Shooting percentage is a sample-size mirage
A skater on a heater — goals on what feels like every other shot — gets described as a finisher who has figured something out, and his shooting percentage is the number trotted out to prove it. It is an intuitive stat: goals divided by shots, the share of a player’s attempts that beat the goalie. The trouble is that for individual skaters, shooting percentage is one of the least stable numbers in hockey. It bounces violently on the small number of shots a player takes in a season, it regresses hard toward the pack, and a career-best percentage is more often a temporary run of puck luck than a new level of skill. Read as a snapshot of how a stretch has gone, it is fine. Read as a measure of finishing talent or a predictor of future scoring, it lies more than almost any stat the sport keeps.
What shooting percentage actually records
Shooting percentage records the fraction of a player’s shots on goal that went in over some span. That is all. It does not know whether the goals came from high-danger chances at the net front or from low-percentage shots that happened to find a hole; it does not know whether the goalie was screened, whether the puck deflected, or whether a defender’s stick redirected it. It also says nothing about how many shots the player generated in the first place — a player who takes 250 shots and a player who takes 80 can post the same percentage while contributing wildly different amounts of offense. The number is a ratio stripped of context: it tells you how often the puck went in, not why, not from where, and not whether it will keep happening.
The sample is tiny, so the swings are huge
This is the heart of the problem. A typical forward might put 150 to 250 shots on net across a full season; many skaters take far fewer. Percentages built on samples that small are inherently jumpy. The difference between a 9 percent season and a 15 percent season can be a handful of pucks over eighty-two games — a few deflections that went in instead of wide, a few breakaways converted rather than stoned. Flip five or six of those outcomes, which is well within normal variance, and a player’s shooting percentage moves from pedestrian to elite or back again with no change in how he actually shoots. When the underlying sample is this thin, the year-to-year number is dominated by noise, and a single hot or cold season tells you very little about the player underneath it.
It regresses, and the best teams plan for it
Because so much of a skater’s shooting percentage is variance, it regresses strongly toward the average from year to year. A player who shot 18 percent last season is overwhelmingly likely to shoot closer to his career rate next season, and a player who shot 6 percent is likely to bounce back up. This is not a guess; it is the most reliable pattern in the stat. It is why a career year built on an unsustainable percentage is so often followed by a “disappointing” one, and why smart front offices are wary of paying for a contract year that rode a hot finishing run. They separate the player’s shot generation, which is stable and repeatable, from his finishing rate, which mostly is not, and they price the former. Confusing the two is how teams end up paying star money for a percentage that was always going to come down.
Role, linemates, and deployment color the number
Shooting percentage is not purely random — there is some real skill in it — but the part that looks like skill is often role and circumstance. A player who parks at the net front and lives off tips and rebounds will shoot a higher percentage than a defenseman blasting point shots into shin pads, because his shots come from dangerous areas, not because he is a better finisher. A skater playing with elite playmakers gets fed cleaner looks and converts more of them; the same player on a checking line takes tougher shots and converts fewer. Powerplay time, zone starts, and quality of linemates all push the percentage around. So even the non-random portion is partly a story about where a player is deployed and who sets him up, not a clean readout of his hands.
Why the highlight reel fools you
The eye is drawn to goals, and a player riding a high percentage is, by definition, scoring on a lot of his chances right now, so he looks unstoppable on the highlight reel. But the reel only shows the pucks that went in; it never shows the regression coming. A genuinely elite finisher and a player on a lucky run look identical for weeks at a time, which is exactly why shooting percentage is such a tempting and such a misleading way to anoint a breakout. The skaters who keep scoring are usually the ones generating a high volume of good chances, not the ones converting a freakish share of a normal volume. Volume persists; conversion rate drifts back. The highlight reel cannot tell you which one you are watching.
What to read instead
The fix is to lead with shot and chance generation and treat finishing rate as the volatile add-on it is. Shots and shot attempts, high-danger chances, and expected goals describe how much dangerous offense a player creates — and those numbers stabilize far faster than shooting percentage, so they tell you more after twenty games and far more after a full season. Comparing a player’s goals to his expected goals flags who is riding luck and who is genuinely finishing above the rate his chances imply, and over multiple seasons that gap is where the small, real finishing skill shows up. StatLine’s per-shot and expected-goals views are built to make that separation, so a hot percentage does not get mistaken for a new gear.
The honest read
Finishing is a real skill at the margins — a few players genuinely sustain above-average shooting percentages over long careers, and at the extremes the number carries some signal. But for almost everyone, over any single season, shooting percentage is a sample-size mirage: a thin, jumpy ratio that regresses hard and reflects luck, role, and linemates more than a player’s hands. Read a hot percentage as a description of a stretch that is going well, not a promise that it will continue, and look underneath it at how many good chances the player is actually creating. The skater who keeps scoring is the one generating the volume, not the one whose puck luck has not run out yet.