What is a good save percentage for a goalie?

The three columns are meant for a full season, never for a tournament weekend. Only the NHL publishes a league average that is easy to double-check: .900 in 2024-25 (66,294 saves on 73,668 shots), and NHL.com noted .902 at the quarter mark of that same season. For the AHL, the NCAA and the CHL, the only public figures are the leaders’; the “below average” column there is an order of magnitude. For youth hockey, no federation publishes an average: the ranges come from what shows up over full seasons, not from an official reference.

Save percentage throws every shot into the same bag: a soft wrist shot from the blue line and a deflection from three feet out weigh exactly the same. A goalie who faces 15 shots, 8 of them from the slot, plays a much harder game than one who faces 30, almost all from the perimeter. The first will often end up with the lower SV% of the two, and yet he is the one who did the better job.

Quantity matters as much as quality. Behind a team that gets outplayed, a goalie sees a lot of pucks, stays warm, and his SV% settles over a large number of shots. Behind a dominant team, he can go eight minutes without touching the puck and get beaten on the only real chance of the period: one goal on 12 shots is .917; one goal on 6 shots is .833. Same goal, same goalie, two numbers that have nothing in common.

Level, finally, sets the floor. The higher you go, the better defenses protect the slot, the farther shots come from and the higher SV% climbs mechanically. That is why .900 is solid in junior and ordinary in the NHL. To judge a goalie without being fooled, you have to look at where the shots come from and what he does with the most dangerous ones: our guide on how to evaluate a goalie walks through that reading.

A save percentage taken over one, two or three games means close to nothing. On 60 shots, a single extra goal moves the number from .933 to .917; on 25 shots, it moves it from .960 to .920. A bad rebound, a deflection off a skate, a five-minute penalty late in the game: any one of those events is enough to turn a good weekend into a “bad” one on paper.

The prudent rule: wait at least 10 to 15 games, or 200 shots and more, before reading a SV% as a level of play rather than as noise. At that volume, one goal moves the number by about 5 thousandths instead of 15 to 40, and the big swings between opponents start to cancel out. The fewer shots a goalie sees in his league, the more games it takes to reach the same number of pucks.

That is why the right question is not “what is his SV% this week?” but “where has his curve been going over the last 15 games?”. A goalie who goes from .870 to .895 over a season is improving, even if no single game of his shows it. RinkLog’s Progression section draws exactly that curve game after game, with the comparison to goalies of the same age.

Labels change from one federation to the next, but the idea is the same: U9 to U18 at Hockey Canada, M9 to M18 at Hockey Québec, 8U to 18U at USA Hockey. Our page on hockey age groups maps them all. What matters here is that SV% only compares within the same age group and the same tier.

A U9 goalie faces slow shots, often from close range because nobody lifts the puck yet, sometimes on a reduced ice surface: his SV% can swing between .800 and .950 from one game to the next without meaning anything. At U13, body checking arrives in some tiers, shots are harder and defenses start protecting the slot: the number settles, but stays sensitive to the tier (a U13 A and a U13 AAA do not live the same season).

At U15 and U18, goalies aiming for junior start posting numbers close to Junior A, around .900 in the strong tiers. Comparing a U15 AA’s .880 to a U9’s .880 makes no sense: for the goalie, those are two different sports. The only comparison that holds is the goalie against himself, and against goalies of the same age in a similar tier.

SV% is a result, not an explanation. To understand why it goes up or down, you need to know where the shots come from, when they arrive and what the goalie does with rebounds. All of that gets logged from the bench, one tap at a time, and gives a far fairer reading than a single number at the end of the game.

Here is what RinkLog actually tracks for a goalie, and what rounds out SV% without replacing it. Each measure is detailed on the goalie statistics page, and our guide on a young goalie’s progression explains how to read them season after season.

During the game, you tap every save and every goal against from the bench or the stands, with the shot location on the map. RinkLog turns that into the game’s SV%, SV% per period, HDSV% and GAA, then rolls them up over the season. An empty-net goal is logged as an ordinary team goal, never against the goalie: the SV% stays right without you fixing anything.

After 10 games, the Progression section draws the curve and compares it, by percentile, to goalies of the same age and level. Everything the app logs to track a young goalie’s save percentage is detailed on its page. The permanent free plan lets you start tonight; the 7-day Premium trial, no card needed, unlocks the shot map and HDSV%.