A fan's guide · NHL injury study

Does a star make his linemates better?

When a star is out injured, his linemates allow the other team more quality scoring chances. On offense, Sidney Crosby stands out: without him, his linemates create far fewer chances of their own.

Hugh Allen · b.hugh.allen@gmail.com · October 2026
For the full statistical write-up, switch to the technical version.

The short answer

What we found

Every result controls for home ice, opponent strength, where shifts start, lineup strength and roster changes, and compares each linemate only with himself in the same season.

Defense · all six stars

4fewer chances against

Controlling for these, the star being in the lineup is associated with a 0.18 drop in his linemates' expected goals against per 60 minutes.1 That's about 4 fewer goals' worth of opponent chances per linemate over a season.

Clear evidence

Offense · Crosby

14more chances created

Controlling for these, Crosby being in the lineup is associated with a 0.66 increase in his linemates' expected goals for per 60 minutes.2 That's about 14 more goals' worth of chances per linemate over a season.

Clear evidence

Defense · Kaprizov

10fewer chances against

Controlling for these, Kaprizov being in the lineup is associated with a 0.44 drop in his linemates' expected goals against per 60 minutes.3 That's about 10 fewer goals' worth of opponent chances per linemate over a season.

Clear evidence

For most of the stars individually, the data can't say much either way. Taken together, the clearest pattern is defensive: linemates keep the other team's chances down when the star plays. Offense improves too, but mostly because of Crosby. All numbers are 5-on-5; "per 60 minutes" means per 60 minutes of the linemate's own ice time.

  1. 95% confidence interval: 0.02 to 0.34 fewer; p = 0.03. Combined across all six stars.
  2. 95% confidence interval: 0.25 to 1.07; p = 0.002. Without the lineup-strength and roster-change controls the estimate is 0.26 (p = 0.04), about 5 goals' worth of chances a season; a simple average of games with and without him shows about 3 or 4. Most of the effect comes from 2011-12; the 2010-11 and 2019-20 estimates are smaller and could be chance. So his linemates clearly did better with him, but by somewhere between about 5 and 14 goals' worth of chances a season.
  3. 95% confidence interval: 0.03 to 0.85 fewer; p = 0.04. Based on the 40 games he missed in 2024-25.

How we tested it

Injuries as an accidental experiment

A coach would never bench his best player for 20 games just to see what happens. Injuries do it for us. When a star is hurt, his regular linemates keep playing with the same team, in the same season, just without him. So we compared how those same linemates played in games with the star and in games without him.

One linemate's season (illustration)

Game with the starGame without him (injured)

We studied six stars who missed long stretches: 434 games in total without them, with dozens of their regular linemates.

A team without its star is different in other ways too, so before comparing we accounted for:

  • How strong each opponent was
  • Home or away
  • Where shifts started (offensive or defensive zone)
  • Other players coming and going through injuries, trades and call-ups

The stats, in plain words

What we measured

Expected goals

Not every shot is equal. A one-timer from the slot is far more dangerous than a wrister from the blue line. Expected goals adds up the quality of every shot, measured in goals. "4 goals' worth of chances" means chances that would usually turn into about 4 goals.

Chances created: more is better
Chances against: fewer is better

Shot attempts

Every shot attempt for and against while a player is on the ice. The share his team gets (often called Corsi) is a simple measure of who has the puck more.

Share of attempts: higher is better

Primary assists

The last pass before a goal. We checked whether linemates set up more goals with the star around. Nothing clear showed up, so it isn't charted here.

More is better

All numbers are from even-strength play (5-on-5), so power plays and penalty kills don't skew them. Every number is for one linemate over a full 82-game season of playing with the star.

Player by player

Who makes a difference

Clear evidence A hint, not proof Can't tell: could easily be chance

Defense: opponent chances kept down

Goals' worth of the other team's chances each linemate avoids per season with the star. Bars to the right are good.

Goals' worth of opponent chances each linemate avoids per season with the star ← more against fewer against → All six stars All six stars: +4 (Clear evidence) +4 Clear evidence Kirill Kaprizov Kirill Kaprizov: +10 (Clear evidence) +10 Clear evidence Mark Stone Mark Stone: +6 (A hint) +6 A hint Matthew Tkachuk Matthew Tkachuk: +5 (Can't tell) +5 Can't tell Jack Eichel Jack Eichel: +4 (Can't tell) +4 Can't tell Quinn Hughes Quinn Hughes: +2 (Can't tell) +2 Can't tell Sidney Crosby Sidney Crosby: -2 (Can't tell) -2 Can't tell no change Goals' worth of opponent chances each linemate avoids per season with the star ← more fewer → All six All six: +4 (Clear evidence) +4 Clear Kaprizov Kaprizov: +10 (Clear evidence) +10 Clear Stone Stone: +6 (A hint) +6 Hint Tkachuk Tkachuk: +5 (Can't tell) +5 Can't tell Eichel Eichel: +4 (Can't tell) +4 Can't tell Hughes Hughes: +2 (Can't tell) +2 Can't tell Crosby Crosby: -2 (Can't tell) -2 Can't tell no change

Offense: chances created

Extra goals' worth of chances each linemate's shifts create per season with the star. Bars to the right are good.

Extra goals' worth of chances each linemate's shifts create per season with the star ← fewer more → All six stars All six stars: +4 (Clear evidence) +4 Clear evidence Sidney Crosby Sidney Crosby: +14 (Clear evidence) +14 Clear evidence Matthew Tkachuk Matthew Tkachuk: +5 (Can't tell) +5 Can't tell Kirill Kaprizov Kirill Kaprizov: +4 (Can't tell) +4 Can't tell Mark Stone Mark Stone: +2 (Can't tell) +2 Can't tell Jack Eichel Jack Eichel: -1 (Can't tell) -1 Can't tell Quinn Hughes Quinn Hughes: -6 (Can't tell) -6 Can't tell no change Extra goals' worth of chances each linemate's shifts create per season with the star ← fewer more → All six All six: +4 (Clear evidence) +4 Clear Crosby Crosby: +14 (Clear evidence) +14 Clear Tkachuk Tkachuk: +5 (Can't tell) +5 Can't tell Kaprizov Kaprizov: +4 (Can't tell) +4 Can't tell Stone Stone: +2 (Can't tell) +2 Can't tell Eichel Eichel: -1 (Can't tell) -1 Can't tell Hughes Hughes: -6 (Can't tell) -6 Can't tell no change

Hatched bars can look big, but they mean the data can't separate an effect from luck. Quinn Hughes missed only 14 games, for example, which is too few to judge.

The stories

Six stars, six answers

Sidney Crosby

Pittsburgh · 2010-11, 2011-12, 2019-20

The standout. Concussions cost Crosby most of two seasons, and his linemates' offense fell off without him. It's the strongest result in the study, though its size depends on how the roster changes are handled.2

Offense: clear evidence

Kirill Kaprizov

Minnesota · 2024-25

A defensive surprise from an offensive star. With Kaprizov, his linemates allowed far fewer chances, and their share of shot attempts rose from about 46% to 52%.

Defense: clear evidence

Mark Stone

Vegas · 2021-22 to 2025-26

Known for his defensive game, and the data leans that way: about 6 fewer goals' worth of opponent chances per linemate per season. Not every season agreed, so it's a hint rather than proof.

Defense: a hint

Matthew Tkachuk

Florida · 2024-25, 2025-26

His linemates did a little better on both ends with him, but not by enough to rule out luck.

Can't tell

Jack Eichel

Vegas · 2022-23, 2023-24

Only 30 games without him to compare, and nothing clear in either direction.

Can't tell

Quinn Hughes

Vancouver · 2024-25

Just 14 games missed, the fewest of the six. That isn't enough to judge, so his numbers bounce around.

Can't tell

Being honest about it

What this can't tell us

  • Six players is a small group. The defensive pattern points the same way no matter which star is left out, but more players would make it firmer.
  • Luck can fool you when you look at many numbers. Six stars times several stats means one result could look real by chance. Crosby's is strong enough that luck is very unlikely; Kaprizov's is less certain.
  • Injuries aren't perfectly random. Coaches shuffle lines and call players up when a star goes down. We accounted for the roster moves we could see, but not every change behind the bench.
  • It's about chances, not goals. Expected goals measures the quality of chances. It leaves out hot goalies and finishing luck on purpose, because those swing wildly over short stretches.
  • It only measures linemates on the ice. It says nothing about leadership, the locker room, or players on other lines.