JuniorPuck logoJuniorPuck

Understanding Stats

GSAA: Measuring How Much a Goalie Beats the Average

Goals Saved Above Average adjusts for shot volume and measures a goalie's performance relative to what a league-average goalie would have done. Why wins and GAA mislead, and how to read GSAA on JuniorPuck.

Going deeper5 min readUpdated May 11, 2026

Goalie wins and GAA are popular because they are easy to understand. They are also among the most misleading stats in hockey. GSAA fixes most of their problems by asking a cleaner question: compared to what a league-average goalie would have done with the same shots, how many extra goals did this goalie save?

Why wins and GAA mislead

Wins are a team stat. A goalie on a dominant offensive team wins games because their team scores four goals, not because they are exceptional. A goalie on a weak defensive team allows more goals against because the defense generates more shots, not because they are poor. Raw wins and losses conflate team quality with individual goalie performance.

GAA has the same problem in a different form. A goalie facing 40 shots per game on a weak defensive team has a mechanically higher GAA than a goalie facing 22 shots per game on a defensive powerhouse, even if the former is actually stopping a higher percentage of what they face.

How GSAA is calculated

GSAA starts with a simple question: if a perfectly average goalie (one with the league-average save percentage) faced the same number of shots, how many goals would they allow? The difference between that expected total and the actual goals allowed is GSAA.

Concretely: a goalie who faces 1,200 shots in a season where the league-average SV% is .900 would be expected to allow 120 goals at average performance. If they actually allow 105 goals, their GSAA is +15; they saved 15 more goals than average. If they allow 135, their GSAA is −15.

By the numbers

GSAA is the fastest way to find which CHL goalies are genuinely outperforming expectations. On JuniorPuck's Goalies page, sort by GSAA descending to find the best-performing goalies in the league, adjusted for team quality and shot volume.

Limits of GSAA as shown on JuniorPuck

The GSAA calculation on JuniorPuck uses shot volume and the league-average SV%: it is a volume-adjusted metric, not a shot-quality-adjusted metric. Ideally, GSAA would account for the difficulty of the shots a goalie faces (using xG per shot), so that a goalie facing only slot shots is not compared unfairly to one who faces mostly perimeter shots. That level of granularity requires per-shot xG for goalies, which depends on having shot coordinates for every shot, and WHL shot coordinates are not in the data feed. So the site's GSAA is a solid approximation, not a perfect read.

Related reading