener on a poor-shooting team looks ordinary. The pick did not change. The finishing did, and the finishing belongs to the shooter. There is a further wrinkle: the screen-assist points a player is credited with are worth two or three depending on where the shot came from, so a big who screens for a team of three-point shooters accrues more screen-assist points per identical pick than one whose guards finish at the rim. The screener’s counting line therefore rides on his teammates’ shot selection as well as their accuracy, stacking one context effect on top of another.
Volume comes from the play-call
The number of screens a player even gets to set is a coaching decision. A center in a heavy pick-and-roll or dribble-handoff offense will be sent into dozens of screening actions a game; a big who plays in a system built around isolation or post-ups will set a fraction as many. Screen-assist totals therefore rank offensive systems as much as they rank screeners. A player’s league standing in the stat can swing dramatically with a change of coach or scheme, without his screening skill moving at all. When the raw total depends this heavily on how often the play-call asks a player to screen, reading it as an individual skill ranking confuses the man with the offense he happens to run inside.
It ignores the best screens of all
The deepest flaw is what the stat cannot see. The most valuable screening in modern basketball often produces no shot for the screener’s man at all — it forces a switch that creates a mismatch elsewhere, or draws two defenders to the ball and springs a teammate on the weak side three passes later. A screen that generates a wide-open corner three for a different player, off a scramble the pick started, yields no screen assist for the screener. Neither does the screen so respected that the defense blitzes it and surrenders a 4-on-3 the offense exploits downstream. The stat only rewards the most direct, immediate chain: screen, catch, make. It is blind to the gravity and the chaos that good screening creates two or three actions later, which is where a lot of its real value lives.
Where the number is genuinely useful
None of this means screen assists are noise. As a descriptor of role and involvement, they are excellent: a high total tells you a player is central to his team’s screening game, a hub of its pick-and-roll and handoff actions, and that context matters before you evaluate anything else. Tracked over a full season, screen-assist volume is a fair fingerprint of how a big is being used, and paired with the right context it hints at whether a player is a screening focal point or a bystander. The information is real. The error is only in reading the points total as a measure of screen quality rather than a measure of how often a player screened and how well his teammates shot afterward.
How to read it honestly
Treat screen-assist volume as the useful part and screen-assist points as the misleading part. The count of screen assists tells you involvement; the points attached to them are contaminated by shooter quality and should never be used to rank screeners across teams. To actually judge screening, go to the film and to the tracking data that captures the pick itself — the separation a screen creates, the defender contact, how often a screen forces a switch or a blitz, and the team’s offensive efficiency in the actions a given player screens in. A screener who consistently generates open looks and forces coverage breakdowns is doing the job whether or not his shooters happen to make the shots. That distinction is exactly what the points total erases.
The honest read
Screen assists were a genuine attempt to reward invisible labor, and as a measure of how involved a player is in the screening game they do useful work. But the stat fires only on made shots, which hands the screener credit for his shooters’ accuracy, scales with a coach’s play-calling, and ignores the switch-forcing, gravity-bending screens that never produce a direct basket. Read screen-assist volume as a role marker and go to the tracking data and the tape to judge the picks themselves. The screener set the same screen whether the shot dropped or rimmed out. Only one of those possessions rewards him — which is the surest sign the number is measuring the make, not the man setting the pick.