The secondary assist is hockey’s luckiest point
Hockey hands out up to two assists on every goal, and the point total — goals plus assists — treats all three credits as equal currency. A goal is a point, a primary assist is a point, and the secondary assist, the credit given to the player who touched the puck before the player who set up the goal, is a point too. That equivalence is the problem. The secondary assist is not the same kind of event as the other two. It is the most loosely connected credit in the box score, awarded for a pass that often had little to do with the goal actually going in, and it behaves statistically far more like a coin flip than like a skill. When we rank scorers by points and treat every point alike, we are quietly stuffing a measure of luck into a number we use to decide award votes, contracts, and legacies.
What the second assist actually credits
A goal in hockey is the end of a sequence, and the scorer scores it. The primary assist — the pass directly setting up the shot — is usually a genuine creative act: the cross-ice feed, the saucer pass to the back door, the delay that opens a lane. The secondary assist is the touch before that. Sometimes it is meaningful, a zone entry or a pass that truly started the play. Far more often it is incidental: a routine breakout pass, a puck chipped off the boards, a clearance that happened to find a teammate who then made two more plays before the goal. The official scorer awards it because a rule says up to two assists may be given, not because the second passer demonstrably contributed to the finish. The credit exists to be handed out, and so it gets handed out.
The repeatability test it fails
The cleanest way to tell skill from luck in any statistic is to ask whether it repeats. Skills persist: a player who shoots accurately this year tends to shoot accurately next year. Luck does not. And when analysts split assists into primary and secondary and check year-over-year correlation, the two behave very differently. Primary assists are reasonably sticky — the players who rack them up tend to keep racking them up, because the playmaking ability behind them is real and stable. Secondary assists are far less repeatable, drifting close to the territory you would expect from noise. A player’s secondary-assist total this season tells you remarkably little about his total next season, which is exactly the signature of a stat measuring circumstance rather than ability.
Why it is so noisy
The noise is structural, not accidental. Whether a player earns a secondary assist depends on events that unfold after his touch and entirely outside his control: whether the next teammate makes a good play, whether the shot goes in, whether the scorer of the goal converts a chance that had nothing to do with the original pass. A defenseman can make the identical first pass a hundred times; the handful of times the sequence happens to end in a goal within the allowed window, he collects a point, and the other ninety-some times he collects nothing. The credit is contingent on a downstream outcome he did not influence. That is the definition of a lucky statistic — the same action is rewarded or ignored based on what other people do afterward.
How it distorts the point total
Because points lump everything together, two players with identical totals can have very different underlying value. One might have built his points on goals and primary assists — repeatable, skill-driven events that will likely recur. The other might be riding a wave of secondary assists that will regress next season, leaving his point total to crater for reasons that look mysterious if you were only watching the top line. Every year a player posts a “breakout” fueled partly by secondary helpers, signs a deal on the strength of it, and then disappoints — not because he got worse, but because the luckiest portion of his production reverted to the mean. Reading the point total without decomposing it sets you up to be surprised by the most predictable kind of regression there is.
The case for keeping it — and reading it carefully
None of this means the secondary assist should be abolished. Some second assists are genuinely excellent plays, and over very large samples a high secondary-assist rate can hint that a player is consistently involved in the early phases of offense, which has some value. The error is not in recording the stat; it is in pretending it carries the same weight as a goal or a primary assist when we sum them into points. The fix is simple and already standard in analytics: separate primary from secondary assists, lean on primary assists and individual shot-creation data when judging playmaking, and treat a heavy secondary-assist share as a regression flag rather than a badge. Weighting the credits by how much they actually predict future scoring restores the signal the point total throws away.
The scorer’s discretion underneath it
Layered on top of the structural noise is a layer of human judgment. Whether a second assist is awarded at all, and to whom, is the decision of an official scorer working in real time, and the standard is anything but uniform. How far back the sequence is allowed to reach, whether a defensive deflection in between wipes out the credit, whether a routine breakout counts as “contributing” — these calls vary from building to building and scorer to scorer. The same play can produce a secondary assist in one arena and nothing in another. That discretion is tolerable for a credit we treat as decorative, but we do not treat it as decorative; we fold it straight into the point total and then compare players across the whole league as though the credits were measured by a single consistent instrument. They are not. A portion of the secondary-assist column is the residue of who happened to be keeping score that night.
The honest read
The point total is a fine headline and a poor verdict. Goals and primary assists are the durable, skill-driven core of a scorer’s production; the secondary assist is the loose change that happens to be denominated in the same coin. Treat a player’s points as a starting question, not an answer, and ask how they were built before you conclude anything about how good he is or how he will age. StatLine’s NHL player tables carry the scoring lines that let you see the composition behind a total, and the composition is where the truth lives. A player whose points lean on goals and primary helpers is standing on solid ground; one leaning on secondary assists is standing on a number that, more than any other in the box score, was handed to him by the bounces.