Skip to content

Explainers / TermUpdated September 30, 2026

Wins above replacement (WAR)

WAR estimates how much a player contributes beyond a depth replacement, expressed in wins. A skater with 2 WAR has contributed roughly two wins more than that replacement baseline over the games measured.

“Replacement” means the depth players a team turns to when it runs short of regulars. That is a lower bar than the average NHL player: being an average regular still has value.

A simple example

Imagine a skater whose play adds 1.5 wins above average. A depth replacement, given the same workload, would be expected to finish 0.5 wins below average.

The gap between them is 2 WAR:

1.5 − (−0.5) = 2 wins above replacement.

These are hypothetical numbers. The comparison puts contributions from different parts of the game on one scale; it isn’t a count of games the player personally won.

What goes into wRink’s WAR?

For skaters, we combine five parts of the game:

  • Process (five-on-five play): creating chances and limiting chances against, with adjustments for teammates and opponents.
  • Finishing: scoring above or below the expectation for the chances taken.
  • Power play: contribution with the man advantage.
  • Penalty kill: contribution while shorthanded.
  • Penalties: the value of drawing penalties versus taking them.

Those contributions add up to wins above average. We then change the comparison from average to replacement, making that adjustment once for the whole player. Different WAR models make different choices; these are wRink’s building blocks.

Where replacement comes from

For each team and season, we set aside the top 13 forwards and top seven defencemen by five-on-five minutes. The remaining depth players define a separate replacement baseline for each position. The gap below average, applied to a player’s five-on-five minutes, supplies one replacement adjustment.

This baseline answers a different question from a player-page rating: contribution above depth replacement, rather than standing among current NHL peers.

Reading the number

Positive WAR means above the replacement baseline; negative WAR means below it. The total reflects both performance and playing time, so the components help explain how a player got there.

WAR per 82 covered games puts recent contribution on a common scale. Covered games are the games our records can measure, which can be fewer than games played. Player listings pool available WAR and covered games across 2023–24 to 2025–26, then multiply by 82. A hypothetical player with 2 WAR in 82 covered games has 2 WAR per 82: 2 ÷ 82 × 82 = 2. Listings require at least 60 covered games across those three seasons. Shorter histories use their available games only when they meet that minimum; otherwise recent rates and ratings stay blank. This is a rate in completed games, not a prediction of a full season.

Full method: Replacement and WAR