What The Hell Are "Timely Saves"?

Goalie Memo #4

“Timely Saves” is a phrases I’m sure most hockey fans have heard at sometime in their life. And I’ll just cut to the chase here: it’s an entirely stupid concept.

Now, I can only speak for myself as a goaltender, but I was under the distinct impression that every puck in a game situation is worth stopping to the best of your abilities — with as much technique and/or reckless abandon as deemed necessary. Who knows, though. I never made it out Jr. B. Maybe “timeliness” is why I never made that next step and not being 5’7” like I always assumed it was.

Now, I’ve never asked someone to define a timely save versus a regular save, but we can make some inferences based on what context this comes up.

Let’s try to define this in the best-faith, least straw-man-y way possible. We can assume that an underlying understanding behind the timely save is that score influences hockey games and participant behaviour greatly. This is an extremely uncontroversial statement. Another understanding is that holding leads/preventing goals against is more conducive to winning than not doing so. Big if true. So then, if both of these assertions are true (and I strongly believe they are) then you could say that saves that maintain the current score and prevent goals against are more valuable than those that don’t.

Wait, that’s all saves.

Okay, so we need to refine some parameters here. Well, hockey games are usually considered close if the absolute difference of the score is no more than 2. So we can call saves “timely” if they prevent a goal against while the score differential is no greater than 2. But that’s also kinda ridiculous because saves made by losing goalies while their team is leading are typically never commended for their “timeliness” — regardless of xGoal quantities or eye-test qualities or flow of play context.

So, I don’t think I can strong man this argument. It reminds me a lot of game-winning goals. You definitionally can’t win without a GWG, but are they all created equal? God, no —some aren’t even created when they’re scored! If a team jumps out to a 4-0 lead, but holds on to win by 4-3, that 4-0 goal turns from stat-padding to something tantamount to an OT-winner. Let’s take this Stars-Oilers game from March 26th, 2025 that matches my above description. By the time of Jason Robertson’s hat-trick, it only caused a 4.3% swing in win probability according to MoneyPuck. Keep in mind that in the MoneyPuck model, as time in a tie game approaches 60:00, the model approaches 50.0% win probability at 5-on-5 (always measure from the home team’s POV), just as it approaches either 0% or 100% for whichever team is leading.

But flipping this back to Jake Oettinger’s point of view, which saves here are “timely? Are the ones he made when the score was 4-0 timely? He ended up needing those ones due to the eventual 1-goal lead. Or how about the 2 shots Oettinger faced after the Oilers made it 4-3? Were those the “timely saves” that ended up being the difference? Are any of these shots more valuable than the rest? How can you point to any of Oettinger’s 41 saves that night and say, “well if he didn’t make that one, they would have lost or had to go to OT”?

I want to close by showing this goal against Joseph Woll from Game 6 of the 2024 first round between Boston and Toronto. Toronto has a 2-0 lead with less than 5 seconds to go, and Joseph Woll attempts to execute a desperation save that injures him and renders him unable to play in the deciding Game 7, which (despite a valiant effort from his replacement, Samsonov) his team ends up losing. In the big picture, this is a foolish attempt. Had Woll known perfectly the outcome of this play — a functionally meaningless goal against and his being injured — he probably would not have attempted the save. But he didn’t know that— perhaps he suspected the former, but certainly not the latter. It’s only from knowing the outcome that one might see this save attempt as pointless. There is no conscious elevation of performance in clutch scenarios because, as Woll brutally demonstrates here, there is no conscious laxing of performance in low-leverage scenarios.

Seeing that play for the first time, I understood at a fundamental level that the end result was entirely unavoidable. Woll was going to dive for that puck no matter what the scoreboard said. Goalies can’t pull punches in the moment. Our margins of success are thin enough as is — at our best, we still get beat. We are crisis managers and damage mitigators. When the net is empty, all shots on goal are goals. Thus, every save is important and all saves are timely.

We’re not quarterbacks — we don’t call the plays and distribute the offense. We’re not pitchers — we don’t have a monopoly on the ball and start each play with it. We’re just the folks in pads throwing ourselves in front of someone else’s offence.

Every time.