Ranking the top-40 goaltenders by Goals Allowed Per Expected Goal Against

[Captain Creampuff]

I like this math because it shows that Jake Allen sucks at hockey

 

[TomasHertlsRooster]

@Bleedred do you happen to have the stoppable goals data handy? I'd like to see whether or not there is a correlation between stoppable goals percentage and goals allowed per expected goal against, and just how strong that correlation is.

 

[Bleedred]

@Bleedred do you happen to have the stoppable goals data handy? I'd like to see whether or not there is a correlation between stoppable goals percentage and goals allowed per expected goal against, and just how strong that correlation is.

Definitely will get it for you later. Have it on files at home but am at work at the moment.

 

[Lshap]

I thought you left?

Anyway, Halak had the best save percentage and goals against on the Islanders that year.

Garth Snow was quoted as saying that he did it to spark the team. When the Islanders desperately needed wins at the end of the season to make the playoffs, he called up Halak. The Islanders beat described the incident as bizarre and everyone involved in it was fired in the offseason.

Halak being bumped up by this model illustrates one of its limits. I'm very familiar with Halak. He's a great 'relief pitcher' who can perform very well in regulated doses. However, he can't handle the workload of a #1 starter. When given too many games over too long a period, his game crashes and he's plain awful.

This model doesn't prove Halak is a superior goalie; it plays to his strength as an erratically effective temp employee.

 

[romba]

HF: Please! I KNOW Rask and Vasi are the shit because *reputations*, f*** this list.
Data: Not so fast.
HF: YOU'RE SPREADSHEETS MUST BE WRONG OTHERWISE IT MEANS I MIGHT BE.

 

[Machinehead]

Halak being bumped up by this model illustrates one of its limits. I'm very familiar with Halak. He's a great 'relief pitcher' who can perform very well in regulated doses. However, he can't handle the workload of a #1 starter. When given too many games over too long a period, his game crashes and he's plain awful.

This model doesn't prove Halak is a superior goalie; it plays to his strength as an erratically effective temp employee.

Well would I call Halak the 12th best goalie in the league? Probably not.

But he does get unnecessarily crucified at times.

For example, he's better -right now- than Lundqvist who a lot of people still think is a top 10 goalie.

 

[Bleedred]

@Bleedred do you happen to have the stoppable goals data handy? I'd like to see whether or not there is a correlation between stoppable goals percentage and goals allowed per expected goal against, and just how strong that correlation is.

It's cluttered and everything. I don't have the total number of goals against that every goalie allowed this year, just the number of goals I counted on them that I thought they could have or should have had. There were some deviations in save percentage, but I did feel like some of the more poor goalies in save percentage, also allowed poor goals on my eye test and vice versa. Fleury and Gibson both had a low percentage of goals I counted stoppable, yet their save percentages weren't amazing. Aaron Dell had a very low percentage, but I think this was a usage thing. And as I've said on the Sharks forum, I think there's a lot of goals that a guy like Aaron Dell allows that don't really look stoppable, but hit many of the goalies in the league because they're bigger and more technically sound.


Luongo 34
Brossoit 14
Kuemper 33
Varlamov 39
Halak 22
Rittich 33
Gibson 35
Dubnyk 46
Ullmark 35
Vasilevskiy 33
Greiss 23
DeSmith 22
Miller 12
Rinne 35
Lehner 20
Raanta 10
Bishop 24
Dell 16
Andersen 41
Saros 19
Lundqvist 45
Grubauer 23
Price 40
Howard 43
Campbell 11
Nilsson 28
Markstrom 46
Hutton 38
Kinkaid 41
Budaj
Crawford 31
Anderson 49
Hellebuyck 51
Talbot 34
Darling 12
Jones 53
Rask 30
Khudobin 23
Niemi 19
Fleury 36
Stalock 15
Murray 34
Sparks 20
McElhinney 19
Ward 37
Elliott 21
Koskinen 53
Holtby 44
Domingue 21
Mrazek 25
Georgiev 24
Copley 26
Bobrovsky 41
Subban 20
Korpisalo 17
Reimer 28
Smith 42
Allen 35
Johnson 16
Bernier 27
Quick 48
Hutchinson 10
Schneider 22
Pickard 16
Condon 3
Neuvirth 8
McKenna 13
Miska 1
Bachman 1
Petersen 6
Lyon 2
Jarry 2
Stolarz 16
Hill 7
Pasquale 3
Gibson 1
Binnington 12
Hart 23
Blackwood 15
Delia 13
Francouz 1
Hogberg 3
Demko 9
Nedeljkovic
Legace 2
Boyle 4
DiPietro 2
Bow
Montembeault 9
Comrie 3
Daccord 2
Lindgren 2
Fulcher 1

This is every goalie that played last year that allowed at least one goal. If there's no number next to their name, it means I didn't count any goals against them as stoppable. I think Budaj and Bow are the only two goalies without a number after their name. Nedeljkovic didn't have any either.

I have a separate count I did for the playoffs and Bobrovsky and Bishop both had very high percentages, despite very good save percentages.

I noticed Carter Hutton didn't look that bad, but Ullmark was terrible and I swear he probably led the league in most wraparound goals against. The disparity in save percentages between the two wasn't as great as the disparity between the stoppable goal percentages that I recorded. Pheonix Copley also allowed a significantly high percentage of goals that I counted stoppable, while Holtby did not. Holtby's save percentage was higher, but not to the degree it was on my eye test.

 

[Bleedred]

Halak being bumped up by this model illustrates one of its limits. I'm very familiar with Halak. He's a great 'relief pitcher' who can perform very well in regulated doses. However, he can't handle the workload of a #1 starter. When given too many games over too long a period, his game crashes and he's plain awful.

This model doesn't prove Halak is a superior goalie; it plays to his strength as an erratically effective temp employee.

I agree with you on Halak.

He's a good 1b option, but I don't think he'd maintain these kinds of numbers as a clear cut #1 starter. He's had a couple seasons where he did, but I don't think that'll happen anymore, if you were to give him that role.

 

[DistantThunderRep]

HF: Please! I KNOW Rask and Vasi are the **** because *reputations*, **** this list.
Data: Not so fast.
HF: YOU'RE SPREADSHEETS MUST BE WRONG OTHERWISE IT MEANS I MIGHT BE.

When your spreadsheets say one thing and another one says another then yes, situations where discussions happen. It's like you are absolutely baffled at the concept of a discussion of minds and opinions are happening on a discussion board.

 

[bsu]

LOL!!! This is just full of crap....

-- Bob Mackenzie #Hey guys, who do you think is the best goaltender in the NHL?

-- NHL Players: Carey Price(79%)

-- This article writers and HF posters : no guys Carey Price is probably the worst starter in the NHL!

-- NHL Players: Hahahahahaha good job clowns!!!


Rask Price Bishop and Vasi are the best goaltenders in the world and everybody who watch hockey knows this, Sorry folks.

Gibson is.

 

[Gm0ney]

It's cluttered and everything. I don't have the total number of goals against that every goalie allowed this year, just the number of goals I counted on them that I thought they could have or should have had. There were some deviations in save percentage, but I did feel like some of the more poor goalies in save percentage, also allowed poor goals on my eye test and vice versa. Fleury and Gibson both had a low percentage of goals I counted stoppable, yet their save percentages weren't amazing. Aaron Dell had a very low percentage, but I think this was a usage thing. And as I've said on the Sharks forum, I think there's a lot of goals that a guy like Aaron Dell allows that don't really look stoppable, but hit many of the goalies in the league because they're bigger and more technically sound.


Luongo 34
Brossoit 14
Kuemper 33
Varlamov 39
Halak 22
Rittich 33
Gibson 35
Dubnyk 46
Ullmark 35
Vasilevskiy 33
Greiss 23
DeSmith 22
Miller 12
Rinne 35
Lehner 20
Raanta 10
Bishop 24
Dell 16
Andersen 41
Saros 19
Lundqvist 45
Grubauer 23
Price 40
Howard 43
Campbell 11
Nilsson 28
Markstrom 46
Hutton 38
Kinkaid 41
Budaj
Crawford 31
Anderson 49
Hellebuyck 51
Talbot 34
Darling 12
Jones 53
Rask 30
Khudobin 23
Niemi 19
Fleury 36
Stalock 15
Murray 34
Sparks 20
McElhinney 19
Ward 37
Elliott 21
Koskinen 53
Holtby 44
Domingue 21
Mrazek 25
Georgiev 24
Copley 26
Bobrovsky 41
Subban 20
Korpisalo 17
Reimer 28
Smith 42
Allen 35
Johnson 16
Bernier 27
Quick 48
Hutchinson 10
Schneider 22
Pickard 16
Condon 3
Neuvirth 8
McKenna 13
Miska 1
Bachman 1
Petersen 6
Lyon 2
Jarry 2
Stolarz 16
Hill 7
Pasquale 3
Gibson 1
Binnington 12
Hart 23
Blackwood 15
Delia 13
Francouz 1
Hogberg 3
Demko 9
Nedeljkovic
Legace 2
Boyle 4
DiPietro 2
Bow
Montembeault 9
Comrie 3
Daccord 2
Lindgren 2
Fulcher 1

This is every goalie that played last year that allowed at least one goal. If there's no number next to their name, it means I didn't count any goals against them as stoppable. I think Budaj and Bow are the only two goalies without a number after their name. Nedeljkovic didn't have any either.

I have a separate count I did for the playoffs and Bobrovsky and Bishop both had very high percentages, despite very good save percentages.

I noticed Carter Hutton didn't look that bad, but Ullmark was terrible and I swear he probably led the league in most wraparound goals against. The disparity in save percentages between the two wasn't as great as the disparity between the stoppable goal percentages that I recorded. Pheonix Copley also allowed a significantly high percentage of goals that I counted stoppable, while Holtby did not. Holtby's save percentage was higher, but not to the degree it was on my eye test.

Does your goals-they-should've-had count cover the entire 2018-19 season? If so...removing < 1000 minute goalies, expressed as a rate, and sorted on Bad Goals Against per 60 minutes:

Goalie	Bad GA	Mins	BGA/60
Binnington 12	12	1876	0.384
Campbell 11	11	1593	0.414
Lehner 20	20	2616	0.459
Bishop 24	24	2637	0.546
Halak 22	22	2308	0.572
McElhinney 19	19	1978	0.576
Fleury 36	36	3635	0.594
Greiss 23	23	2294	0.602
Kuemper 33	33	3251	0.609
Vasilevskiy 33	33	3204	0.618
Price 40	40	3880	0.619
Khudobin 23	23	2219	0.622
Mrazek 25	25	2387	0.628
Miller 12	12	1109	0.649
Gibson 35	35	3233	0.650
Rinne 35	35	3220	0.652
Saros 19	19	1696	0.672
DeSmith 22	22	1944	0.679
Grubauer 23	23	2021	0.683
Rask 30	30	2635	0.683
Bobrovsky 41	41	3557	0.692
Andersen 41	41	3510	0.701
Murray 34	34	2880	0.708
Blackwood 15	15	1264	0.712
Dubnyk 46	46	3855	0.716
Brossoit 14	14	1165	0.721
Dell 16	16	1323	0.726
Korpisalo 17	17	1361	0.749
Markstrom 46	46	3599	0.767
Georgiev 24	24	1874	0.768
Holtby 44	44	3407	0.775
Rittich 33	33	2503	0.791
Hutton 38	38	2840	0.803
Hart 23	23	1717	0.804
Domingue 21	21	1561	0.807
Allen 35	35	2568	0.818
Nilsson 28	28	2040	0.824
Varlamov 39	39	2839	0.824
Hellebuyck 51	51	3704	0.826
Crawford 31	31	2213	0.840
Howard 43	43	3053	0.845
Stalock 15	15	1065	0.845
Luongo 34	34	2347	0.869
Bernier 27	27	1860	0.871
Lundqvist 45	45	3089	0.874
Jones 53	53	3597	0.884
Elliott 21	21	1397	0.902
Reimer 28	28	1805	0.931
Schneider 22	22	1371	0.963
Subban 20	20	1227	0.978
Ullmark 35	35	2102	0.999
Copley 26	26	1529	1.020
Smith 42	42	2400	1.050
Anderson 49	49	2785	1.056
Koskinen 53	53	2992	1.063
Kinkaid 41	41	2302	1.069
Talbot 34	34	1905	1.071
Sparks 20	20	1105	1.086
Quick 48	48	2648	1.088
Ward 37	37	1882	1.180

[TBODY] [/TBODY]

 

[Bleedred]

Does your goals-they-should've-had count cover the entire 2018-19 season? If so...removing < 1000 minute goalies, expressed as a rate, and sorted on Bad Goals Against per 60 minutes:

Goalie	Bad GA	Mins	BGA/60
Binnington 12	12	1876	0.384
Campbell 11	11	1593	0.414
Lehner 20	20	2616	0.459
Bishop 24	24	2637	0.546
Halak 22	22	2308	0.572
McElhinney 19	19	1978	0.576
Fleury 36	36	3635	0.594
Greiss 23	23	2294	0.602
Kuemper 33	33	3251	0.609
Vasilevskiy 33	33	3204	0.618
Price 40	40	3880	0.619
Khudobin 23	23	2219	0.622
Mrazek 25	25	2387	0.628
Miller 12	12	1109	0.649
Gibson 35	35	3233	0.650
Rinne 35	35	3220	0.652
Saros 19	19	1696	0.672
DeSmith 22	22	1944	0.679
Grubauer 23	23	2021	0.683
Rask 30	30	2635	0.683
Bobrovsky 41	41	3557	0.692
Andersen 41	41	3510	0.701
Murray 34	34	2880	0.708
Blackwood 15	15	1264	0.712
Dubnyk 46	46	3855	0.716
Brossoit 14	14	1165	0.721
Dell 16	16	1323	0.726
Korpisalo 17	17	1361	0.749
Markstrom 46	46	3599	0.767
Georgiev 24	24	1874	0.768
Holtby 44	44	3407	0.775
Rittich 33	33	2503	0.791
Hutton 38	38	2840	0.803
Hart 23	23	1717	0.804
Domingue 21	21	1561	0.807
Allen 35	35	2568	0.818
Nilsson 28	28	2040	0.824
Varlamov 39	39	2839	0.824
Hellebuyck 51	51	3704	0.826
Crawford 31	31	2213	0.840
Howard 43	43	3053	0.845
Stalock 15	15	1065	0.845
Luongo 34	34	2347	0.869
Bernier 27	27	1860	0.871
Lundqvist 45	45	3089	0.874
Jones 53	53	3597	0.884
Elliott 21	21	1397	0.902
Reimer 28	28	1805	0.931
Schneider 22	22	1371	0.963
Subban 20	20	1227	0.978
Ullmark 35	35	2102	0.999
Copley 26	26	1529	1.020
Smith 42	42	2400	1.050
Anderson 49	49	2785	1.056
Koskinen 53	53	2992	1.063
Kinkaid 41	41	2302	1.069
Talbot 34	34	1905	1.071
Sparks 20	20	1105	1.086
Quick 48	48	2648	1.088
Ward 37	37	1882	1.180

[TBODY] [/TBODY]

Yes.

This is for the entire 18-19 regular season. I have a separate count for the playoffs where I reset everything from the regular season.

 

[VoluntaryDom]

@JoeThorntonsRooster are you using all situations? because vasy is 4th in the league in GSAA on natural stat trick in the 3 year sample and 8th in GSAA/60 (3000 minute minimum) at 5v5. i do think NST's model might overrate the slot and the rush a tad and thus i dont have vasy top 8 but your metric showing him as below average by NST numbers is...strange

 

[VoluntaryDom]

by NST at 5v5, 2 year sample with 2000 min minimum vasy is 5th in GSAA/60 and 2nd in total GSAA. honestly no idea what you did @JoeThorntonsRooster

 

[TomasHertlsRooster]

@JoeThorntonsRooster are you using all situations? because vasy is 4th in the league in GSAA on natural stat trick in the 3 year sample and 8th in GSAA/60 (3000 minute minimum) at 5v5. i do think NST's model might overrate the slot and the rush a tad and thus i dont have vasy top 8 but your metric showing him as below average by NST numbers is...strange

by NST at 5v5, 2 year sample with 2000 min minimum vasy is 5th in GSAA/60 and 2nd in total GSAA. honestly no idea what you did @JoeThorntonsRooster

Yes, I used all situations.

The way that Natural Stat Trick calculates GSAA does not factor in xGA. It only factors in SV% relative to the league average SV%.

 

[VoluntaryDom]

Yes, I used all situations.

The way that Natural Stat Trick calculates GSAA does not factor in xGA. It only factors in SV% relative to the league average SV%.

ah ok that makes more sense