No. Across 588 rookie-to-sophomore pairs since 2007-08, scoring goes up in year two — points per game rise 8.6% and total points rise 28.4%. 55.6% of players improve on a per-game basis.
But there is one group that really does fall off, it falls off hard, and you can identify it in September from a single number. That is the useful half of this post.
What happens to an NHL player in his second season?
Read the first two rows together, because they are the whole story. Points per game jump 8.6%. Points per sixty minutes — the rate that measures the player rather than his opportunity — move 0.6%, which is nothing.
Second-year players are not better at hockey than they were as rookies. They are on the ice more (14.56 minutes a night to 15.39) and they are in the lineup more (50.1 games to 60.0). That is where every point of the gain comes from. A rookie spends half his season earning a coach's trust; a sophomore starts with it.
So why does everyone believe in the slump?
Because of who you remember, and because of regression to the mean.
Split every rookie season into five groups by scoring rate and the pattern is textbook:
The rookies who scored least improve the most (+0.079 points per game). The rookies who scored most improve the least (−0.013). Nobody writes a column about a fourth-liner going from 0.125 to 0.205. Everybody writes one about the kid who put up 0.623 and then didn't.
That is not a slump. That is an unusually good number being an unusually good number.
Which rookies actually do slump?
Here is the part that is actually actionable. Take only the rookies who scored at a high rate — 0.45 points per game or better — and split them by how many games their rookie season lasted:
A high-scoring rookie who did it over a full season gets better in year two, by +0.039 points per game. A high-scoring rookie who did it in under 45 games loses 0.083 — a drop of 15.0%, and the only one of the three that clears statistical significance on its own.
And it is specific to the high scorers. Run the same split across all 588 pairs and every group improves: +0.023 for under 45 games, +0.030 for 45 to 64 games, +0.038 for 65 or more games. Short rookie seasons are not the problem. Short rookie seasons with a hot scoring rate attached are, because the rate was never real to begin with — it was 30-odd games of shooting luck and favorable matchups, and 82 games of hockey finds that out.
Who slumped, and who did not?
The hot short-season rookies:
Two things are worth saying honestly about that table. First, the rate collapse is real and it is large. Second, several of those players were still good — Jake Guentzel's rate fell by a quarter of a point a game and he went on to a very long, very productive career. A sophomore whose rate regresses is not a bust; he is a player whose true level was lower than his first 40 games suggested. In fantasy terms, this group averaged 0.552 points per game as rookies and 0.469 as sophomores. You are asked to draft the first number and you get the second.
There is at least one player who broke the pattern completely: William Nylander went 0.591 to 0.753 off a 22-game debut. The bucket is a tendency, not a law.
Now the hot full-season rookies:
Macklin Celebrini is the cleanest recent example of what the data predicts: 63 points in 70 rookie games, then 115 in 82. Andrei Kuzmenko is the counterexample that keeps this honest — a full 81-game rookie season at 0.914 followed by 0.639. The full-season group improves on average; individual players still bust.
Which 2026-27 sophomores does this apply to?
Four players from last season's rookie class scored at least 0.50 points per game, and the rule above splits them cleanly.
Beckett Sennecke and Matthew Schaefer each did it over a full season, which puts them in the group history says improves. Igor Chernyshov and Matvei Gridin did it in 28 and 37 games, which is the group that regresses — and both are being drafted on a rate that 28 or 37 games cannot support.
Two honest notes about our own numbers, because they do not perfectly agree with the history above.
Our fantasy projection has Matthew Schaefer going from 0.720 to 0.785 points per game, a bigger jump than the +0.039 the full-season group averages. That one is defensible: at 19.1 he is young enough that the age curve and a rising ice-time curve both point up, and teenagers are the subgroup most likely to beat the average.
The harder one is Matvei Gridin. History says a 37-game rookie rate should come down; our projection takes him up, from 0.541 to 0.611. It gets Igor Chernyshov right, moving him from 0.679 down to 0.607. Where the model and this page disagree, the 37-game sample is the reason to be careful, and we would rather point that out than let it sit in a table unmentioned.
Frequently asked questions
Is the sophomore slump real in the NHL?
Not as a general rule. Across 588 rookie-to-sophomore pairs since 2007-08, points per game rose 8.6% and 55.6% of players improved. The exception is rookies who posted a high scoring rate in a short season — under 45 games — who lost 0.083 points per game on average.
Why do second-year NHL players score more?
Almost entirely opportunity rather than improvement. Points per 60 minutes change by only 0.6% from year one to year two, while ice time rises from 14.56 to 15.39 minutes a game and games played from 50.1 to 60.0.
Which rookies should I expect to regress?
The ones whose scoring rate came from a small sample. A player who scored 0.60 points per game over 30 games is a much worse bet than one who scored 0.60 over 75 games, even though the rate on the page is identical. Rookie games played is the number to check.
Does a regressing sophomore mean the player is a bust?
No. The rate coming down means his first sample overstated him, not that he is finished. Several players in the short-season group above lost a quarter of a point per game in year two and still had long, productive careers.
How many games does a rookie season need for the rate to be trustworthy?
On this evidence, about 65. High-rate rookies above that threshold improved the following year; those below it did not. Between 45 and 64 games the effect is negative but small enough that it does not separate from noise on its own.
Who are the 2026-27 NHL sophomores to watch?
Beckett Sennecke (60 points in 82 rookie games), Matthew Schaefer (59 points in 82 rookie games), Igor Chernyshov (19 points in 28 rookie games), Matvei Gridin (20 points in 37 rookie games). The table above has each one's rookie rate, how long the rookie season was, and what history says happens next.
Where these numbers come from
Every regular-season box score from 2007-08 through 2025-26, 4,338 player-seasons in all. A rookie season is the first season a skater appears in, and both seasons in a pair need at least 20 games so that call-ups and injury-wrecked years are not compared to full ones. Goalies are excluded. Rates are points per game and points per sixty minutes of ice time, never raw totals, because a total mostly measures how often a coach used a player.
The 2026-27 projections are from our fantasy model; the full boards are in the draft kit and every skater's projection is on the fantasy page. If you want the other thing that moves a young player's point total more than development does, it is power-play time.