Shot-creating actions bundle unlike skills
Shot-creating actions arrived as a fix for a real problem. Assists only reward the final pass, and only when a teammate scores, so they miss most of the work that builds a chance. SCA widened the lens: it counts the two offensive actions immediately before any shot — the pass, the dribble, the drawn foul, whatever moved the ball into a shooting position — whether or not the shot goes in. As a way of catching creative contribution that the assist column throws away, it is a genuine step forward, and it deserves a place in any modern player profile. But the tidy per-90 that results hides a design decision worth arguing with. SCA adds together actions that require completely different skills, credits two players for every shot regardless of who did the hard part, and inherits the noise of whether shots happened to follow. The single number is cleaner than the thing it measures.
Different actions, one column
The core problem is additive. A shot-creating action can be a defense-splitting through ball, a mazy dribble past two defenders, a drawn foul in a dangerous area, a successful shot that rebounds to a teammate, or a simple lay-off to a runner. These are not the same talent. Vision and weight of pass have almost nothing in common with the balance and acceleration behind a dribble, which in turn have nothing to do with the body positioning that wins a foul. SCA sums them as if they were interchangeable units of creation, so two players with identical totals can be doing entirely different jobs — one a passer, one a carrier, one a foul-magnet — and the metric flattens that difference into a single rank. It is the same category error that makes goal contributions a misleading sum: adding unlike skills produces a number that describes none of them.
Two players credited per shot
By construction SCA credits the last two actions before a shot, which means most shots hand out two shot-creating actions to two different players. That is defensible as bookkeeping, but it dilutes the link between the stat and the skill. The player who threaded the killer pass and the player who made an ordinary lay-off that preceded it both get a shot-creating action, weighted the same. Over a season the totals fill up with these secondary, low-difficulty touches — the equivalent of a hockey secondary assist — and a player can accumulate a healthy SCA rate largely by being the penultimate touch in a fluent side’s buildup, without ever producing the pass that actually broke the defense. The column rewards proximity to the shot, not authorship of the chance.
It still depends on shots happening
SCA improves on the assist by not requiring a goal, but it still requires a shot — and whether a shot follows an action is not fully in the creator’s hands. A perfectly weighted pass into the box that a striker declines to shoot, or takes a touch too many on, produces no shot and therefore no shot-creating action, even though the creative work was identical to one that did draw a strike. The metric quietly outsources part of its own credit to the decisions of teammates downstream. A creator behind a hesitant, low-volume forward line will post a lower SCA than the same creator behind shoot-on-sight strikers, and the gap says more about who he plays with than about how well he creates.
Volume masquerading as quality
Because it is a count of actions, SCA rewards teams and players who generate many shots regardless of how good those shots are. A side that peppers the box with low-value efforts will manufacture shot-creating actions in bulk, and its midfielders’ per-90s will glow, even if most of those chances are hopeful strikes from distance. This is the trap that catches every volume-based creation stat, from key passes to shots on target: the denominator is opportunity, not danger, so a busy but wasteful attack inflates the number. Two playmakers with the same SCA can be feeding a stream of half-chances and a trickle of gilt-edged ones respectively, and the raw total cannot tell you which is which.
What xG-based creation adds back
The honest correction is to weight each created chance by how good it was, which is exactly what expected-assist-style models do — they ask not how many shots a player set up but how much scoring value those shots carried. Expected assists is itself a shot-quality stat, and reading SCA next to an xG-weighted creation number immediately separates the players building genuine danger from the ones running up the count on speculative efforts. Where SCA says a midfielder created four chances, the weighted view says whether those four were tap-ins waiting to happen or thirty-yard prayers. The count tells you the creator was involved; the value tells you whether the involvement mattered.
Reading it on the StatLine tables
StatLine’s Premier League player tables now carry full season lines — goals, assists, shots, minutes and more — and they are the right place to sanity-check any creation stat before trusting it. A midfielder with a shiny SCA reputation but modest assists and a low shot-assist output is probably living on secondary touches and volume; a player whose end-product holds up across shots and assists is more likely doing the authorship the headline number implies. The point is not to discard SCA but to refuse to read it alone. Put the bundle back into its parts, ask which skill is actually driving the total, and check whether the chances created were worth creating.
A tidy number for an untidy thing
Part of SCA’s appeal is aesthetic: it turns the sprawling, improvised business of creating a chance into one clean figure that sorts a leaderboard. That neatness is also the warning sign. Chance creation is genuinely untidy — a mix of vision, technique, movement, and the willingness of teammates to shoot — and any single number that makes it look orderly has hidden the seams rather than removed them. The best analysts treat SCA the way a good scout treats a highlight reel: as a prompt to go look, not as the finding itself. When the total is high, the useful question is always which component drove it and whether the chances were worth anything, and answering that means unbundling the stat back into the different skills it quietly added together. The clean number is where the inquiry starts, never where it ends.
The honest read
Shot-creating actions solved a real blind spot in the assist and widened the credit for building chances, and for that it belongs in the modern toolkit. But the neat per-90 conceals a pile of decisions: it adds together passing, dribbling and foul-drawing as if they were one skill, spreads credit across two players per shot, leans on teammates to actually shoot, and counts chances without weighing their danger. Read as a single creativity ranking, it will hand you players who are near the ball a lot and authors of very little. Read it split into its components, weighted by chance quality, and checked against a player’s underlying output, and it becomes what it was meant to be — a wider window on creation, rather than a tidier way to be wrong about it.