After four advanced stats articles breaking down the prowess of skaters, it’s time to head to the crease and help you further examine goaltenders. The next stat up in our series of Advanced Stats for VGK Dummies is GSAA or Goals Saved Above Average.
For those who are baseball fans, this is essentially WAR (wins above replacement) for goalies. For those who aren’t baseball fans, smart thinking, baseball is boring.
Usually, we first try to explain what the stat is before we get into the formula of how it’s calculated, but this one is a bit convoluted when explained that way that we’ll start by simply saying, GSAA measures individual goalies against the league average goalie. That’s all you need to know, now follow through how it’s calculated.
First, we need to calculate the league average goalie. To do so, we take every save made by every goalie in the NHL and divide it by every single shot on goal over the course of a season.
To simply the numbers let’s use a hypothetical using just one game. Say the first game of the season there are 50 total shots on net, 25 by each team. A total of five goals are scored, so 45 shots are saved. Thus, the league average is 45/50=0.900.
Now, we take an individual goalies stats. Let’s say he gave up one goal on 25 shots, meaning he saved 24.
GSAA = [Shots against x (1 – league-average save percentage)] – goals allowed
So, our goalie faced 25 shots, he allowed 1 goal, and the league average is 0.900.
[25 x (1-0.900)] – 1[25 x 0.1] – 1
2.5-1
GSAA = 1.5
In 2016-17 there were 74,046 total shots on goal. A total of 67,638 were saved. Thus, the league average was 67638/74046=0.913459201. (Note: Rounding only occurs when the formula is completed. Hence the ridiculously long decimals, and to answer your question, yes, it does make a difference.)
Marc-Andre Fleury faced 1,181 shots, allowed 107 goals, and the league average is approximately 0.913.
[1181 x (1-0.913459201)] – 107[1181 x 0.086540799] -107
102.204683619 – 107
GSAA= -4.795316381 (Rounded to -4.80)
This stat is graded off a basis of 0. If a goalie is exactly average his GSAA would come out to 0.00. So, a goalie who has a positive number is above average, a goalie that has a negative number is below average. Here’s a quick look at the top and bottom goalies from 2016-17.
Player | GSAA |
---|---|
Sergei Bobrovsky | 33.45 |
Braden Holtby | 19.25 |
Devan Dubnyk | 18.41 |
Carey Price | 17.25 |
Craig Anderson | 15.92 |
Player | GSAA |
---|---|
Antti Nieme | -18.84 |
Petr Mrazek | -18.48 |
Kari Lehtonen | -17.21 |
Cam Ward | -13.93 |
Calvin Pickard | -13.56 |
The numbers don’t look great for either of the two Vegas goaltenders (Fleury -4.80, Pickard -13.56). The good news is, Fleury had an off-year based on his career, which could’ve been caused by a lack of ice time in Pittsburgh. We know that won’t be the case in Las Vegas.
Last season, the Vegas goaltender played the least amount of minutes and games in his past ten. Likely an indication why Fleury’s GSAA fell below league average for only the fifth time in his 13-year career. Many argue Fleury never found a rhythm last season sharing a spotlight with Matt Murray. The 32-year-old had gaps of five, 14, 20+ days off in-between starts. That’s uncommon for an All-Star like Fleury. In 2011, he started 23 consecutive games. As a Golden Knight, Fleury will have the chance to start his normal amount of games per season. Hopefully, this will create a more positive flow for the three-time Cup winner.
As mentioned before, #29 has only fallen below the league average five times; -9.89, -5.38, -10.77, -0.13, and -4.80. Two seasons he split starts, and another was marred by Crosby and Malkin injuries. With all of that in mind, Fleury closed out in Pittsburgh with a 1.99 GSAA in 15 postseason starts during the 2017 postseason, a good sign the Golden Knight’s backstop is coming in hot.
As for Calvin Pickard‘s disastrous -13.56, we’ll just chalk that up to Colorado being Colorado. In 16 games his rookie year though he put up an outstanding 8.62 GSAA and then posted a nice 3.89 number the following year in 20 starts. Hopefully we’re getting his first two seasons, and not the last one.

RJ
As always, great work breaking down a complex equation into its component parts then putting them back together again in a way that many new fans will be able to understand.
As far as Fluery, I’m pretty okay with him being an average goalie over the two seasons he will be in town. He is a big name, he is great with fans, he has tons of great experience, and he can clearly still play at an adequate NHL level. Putting an elite goalie behind this line up wouldn’t help too much, and there wasn’t an elite goalie available anyway.
I am curious what the goalie plan is for year three on. Does GMGM think Fluery is going to get better with age and sign the guy to another deal? Is Calvin Pickard our playoff goalie of the future? Did we get the steal of the draft with Maxim Zhukov? I guess that’s a question for another post.
James
@RJ
‘I am curious what the goalie plan is for year three on. Does GMGM think Fluery is going to get better with age and sign the guy to another deal? Is Calvin Pickard our playoff goalie of the future? Did we get the steal of the draft with Maxim Zhukov? I guess that’s a question for another post.’
Goalies are Voodoo. It’s tough to make goalie predictions, especially about the future. The answer may be “none of the above” ..
James
@Jason Pothier
For those who are baseball fans, this is essentially WAR (wins above replacement) for goalies.
How many wins above replacement is Sergei Bobrovsky? Unless I’m missing something, Goals Saved Above Average doesn’t tell me.
I wasn’t familiar with Goals Saved Above Average. GSAA kind of reminds me of PER. The player efficiency rating (PER) is John Hollinger’s all-in-one basketball rating, which attempts to boil down all of a player’s contributions into one number.
The GSAA numbers would mean more to me if there was a Reference guide. For example:
All-time great season 35.0+
Runaway MVP candidate 30.0-35.0
Strong MVP candidate 27.5-30.0
Weak MVP candidate 25.0-27.5
Definite All-Star 22.5-25.0
Borderline All-Star 20.0-22.5
Second offensive option 18.0-20.0
Third offensive option 16.5-18.0
Slightly above-average player 15.0-16.5
Rotation player 13.0-15.0
Non-rotation player 11.0-13.0
Fringe roster player 9.0-11.0
Player who won’t stick in the league 0-9.0