HOH Top 60 Goaltenders of All Time (2024 Edition) - Round 2, Vote 1

[nabby12]

In his first and only full season with Boston they didn't make the playoffs. The next year I guess he got hurt and the unreliable Don Simmons put up similar numbers and somehow got them to the Final (losing to Montreal).

I guess one guy that might think a little less of Sawchuk is Jack Adams or whoever was running the Wings in '55. Is it well known why that deal (blockbuster as it was) actually went down...?

Sawchuk was very down and depressed the entire time he was with Boston. He even retired from hockey due to mononucleosis during the second season which is why Simmons was playing.

Detroit eventually brought him back out of retirement, but the damage was done. The initial trade to Boston was a massive shock and the worst day in his life.

Jack Adams thought that goalies had a short shelf life, and with Glenn Hall waiting in the wings, he decided to trade Sawchuk as he did with Lumley a few years back. Before he died, Adams regretted trading Sawchuk and called it the biggest mistake of his managing career and noted how it cost his team multiple Stanley Cups.

 

[ContrarianGoaltender]

In his first and only full season with Boston they didn't make the playoffs. The next year I guess he got hurt and the unreliable Don Simmons put up similar numbers and somehow got them to the Final (losing to Montreal).

I guess one guy that might think a little less of Sawchuk is Jack Adams or whoever was running the Wings in '55. Is it well known why that deal (blockbuster as it was) actually went down...?

I'm reading Todd Denault's book on Jacques Plante right now and just went past that part. Here's his quote straight from Jack Adams himself:

"We let Sawchuk go because we found ourselves with two top goalies. Hall is more advanced now than Sawchuk was when he joined us and all the players insist Glenn has been NHL material for the past year...It was a case of trading one of them and Sawchuk is the established player. Consequently, he brought a better offer."

Just going to also echo a few others here that I think Sawchuk is the most interesting guy this round and the one with the most room to move for me.

 

[jigglysquishy]

Gordie Howe on the trade

"Hall was a terrific goaltender, no question, but trading Sawchuk, who was coming off another Vezina Trophy, was hard to swallow. Most general managers spent their entire career waiting for a goalie like Sawchuk, who they could build a team around, to come along. Not Mr. Adams. When Terry was on his game, it's hard to think of anyone better. He also always seemed to save his best for when the stakes were the highest. An NHL team can't ask for much more than having a goalie who heats up every spring "

 

[TANK200]

Quoting myself on Hasek vs Roy:

Here's how Roy and Hasek rank in Goals Versus Average (which combines save percentage with workload):

• Roy - 1st (1989), 1st (1990), 1st (1992), 2nd (1988), 2nd (1991), 2nd (2002), 3rd (1987), 3rd (1994), 3rd (1997), 5th (1998), 6th (2000), 6th (2003), 7th (1993), 7th (1996), 7th (1999), 10th (1995)
• Hasek - 1st (1995), 1st (1996), 1st (1997), 1st (1998), 1st (1999), 2nd (2001), 2nd (1994), 7th (2006), 8th (2000), 10th (2002)

If we cancel out similar finishes (and I'll also cancel out a 7th for Roy with an 8th for Hasek), and we're left with:

• Roy - 2nd, 3rd, 3rd, 3rd, 5th, 6th, 6th, 7th
• Hasek - 1st, 1st

There's a much wider range in how goalies perform from year to year, compared to skaters. Roy brought enormous value to his teams by being consistently excellent year after year. From 1987 to 2003 (a span of 17 seasons), he placed lower than 7th in GVA only twice. I'm not saying that's the same as Gordie Howe with 20 straight years in the top five in scoring, or Ray Bourque with 19 out of 22 seasons placing in the top five in Norris voting - but it's not far off either. Would you trade someone with two scoring titles and little else (ie Dickie Moore) for someone who was a top 7 scorer eight times (ie Joe Sakic)?

Just to show Roy's consistency. From 1980 to 2022, here's how many times each goalie placed 7th or higher in GVA (granted an arbitrary cutoff) - Roy did it 15 times. Hasek and Luongo are next at 8 times. Lundqvist and Vanbiesbrouck did it 7 times. Brodeur 6 times. Then Joseph, Barrasso, Hrudey and Moog five times each. It's a very tough standard (Belfour and Price don't even qualify).

You can say I'm being disingenuous because Hasek clearly peaked higher. And that's true. Hasek has all five of the highest seasons in GVA between them (then Roy has 9 of the next 12). Can you argue that Hasek's peak trumps anything Roy ever did? Possibly. If the question is who's the "better" or more naturally talented goalie, I think it's almost impossible to choose Roy.

But if the question is "greatest" (which is generally how we make our top X lists), I'd go with Roy. Hasek has the higher peak, but Roy ends up ahead in career GVA. He was much more consistent from year to year (which is rare for a goalie). He still led the league in GVA three times (so he wasn't a compiler - and nobody else from 1980 onwards has done this more than twice). And Roy's playoff resume is vastly superior (where, adjusted for era, he stopped the puck more effectively than Hasek, over a much larger number of games).

Roy's consistent excellence relative to his peers is certainly impressive, but it is important to consider that Hasek was behind the Iron Curtain at the start of his career. Before playing a singly game in the NHL, he had already accomplished the following:

• World Championship Best Goalie: x2
• World Championship All-Star Team: x3
• Czechoslovakia Goalie of the Year: x5 (consecutively)
• Czechoslovakia Best Player: x2

It's almost certain that he would have dominated the NHL from an earlier age if he was born in North America. And when he eventually came to the NHL, he was battling for a starting position with an already-established Eddie Belfour.

 

[Professor What]

No matter where I vote Sawchuk, I'm sympathetic to him. Severe depression is horrible to deal with. If anything, I respect him for overcoming it to the degree he did, even if it almost certainly contributed to his end.

 

[MadArcand]

Shot undercounting at Devils' home games plays a large part in the SV% data. Brodeur's H/R split from 1999-2004 is .904/.914 in the RS and .913/.920 in the PO. There were 3.83 fewer shots in Devils' RS home games than in their road games over the time span, and it actually climbs to 5.63 in the PO. S% H/R splits for both teams combined are 9.75/8.85 RS and 9.02/8.04 PO.

Adding 410 shots/saves (3.83/2*215) to Brodeur's home totals raises his SV% to.912 at home and .913 overall.

Roy played with 2 Hart winners, 3 Selke winners, and 3 1st team all-star defensemen in his career. Brodeur played with 2 1st team all-star defensemen and 1 Selke winner in his career. Why is Brodeur the one knocked for playing on a stacked team?

A random question, but do we have the Devils' shots for and shooting percentage home/road splits for the timeframe as well? And how do they compare to other teams' home/road splits?

 

[Dennis Bonvie]

Very excited to get this project started.

The first 8 up is the "canon" 8. How confident are we that the top 8 should be the same top 8 as 12 years ago? It's 6 birthyears representing 3 eras. I guess we will wait until another name comes up.

The only things I'm confident on are Dryden last and Plante first of the O6 guys. Other than that I'm open to arguments for anything.

I'll be voting Hasek first, but I think it'll be close.

Going strictly by contemporary opinion, Hasek is king. That multiyear run where the public consensus was that he was the best player in the world (or close to it) is hard to beat.

I recently rewatched the CSSR/Canada semi final game from 1987. Hasek is a baby but does a good job of dealing with the Canadian onslaught. For being 7 years pre NHL, he looks really good.

Dryden wasn't an O6 guy. He started after expansion.

That said, Dryden is kind of a Bill Durnan with more rings. Very short career, but very impressive.

Yes, I realize who he played for. But 6 Cups in 8 seasons means he must have done his job for them. Consider the Canadiens didn't win the year before or the year after Dryden played. And they didn't win the year he sat out (1973-74).

 

[jigglysquishy]

Dryden wasn't an O6 guy. He started after expansion.

That said, Dryden is kind of a Bill Durnan with more rings. Very short career, but very impressive.

Yes, I realize who he played for. But 6 Cups in 8 seasons means he must have done his job for them. Consider the Canadiens didn't win the year before or the year after Dryden played. And they didn't win the year he sat out (1973-74).

Ya, the sentence is worded poorly but it's two separate thoughts. Dryden is last of the 8 for me. Plante is ahead of Sawchuk/Hall.

 

[Dennis Bonvie]

Ya, the sentence is worded poorly but it's two separate thoughts. Dryden is last of the 8 for me. Plante is ahead of Sawchuk/Hall.

I've always felt Hall gets undervalued.

 

[Professor What]

I've always felt Hall gets undervalued.

I feel like Hall deserves to be higher than fifth, but if we voted right now, he'd be fifth for me.

 

[ContrarianGoaltender]

So we're getting into the weeds again of some well-worn debate territory, but I wanted to bring up some general discussion topics that are not only relevant to the discussion, but might potentially help people think about their rankings and what they value.

First up is something that has changed significantly even since we did the last goalie list on HOH: Goalie Workloads.

Braden Holtby and Jonathan Quick started 72 and 71 games in 2014-15. Since then, Cam Talbot in 2016-17 is the only guy with 70 or more starts, and I wonder whether he might just be the last goalie to ever start 70 games in a season in the NHL. In the last two seasons, nobody even played in 65 games. Pretty much everyone in hockey has now come to an agreement on the fact that load management for goalies is the objectively correct decision, and that having a goalie that plays 70 games in a season isn't a virtue, it's a mistake.

Some of the reasons that this is true:

- Playing lots of games increases wear and tear and injury risk

- A starter who plays a lot of games is also necessarily taking a greater share of the easy games, in which he's less likely to have a significant impact on win probability

- There are fatigue effects, especially in the playoffs. There appears to be a pretty strong negative correlation between regular season workload and playoff success. Only two goalies since the 2005 lockout have won the Cup while playing at least 60 regular season games, for example, and many elite goalies have pretty mediocre playoff records in their high-workload seasons.

As a result of this development, I don't see any reason to give goalies credit for a high number of games. In my current evaluation system, I don't give anyone credit for more than 65 games played, because I don't think that represents additional value to a team.

Also, I think we should pay attention to a goalie's workload and how that may have impacted their results, whether it was padding their stats from lots of games, or being someone who was forced to play more games than they should have despite physical concerns, or benefiting from a more optimal usage which helped set them up for success.

2. Non-NHL Value

I think it is very important to consider relevant non-NHL performance. There were some good discussions in the preliminary thread about the difficulty of getting a goalie job in the Original 6 pre-1965, and that shouldn't be forgotten.

Jacques Plante, for example, turned pro with the Montreal Royals in 1949-50 in his age 21 season, and already in that very first year he was getting paid to be the practice goalie for the Montreal Canadiens and was clearly viewed as third in line on their organizational depth chart, behind Durnan and McNeil. I'm quite confident he would have been an NHLer in any post-expansion scenario at that age, but he came up in the one-goalie era on the team with the best goaltending depth in the league at a time when teams completely controlled their players' destinities, and as a result he had to wait 5 years to gain the starting job.

For the sake of comparison:

10 goalies played in the NHL in Terry Sawchuk's first season as a starting goalie.
41 played in Ken Dryden's first season as a starting goalie.
80 played in Martin Brodeur's first season as a starting goalie.

Taking everything into account, it's far from obvious to me that Martin Brodeur has more elite longevity than Jacques Plante.

And has already been pointed out in this thread, comparing Hasek to Roy without factoring in Europe at all is simply an incomplete analysis of the relevant resumes of those two goaltenders.

I'm certainly not saying that everything matters and we should consider ECHL seasons or junior careers or count every time somebody retired to go play semi-pro in Europe. It's fine to have a more exclusive NHL focus after the world opened up and the best talent worldwide had an open door to the NHL (say, anybody starting their careers in the late '90s or later). But if we're trying to fairly compare across eras and adjust for different team environments, we should recognize the relative difficulty of goalie paths to the NHL as well, and give proper credit for those performances.

 

[Professor What]

So we're getting into the weeds again of some well-worn debate territory, but I wanted to bring up some general discussion topics that are not only relevant to the discussion, but might potentially help people think about their rankings and what they value.

First up is something that has changed significantly even since we did the last goalie list on HOH: Goalie Workloads.

Braden Holtby and Jonathan Quick started 72 and 71 games in 2014-15. Since then, Cam Talbot in 2016-17 is the only guy with 70 or more starts, and I wonder whether he might just be the last goalie to ever start 70 games in a season in the NHL. In the last two seasons, nobody even played in 65 games. Pretty much everyone in hockey has now come to an agreement on the fact that load management for goalies is the objectively correct decision, and that having a goalie that plays 70 games in a season isn't a virtue, it's a mistake.

Some of the reasons that this is true:

- Playing lots of games increases wear and tear and injury risk

- A starter who plays a lot of games is also necessarily taking a greater share of the easy games, in which he's less likely to have a significant impact on win probability

- There are fatigue effects, especially in the playoffs. There appears to be a pretty strong negative correlation between regular season workload and playoff success. Only two goalies since the 2005 lockout have won the Cup while playing at least 60 regular season games, for example, and many elite goalies have pretty mediocre playoff records in their high-workload seasons.

As a result of this development, I don't see any reason to give goalies credit for a high number of games. In my current evaluation system, I don't give anyone credit for more than 65 games played, because I don't think that represents additional value to a team.

Also, I think we should pay attention to a goalie's workload and how that may have impacted their results, whether it was padding their stats from lots of games, or being someone who was forced to play more games than they should have despite physical concerns, or benefiting from a more optimal usage which helped set them up for success.

I think I agree with this for the most part, but I think there are exceptions. Martin Brodeur is one that comes to mind. To me, he's better than his numbers say, and one big reason is that you could ride that horse all season long and into the playoffs, and he didn't break down. A guy that can handle that kind of workload without the ill effects and still be a key contributor to Cup wins deserves credit for that. A guy that breaks down and has limited playoff success after a heavy workload, much less so. That's one of the big differences between Brodeur and Esposito for me. Esposito would play almost all the games, but then one can't help but wonder if his playoff record would be better had he been platooned.

 

[Dr John Carlson]

So we're getting into the weeds again of some well-worn debate territory, but I wanted to bring up some general discussion topics that are not only relevant to the discussion, but might potentially help people think about their rankings and what they value.

First up is something that has changed significantly even since we did the last goalie list on HOH: Goalie Workloads.

Braden Holtby and Jonathan Quick started 72 and 71 games in 2014-15. Since then, Cam Talbot in 2016-17 is the only guy with 70 or more starts, and I wonder whether he might just be the last goalie to ever start 70 games in a season in the NHL. In the last two seasons, nobody even played in 65 games. Pretty much everyone in hockey has now come to an agreement on the fact that load management for goalies is the objectively correct decision, and that having a goalie that plays 70 games in a season isn't a virtue, it's a mistake.

Some of the reasons that this is true:

- Playing lots of games increases wear and tear and injury risk

- A starter who plays a lot of games is also necessarily taking a greater share of the easy games, in which he's less likely to have a significant impact on win probability

- There are fatigue effects, especially in the playoffs. There appears to be a pretty strong negative correlation between regular season workload and playoff success. Only two goalies since the 2005 lockout have won the Cup while playing at least 60 regular season games, for example, and many elite goalies have pretty mediocre playoff records in their high-workload seasons.

As a result of this development, I don't see any reason to give goalies credit for a high number of games. In my current evaluation system, I don't give anyone credit for more than 65 games played, because I don't think that represents additional value to a team.

Also, I think we should pay attention to a goalie's workload and how that may have impacted their results, whether it was padding their stats from lots of games, or being someone who was forced to play more games than they should have despite physical concerns, or benefiting from a more optimal usage which helped set them up for success.

That's definitely something I had to wrestle with on my initial list, and will become an even bigger factor once we start getting into tandem goalies.

I've sort of reevaluated how I consider award voting for goalies, particularly Hart voting, when taking into consideration games played. Jacques Plante led the league in GP a single time - coincidentally, his Hart season. Terry Sawchuk led in GP a few more times, and received a bit more Hart consideration. And then Glenn Hall routinely led the league in GP, and basically lived on the Hart ballot.

I wonder if there's something to be done in looking at a ratio of games played to Hart voting. Which guys' MVP support is tied closely to games played, and which are tied less closely? Patrick Roy had top-5 Hart finishes in years in which he had the 15th and 12th most games played among goalies... Hasek nearly won in a year where he finished 14th in games played, not bad... obviously the O6 goalies wouldn't be able to finish that low in GP, but still, there's surely a way to account for era here.

 

[ContrarianGoaltender]

3. Goalie Talent Pool

We should think in a serious way about the depth of talent pools, sources of talent, and how that has impacted goaltending over the years.

Here's how the number of goalies in the league hitting various games played cutoffs and the percentage of them that were born in Canada changed over time (picking years ending in 9 rather than a nice round 0, just because that gives us Dryden's last season, a Roy peak season, a Hasek peak season, and just misses Brodeur's last Vezina):

Year	1979​	1979​	1989​	1989​	1999​	1999​	2009​	2009​
Games Played	All	Canada	All	Canada	All	Canada	All	Canada
>=1 GP	50​	49​	68​	59​	85​	62​	89​	43​
20+ GP	32​	32​	35​	41​	46​	29​	53​	25​
40+ GP	14​	14​	20​	16​	25​	12​	33​	15​

Converting everything to percentages:

	1979​	1989​	1999​	2009​
Games Played	% Canadian	% Canadian	% Canadian	% Canadian
>=1 GP	98%​	87%​	73%​	48%​
20+ GP	100%​	85%​	63%​	47%​
40+ GP	100%​	80%​	48%​	45%​

I think the competitive environment changed a whole lot over those 40 years, and the international goalie talent pool in particular exploded. I tried to adjust for this in putting together my list, but I think if anything I didn't go heavy enough on modern guys. This is one of the reasons why Ken Dryden didn't make the top 8 on my initial list for this project, and why I continue to argue that Hasek had the best peak in goaltending history. It is also why you should be very careful with comparing league averages across time.

I think the current terrible state of Canadian goaltending is also an interesting point to consider. Canadian goalies have combined for one Vezina third-place vote in the last three seasons combined, but if there were only Canadian goalies in the league, one of them would be winning the Vezina. I think it also shows that talent pools can swing pretty widely, especially if we're only looking at the top end at one fairly niche position like goaltending (and for much of history, it was also mostly one country supplying all the relevant talent).

4. Coaching Effects

One factor that I'm really focusing on this time around is coaching. I'm looking at the career defensive stats for coaches to try to determine which ones had the largest impact on their netminders' success, and using that as part of my estimate of the team effects on each goalie's numbers.

It's also interesting to note goalies who had standout success for one coach vs. many. Obviously everyone in this round was elite for the vast majority of their careers, but not all of them were able to win awards for multiple coaches. For example, here's the current pool (Tretiak excluded) with the number of different coaches they have had for a 1AST team (pre-1982) or Vezina (post-1982) season:

Glenn Hall (4): Jimmy Skinner, Rudy Pilous*, Billy Reay, Scotty Bowman
Dominik Hasek (3): John Muckler, Ted Nolan, Lindy Ruff
Martin Brodeur (3): Pat Burns, Jacques Lemaire, Brent Sutter
Ken Dryden (1): Scotty Bowman
Patrick Roy (1): Pat Burns
Terry Sawchuk (1): Tommy Ivan
Jacques Plante (1): Toe Blake

(*-Hall had the 1957-58 season that was split between Pilous and Tommy Ivan, but I'm only counting the coach with the most games in each season for simplicity's sake)

The top three on that list had success in a few different coaching environments, although Hall was the only one to accomplish this for different franchises (and he impressively did it for 3 of them). Everybody knows about Sawchuk's five-year peak which came in one consistent coaching environment, while Dryden and Plante both had one coach for almost their entire tenures in Montreal. Roy is the one that really stands out here, given that he had Pat Burns for only 4 of his 19 seasons, but those 4 years accounted for all 3 of his Vezinas and one of his Vezina runner-up finishes, while he went 2-3-3-4-5-5 in Vezina voting over the rest of his career. Which raises the obvious question, how much of that great 1989-1992 run was Roy peaking, and how much of it was Burns?

Similarly, how many of these guys faced coaching situations which might have negatively impacted their stats, awards, or team success? For this round, I'm particularly interested in trying to get a good defensive estimate for Sawchuk's Boston and Detroit 2.0 years.

 

[Michael Farkas]

There are fatigue effects, especially in the playoffs.

Porting this over from the round 1 discussion...

Guys that won often/a lot:
Plante
Bower
Sawchuk
Parent
Smith
Dryden

Guys that famously did not win often/a lot:
Hall
Vachon
Giacomin
Esposito
Crozier
Worsley until he was in his late 30's on a wagon Habs team that leaked into the expansion era

Very clear divide here. The bottom group were butterfly guys and/or guys that threw themselves all over the floor indiscriminately.

Brodeur made the Final with 73, 72, 40 of 48 (~69 normalized to 82), etc. games played. Not a butterfly guy, not a nonsense goalie.

Hasek maxed at 64 and 65 to make the Final. Roy: 62, 61, 62. Above that, it's two first round losses and a second round sweep.

So maybe there's smoke here in terms of the workload and the style. Given how popular the butterfly is these days, we see that maybe there's something to limiting games...

 

[Johnny Engine]

Is there a reason or context that I'm missing that Roy isn't on either list?

 

[overpass]

4. Coaching Effects

One factor that I'm really focusing on this time around is coaching. I'm looking at the career defensive stats for coaches to try to determine which ones had the largest impact on their netminders' success, and using that as part of my estimate of the team effects on each goalie's numbers.

It's also interesting to note goalies who had standout success for one coach vs. many. Obviously everyone in this round was elite for the vast majority of their careers, but not all of them were able to win awards for multiple coaches. For example, here's the current pool (Tretiak excluded) with the number of different coaches they have had for a 1AST team (pre-1982) or Vezina (post-1982) season:

Glenn Hall (4): Jimmy Skinner, Rudy Pilous*, Billy Reay, Scotty Bowman
Dominik Hasek (3): John Muckler, Ted Nolan, Lindy Ruff
Martin Brodeur (3): Pat Burns, Jacques Lemaire, Brent Sutter
Ken Dryden (1): Scotty Bowman
Patrick Roy (1): Pat Burns
Terry Sawchuk (1): Tommy Ivan
Jacques Plante (1): Toe Blake

(*-Hall had the 1957-58 season that was split between Pilous and Tommy Ivan, but I'm only counting the coach with the most games in each season for simplicity's sake)

The top three on that list had success in a few different coaching environments, although Hall was the only one to accomplish this for different franchises (and he impressively did it for 3 of them). Everybody knows about Sawchuk's five-year peak which came in one consistent coaching environment, while Dryden and Plante both had one coach for almost their entire tenures in Montreal. Roy is the one that really stands out here, given that he had Pat Burns for only 4 of his 19 seasons, but those 4 years accounted for all 3 of his Vezinas and one of his Vezina runner-up finishes, while he went 2-3-3-4-5-5 in Vezina voting over the rest of his career. Which raises the obvious question, how much of that great 1989-1992 run was Roy peaking, and how much of it was Burns?

Similarly, how many of these guys faced coaching situations which might have negatively impacted their stats, awards, or team success? For this round, I'm particularly interested in trying to get a good defensive estimate for Sawchuk's Boston and Detroit 2.0 years.

I agree that coaching is a big factor. Looking forward to the discussion!

Re: the numbers above, its true that Patrick Roy was only awarded a Vezina trophy when playing for Pat Burns. But he was a first team all-star and third in Hart voting for Bob Hartley in 2001-02.

We've discussed before that Roy's Colorado environment may not have been favourable for save percentage, although it was favourable for winning. I'd like to examine this a little more. First, although Roy had only the one postseason all-star team in 2001-02 while in Colorado, he had the second best save percentage over that time period, behind Hasek.

1996-97 through 2002-03, regular season and playoffs

Goalie	SA	GA	TOI	GAA	SV%	SA/60
Dominink Hasek	12745​	931​	26823​	2.08​	0.927​	28.5​
Patrick Roy	15066​	1215​	32804​	2.22​	0.919​	27.6​
Jose Theodore	7502​	646​	14942​	2.59​	0.914​	30.1​
Martin Brodeur	14232​	1234​	35811​	2.07​	0.913​	23.8​
Ed Belfour	12604​	1093​	29582​	2.22​	0.913​	25.6​

Roy was a clear #2 in save percentage behind Dominik Hasek for this time period.

(Martin Brodeur is another discussion...maybe he matches Roy when you account for shot undercounting and his ability to play the puck.)

Now I'll present Roy's numbers for this time period in games where Peter Forsberg played and in games where Forsberg didn't play.

1996-97 through 2002-03, regular season and playoffs

Goalie	W	L	T	W%	GAA	SV%	SA/60
Roy with Forsberg	233​	120​	49​	0.641​	2.25​	0.918​	27.6​
Roy without Forsberg	72​	51​	15​	0.576​	2.13​	0.923​	27.5​

Roy played 138 games without Peter Forsberg in this time period, and his save percentage was 5 points higher in those games than it was in the games with Forsberg. I'm not prepared to say this was because Forsberg was a bad defensive player, maybe it just shows the effect of having two scoring lines vs 1 scoring line. But I think it shows that something about the Colorado environment made it harder to have a high save percentage. When Forsberg was out of the lineup, presumably the Avs played more defensively, such as in the 01-02 regular season and the 01 playoffs. And Roy's save percentage increased from 0.919 to 0.923.

That 0.004 change closes half the 0.008 gap between Roy and Hasek in this time period.

Edit: and if you prefer GAA, Roy without Forsberg (2.13) is nearly equal to Brodeur (2.07) and Hasek (2.08) in this time.

 

[Michael Farkas]

Is there a reason or context that I'm missing that Roy isn't on either list?

It stemmed from a piece about the older guys. Equipment was likely much more cumbersome and absorbent back in the day, which may have been a contributing factor that someone like Roy wouldn't have had to deal with quite as much.

 

[BadgerBruce]

So we're getting into the weeds again of some well-worn debate territory, but I wanted to bring up some general discussion topics that are not only relevant to the discussion, but might potentially help people think about their rankings and what they value.

First up is something that has changed significantly even since we did the last goalie list on HOH: Goalie Workloads.

Braden Holtby and Jonathan Quick started 72 and 71 games in 2014-15. Since then, Cam Talbot in 2016-17 is the only guy with 70 or more starts, and I wonder whether he might just be the last goalie to ever start 70 games in a season in the NHL. In the last two seasons, nobody even played in 65 games. Pretty much everyone in hockey has now come to an agreement on the fact that load management for goalies is the objectively correct decision, and that having a goalie that plays 70 games in a season isn't a virtue, it's a mistake.

Some of the reasons that this is true:

- Playing lots of games increases wear and tear and injury risk

- A starter who plays a lot of games is also necessarily taking a greater share of the easy games, in which he's less likely to have a significant impact on win probability

- There are fatigue effects, especially in the playoffs. There appears to be a pretty strong negative correlation between regular season workload and playoff success. Only two goalies since the 2005 lockout have won the Cup while playing at least 60 regular season games, for example, and many elite goalies have pretty mediocre playoff records in their high-workload seasons.

As a result of this development, I don't see any reason to give goalies credit for a high number of games. In my current evaluation system, I don't give anyone credit for more than 65 games played, because I don't think that represents additional value to a team.

Also, I think we should pay attention to a goalie's workload and how that may have impacted their results, whether it was padding their stats from lots of games, or being someone who was forced to play more games than they should have despite physical concerns, or benefiting from a more optimal usage which helped set them up for success.

2. Non-NHL Value

I think it is very important to consider relevant non-NHL performance. There were some good discussions in the preliminary thread about the difficulty of getting a goalie job in the Original 6 pre-1965, and that shouldn't be forgotten.

Jacques Plante, for example, turned pro with the Montreal Royals in 1949-50 in his age 21 season, and already in that very first year he was getting paid to be the practice goalie for the Montreal Canadiens and was clearly viewed as third in line on their organizational depth chart, behind Durnan and McNeil. I'm quite confident he would have been an NHLer in any post-expansion scenario at that age, but he came up in the one-goalie era on the team with the best goaltending depth in the league at a time when teams completely controlled their players' destinities, and as a result he had to wait 5 years to gain the starting job.

For the sake of comparison:

10 goalies played in the NHL in Terry Sawchuk's first season as a starting goalie.
41 played in Ken Dryden's first season as a starting goalie.
80 played in Martin Brodeur's first season as a starting goalie.

Taking everything into account, it's far from obvious to me that Martin Brodeur has more elite longevity than Jacques Plante.

And has already been pointed out in this thread, comparing Hasek to Roy without factoring in Europe at all is simply an incomplete analysis of the relevant resumes of those two goaltenders.

I'm certainly not saying that everything matters and we should consider ECHL seasons or junior careers or count every time somebody retired to go play semi-pro in Europe. It's fine to have a more exclusive NHL focus after the world opened up and the best talent worldwide had an open door to the NHL (say, anybody starting their careers in the late '90s or later). But if we're trying to fairly compare across eras and adjust for different team environments, we should recognize the relative difficulty of goalie paths to the NHL as well, and give proper credit for those performances.

I do not completely agree with your views on goaltenders who consistently carried extraordinarily heavy workloads.

Not my style to rant, but it’s still early morning in central Canada and I’m all out of coffee (don’t ask about the wine, which somehow disappeared late last evening …).

O6 goalies lived with the legitimate belief that coming out of the lineup put their jobs at risk. It’s not an overstatement to say that many of them played very scared — the pressure to play well night after night, lest the mortgage goes into default and the house is gone. In almost all cases, these goalies (like the skaters) played on 1-year contracts. Practice a little career providence and you just might end up playing in Providence.

That’s no joke. What happens if the door is open even a crack? That’s how Al Rollins took the job from Turk Broda, Jacques Plante took the job from Gerry McNeil, Sawchuk took the job from Harry Lumley, etc. Eddie Johnston didn’t start all 70 games for the Bruins in 63-64 just for kicks. No, that had a lot to do with guys like Cheevers, Parent and Favell waiting for EJ to open the door just a crack.

So yeah, I give the O6 guys who played 90% or more of the games year after year a ton of credit. Crappy gear, no masks, no job security.

Imagine this very project you’re all participating in only allowing 6 guys instead of the 20 who submitted lists. The project admins. have the arbitrary authority to replace any of the 6 participants if they believe someone better is waiting in the wings. Now, every post you make matters, right? Want to rest up your brain, skip the Round 2/Discussion 8 to make sure you’re ready for Discussions 9 and 10? Might not be a place for you if your replacement makes solid contributions. Know why Shakespeare wrote about 40 plays over 25 years and never took a break? To make a living during an age when contemporaries (Robert Greene, John Lyly, Christopher Marlowe, Thomas Nashe, Ben Jonson, etc.) were nipping at his heels. Livelihood, right? Gotta protect it.

Rant over. One of my kids just brought me a Timmies.

 

[overpass]

Goalie Workloads.

Braden Holtby and Jonathan Quick started 72 and 71 games in 2014-15. Since then, Cam Talbot in 2016-17 is the only guy with 70 or more starts, and I wonder whether he might just be the last goalie to ever start 70 games in a season in the NHL. In the last two seasons, nobody even played in 65 games. Pretty much everyone in hockey has now come to an agreement on the fact that load management for goalies is the objectively correct decision, and that having a goalie that plays 70 games in a season isn't a virtue, it's a mistake.

Some of the reasons that this is true:

- Playing lots of games increases wear and tear and injury risk

- A starter who plays a lot of games is also necessarily taking a greater share of the easy games, in which he's less likely to have a significant impact on win probability

- There are fatigue effects, especially in the playoffs. There appears to be a pretty strong negative correlation between regular season workload and playoff success. Only two goalies since the 2005 lockout have won the Cup while playing at least 60 regular season games, for example, and many elite goalies have pretty mediocre playoff records in their high-workload seasons.

As a result of this development, I don't see any reason to give goalies credit for a high number of games. In my current evaluation system, I don't give anyone credit for more than 65 games played, because I don't think that represents additional value to a team.

Also, I think we should pay attention to a goalie's workload and how that may have impacted their results, whether it was padding their stats from lots of games, or being someone who was forced to play more games than they should have despite physical concerns, or benefiting from a more optimal usage which helped set them up for success.

I have also been thinking that the findings of modern load management should update our analysis.

But I'm not sure what to make of the actual data. Have you looked at the days by rest report on nhl.com? Most goaltenders' save percentages with 0 days rest are similar to their overall numbers.

For example Martin Brodeur has played in 180 games with 0 days of rest. Of all goalies in the last 50 years*, that's over 50% more than #2. His save percentage in those games is 0.912, identical to his career save percentage of 0.912.

Others with similar numbers. Maybe Roy and Hasek were worse with 0 days of rest, but the statistical difference is only about one standard deviation.

Ed Belfour - 0.906 overall, 0.911 on 0 days rest (106 GP)
Dominik Hasek - 0.922 overall, 0.918 on 0 days rest (103 GP)
Curtis Joseph - 0.906 overall, 0.904 on 0 days rest (93 GP)
Patrick Roy - 0.910 overall, 0.905 on 0 days rest (91 GP)
Henrik Lundqvist - 0.918 overall, 0.926 on 0 days rest (90 GP)
Roberto Luongo - 0.919 overall, 0.919 on 0 days rest (84 G)

It's possible these numbers are skewed by selection effects, where goaltenders have only been played in the easier back to backs. Although Brodeur played 70% of 0 days of rest games on the road, so I'm not sure that's the case for him, at least.

And the 0 days of rest data doesn't speak to the fatigue or wear and tear that a workhorse goalie might accumulate by the end of a season.

With regards to Brodeur, I wondered if his style played a role, and if the modern pro-fly style needs more recovery than some styles of the past. I see @Michael Farkas is thinking along the same lines.

*I wouldn't draw any conclusions from 0 days of rest data from the Original Six. Each team had it's own specific travel schedule where some teams played their back to backs at home and others played their back to backs on the road. Days of rest data from that era can't be compared across teams.