Save percentage is a rumor, not a verdict
Save percentage is the number everyone reaches for, and over a short stretch it is close to meaningless. A goalie can post a gaudy save percentage for a month simply because the shots he faced were easy, or a poor one because he faced a barrage from the slot. To read a goalie honestly, you have to separate what he did from what he was asked to do.
Quality over quantity of shots
Raw shots against tells you how busy a goalie was, not how hard the job was. The better lens is shot quality. Ten point-blank chances from the slot are a far greater test than thirty perimeter wristers a goalie can see all the way. This is why the most useful goalie stat is goals saved above expected, which compares the goals a goalie actually allowed to the goals an average netminder would have allowed on those exact chances.
- Positive goals saved above expected: he is stopping more than the chances suggest he should
- Around zero: he is roughly a league-average finisher of the chances he faced
- Negative: the chances were manageable and he let in more than he should have
That single adjustment strips out the biggest lie in the save percentage: that all shots are equal.
What workload actually does
Workload matters, but not always the way people assume. There are two competing effects:
- A goalie in rhythm, playing often, can find and hold his form, tracking pucks sharply game to game
- A goalie run into the ground with too many starts and no rest tends to fade late in a season, his numbers slipping as fatigue accumulates
The goal is not to play your starter the most. It is to play him enough to stay sharp and rest him enough to stay fresh. The teams that manage that arrive at the playoffs with a goalie who has something left.
The smartest organizations treat the crease like a pitching staff, planning starts weeks ahead to protect their best netminder for the games that matter.
Reading form without fooling yourself
If you want to judge whether a goalie is genuinely hot or merely lucky, ask three questions:
- Is his goals saved above expected strong, or just his raw save percentage?
- How large is the sample? A great two weeks is noise. A great two months is a signal.
- Has his workload been sustainable, or is he due for regression from fatigue?
The honest takeaway
Goaltending is the highest-variance position in the sport, which is exactly why the surface numbers mislead so often. Look past save percentage to the quality of chances and to goals saved above expected, respect the sample size, and account for how many miles are on the goalie's legs. Do that, and you will spot the netminder in real form well before the highlight reel catches up, and you will stop being fooled by the one who is simply getting lucky.