Goalie value: GSAx, High-Danger GSAx & Above-Expected Starts
Our headline goalie numbers, built on the same shot model as our skater work. By design, the model never knows which goalie is in net.
Goals Saved Above Expected (GSAx)
For every shot a goalie faces, our expected-goals model estimates its chance of becoming a goal. It looks at the shot's location and angle, its type and the play that created it. GSAx is this season's goals saved above expected. It is the sum of those expected goals minus the goals actually allowed. Positive means the goalie stopped more than a league-average goalie would have on the same shots. It credits how hard the shots were, not just raw save percentage. It is a scorecard of the season you are looking at. It says what happened, not what comes next.
High-Danger GSAx
Not every save tells you the same amount. We split shots into three danger tiers by expected-goal value. High-Danger GSAx runs the same calculation on the toughest chances only. It is the least settled goalie number on this page. It rests on fewer shots. Treat it as a rough guide, not the sharper read. We are still working on it.
| Danger tier | Expected goals per shot | ≈ Share of shots |
|---|---|---|
| Low | under 0.05 | 52% |
| Medium | 0.05 – 0.12 | 31% |
| High | 0.12 and up | 17% |
Above-Expected Starts
A start counts as an above-expected start when the goalie's GSAx for that game is zero or better. That means he was at least league-average against the shots he actually faced. Above-Expected Start % is the share of a goalie's starts that clear that bar. It deliberately ignores volume. It asks whether the goalie matched or beat expectation on the night, not whether he happened to face few shots. That breaks from the usual quality-start rule, which is based on save percentage. That rule sets an easier bar on nights with few shots.
How we keep it honest
These metrics need only shot locations, so they reach back to 2009-10. That is 18 seasons of goaltending in all situations, measured with one consistent shot-quality yardstick. Season percentiles count only goalies with a real workload of at least 20 starts. So a handful of spot starts can never produce a leaderboard rank.
Before any season publishes, we check that the rankings land where a knowledgeable fan would expect. A known elite goaltending season has to rank near the top. It is a sanity check on how shots are matched to goalies.
| Coverage & discipline | Value |
|---|---|
| Seasons covered (all situations) | 2009-10 through 2026-27 (18) |
| Goalie-seasons published | 1,686 (328 goalies) |
| Cleared the 20-start percentile floor | 883 (177 goalies) |
| Shot mix scored (low / medium / high danger) | 52% / 31% / 17% |
The honest caveat: goalie value is noisy
Single-season goaltending swings far more than anything a skater does. We would rather say so than sell a number as steadier than it is. A season of GSAx rests on a few hundred shots against. A handful of goals either way moves it a long way. How much one goalie season says about the next is being re-measured on today's numbers. Use GSAx to read the season that happened.
Goalie Wins Above Replacement
We also convert a goalie's saves into wins. Goalie WAR starts from what a replacement goalie would have allowed on the same shots. It counts the goals a goalie saved above that, in wins. It counts the season as it happened: nothing is pulled toward average first. It is shown for starters only. A starter has 25 or more starts in a season. That is 741 goalie-seasons in all.
How the goalie WAR is set
A season's saves above expected are added to the credit for beating a replacement goalie on the same shots. That total is turned into wins at the same rate skaters use for that season. Nothing is pulled toward average first. The badge is what the season was. The best goalie season in a year is worth several wins. That puts it above the best skater's WAR in most seasons and below it in a few.
Counts on this page are read from the published data through 2026-27. A figure from a one-time study says when it was measured.