NHL.com article on the drop in save percentage

[FunkySeeFunkyDo]

Theory is SV% is down because shots are down, because low probability of scoring shots are down. Are shots down? Next question—are GAAs up because SV% are down or down because shots are down?

 

[ijuka]

Theory is SV% is down because shots are down, because low probability of scoring shots are down. Are shots down? Next question—are GAAs up because SV% are down or down because shots are down?

The average xGoals per shot is up.

 

[ijuka]

Love the progression of analytics where people have gotten their heads around the concept of shot quality. Still a ways to go such as doing xGF and GSAA with factors other than shot location.

Depending on the model, they do make use of factors other than shot location. For example, zone transitions and the vector between consecutive events.

The thing with corsi is that it's descriptive, not prescriptive. It works if it's a byproduct of what happens naturally, as it's a solid way of estimating possession. It does not work if you intentionally take low quality shots in order to inflate your corsi.

It's sort of like having a nutritional label for a chocolate cake versus having a recipe for it.

 

[FunkySeeFunkyDo]

The average xGoals per shot is up.

This is what I’m getting at—despite SV% going down, GAA is down too. Maybe shooters are getting too picky. NHL League Averages | Hockey-Reference.com

 

[dgibb10]

When Gretzky recorded 215 points in 1985-86, the average save percentage was only 0.872 compared with the current 0.901. This is a 2.9% difference in save percentage.
Based on this, Gretzky would only have had 209 points in 2024.
I know there are other mathematic variables, but regardless, this points towards Gretzky’s inconceivable dominance.

This is an act of terrorism against whoever your teacher was the last time you took a math class ngl

 

[dgibb10]

This is a hilarious comment considering Corsi has largely fallen out of favour to scoring chance and shot-quality based metrics (i.e. expected goals).

Whoever decided to name Corsi and Fenwick after the people who started using them needs to pay for their crimes.

They're really simple stats that the name has convinced a large number of people is something made up by wizards.

It's basically equivalent to looking at OBP instead of Average in baseball.

It's literally just shot attempts and unblocked shot attempts.

 

[johan f]

The league wanted more goals. We now see more goals. With that follows a drop in save percentage. Nothing to see here.

 

[FoxyClean]

The 2010s had the best goaltending the league has ever seen. On average, what we've seen in the 2020s isn't even close in quality.

 

[braunm]

This is an act of terrorism against whoever your teacher was the last time you took a math class ngl

Apologies-this was indeed an act of Mathematical terrorism! Embarrassing

 

[SladeWilson23]

Theory is SV% is down because shots are down, because low probability of scoring shots are down. Are shots down? Next question—are GAAs up because SV% are down or down because shots are down?

Shot totals in game do have an effect on SV%. GAA doesn't go up or down because of SV%. GAA moves up and down from shot totals.

Higher shot games means higher SV% and higher GAA.

Lower shot games means lower SV% and lower GAA.

The league wanted more goals. We now see more goals. With that follows a drop in save percentage. Nothing to see here.

It's not that simple. SV% during DPE was also low.

 

[FunkySeeFunkyDo]

The league wanted more goals. We now see more goals. With that follows a drop in save percentage. Nothing to see here.

Goals per game were down last year and are down again so far this year and sv% is down. Shots were down 1 last year and are down a further 1.5 so far this year.

 

[DanielBrassard]

I don’t think vasilevskiy’s theory is quite correct, if you look at Corsi for/60 it’s actually higher this season and last than it has ever been. Last season was slightly higher than this year. For some reason the shots are not getting to the net or on net quite as often. And while sv% might be down the number of goals are going down because there are fewer shots by a not insignificant amount. So if shooters are buying into the theory of waiting for the perfect shot then it’s not really working.

 

[Anglesmith]

I don’t think vasilevskiy’s theory is quite correct, if you look at Corsi for/60 it’s actually higher this season and last than it has ever been. Last season was slightly higher than this year. For some reason the shots are not getting to the net or on net quite as often. And while sv% might be down the number of goals are going down because there are fewer shots by a not insignificant amount. So if shooters are buying into the theory of waiting for the perfect shot then it’s not really working.

The perfect shot doesn't mean an open net, necessarily. It could mean waiting for traffic. It's possible that shooters are shooting less when they are sure the puck will get through and more when they aren't sure if it will get through.

 

[Nogatco Rd]

Get back to me when they address the drop in Claude Giroux-Adjusted Points

 

[NYRfan85]

These attempts at adjusted stats are flawed because they try to treat the outlier as just another player. Where in these formulas does it factor in Gretzky and the Oilers playing with the two line pass rule versus not playing with it today?

When McDavid scored 153, the Oilers led the league with 325 goals. If I’m being honest, I think modern day Gretzky and the Oilers still hit 400 because Gretzky without the two line pass rule might demolish the league more than he did in reality. It won’t make sense to people trying to do straight calculations that can’t factor in context, but nothing about todays league leads me to believe that just because the overall goals scored are 25% less than they were forty years ago, Gretzky and the Oilers themselves would be held back proportionally.

This right here is something people don't realize. If the two-line pass rule didn't exist back then (SO glad they abolished it, a total garbage rule), then Gretzky could have easily done more damage.

 

[Reality Czech]

Theory is SV% is down because shots are down, because low probability of scoring shots are down. Are shots down? Next question—are GAAs up because SV% are down or down because shots are down?

Interesting observations. I had two random thoughts as well. Blocked shots are way up but most blocked shots come from farther out, which are generally low percentage chances for the shooter. So those are saves a goalie never has a chance to make. Could be one reason why more shots are "high danger" compared to the past.

Also, what percentage of games end in an OT goal? I could see that lowering sv% because 3 on 3 scoring chances are more likely to be high danger. One goalie basically has an extra goal tacked onto his stats every time a game ends in OT.