Roy or Hasek?

[billingtons ghost]

Stats monkey?!

I am a regular monkey, thank you very much!

I get what you’re saying in that we have such a brilliant collection of top goaltenders - I think I have 7 in my top-35 players - that having a conversation between two of them might be too narrowly focused, because it’s like debating Crosby/Ovechkin without Malkin or Gretzky/Lemieux without Messier (don’t @ me).

But of the top goaltenders, I (and probably many others) feel that some of the seven answers you hear regarding best goaltender are more right than others.

I think there’s a reason you’ll hear Roy and Plante in the context of the Beliveau/Harvey/Hull/Richard block of career value or Sawchuk and Hasek in the context of the otherworldly five-year prime discussions, but Brodeur is almost entirely relegated to historical discussions exclusively with other goaltenders. I’d say Glenn Hall and Ken Dryden are in that same boat.

Assuming Martin Brodeur is your selection for greatest goaltender of all-time, how many skaters are we saying are better than literally every goaltender? For me, Martin Brodeur is about as good as Nicklas Lidstrom. Maybe it’s because they played at the same time, but I just see too many parallels between the two.

Now do you see Martin Brodeur at a higher level than this (Lidstrom)? Do you see him as a reasonable comparison point to Jean Beliveau and Maurice Richard the way many do of Patrick Roy? Or do you think the high-water mark for goaltending is more #15-25 range than #5-10?

I think your Lidstrom comparison is fair - I just think he's pretty criminally underrated if you ask the right questions - especially since no one ever SAW New Jersey games unless they were in a playoff run.

Hasek is far and away the most spectacular of any goalie in this discussion, and perhaps had to be - and had, as you say, peak other worldly stats. If you asked me who I thought was the most fun to watch, I'd answer Hasek. But the question becomes - as someone pointed out - Hasek was supposedly the workhorse, yet Brodeur showed up in 70+ games 12 times to Hasek's one and played 500 more games (and at a high level) than Hasek in the NHL.

Roy is undeniably the greatest playoff goaltender, right? Brodeur had 7 of his first 9 seasons with a playoffs GAA of under 2.00. Roy had 3 in his 19 year career. Brodeur had 7 shutouts in one playoff season - neither Roy(4) nor Hasek (6) matched that. He played in 40 (20%) fewer playoff games than Roy, but had a higher SV% and lower GAA than Roy for his playoff career.

Brodeur had almost DOUBLE the regular season shutouts of Roy and had 300 more wins than Hasek - they made a rule change in the way the game is actually played specifically because of him and they may change it back after he's gone. I'm not sure many sports have done that for certain athletes.

Those are pretty big numbers to sweep under the rug - and he went to the Cup Finals playing at a high level even in his 19th season. I'm not sure why he doesn't get the same press -

Someone made the comparison to Gordie Howe based upon longevity - but the reality was that Brodeur didn't play over 5 decades - his career spanned roughly the same time as Roy's and Hasek's - he just was available for more games during each season and put up more impressive numbers during that time and had to be just as good to attain the same stats - just more frequently during more nights in a single season. That alone should be worthy of recognition - not a reason to penalize or belittle the accomplishments.

 

[Merya]

Hasek by a margin bigger than in any other position. Also, Roy is not second.

 

[Hockey Outsider]

Roy is undeniably the greatest playoff goaltender, right? Brodeur had 7 of his first 9 seasons with a playoffs GAA of under 2.00. Roy had 3 in his 19 year career. Brodeur had 7 shutouts in one playoff season - neither Roy(4) nor Hasek (6) matched that. He played in 40 (20%) fewer playoff games than Roy, but had a higher SV% and lower GAA than Roy for his playoff career.

On average, Roy spent his career playing in a league that had a much higher scoring environment than Brodeur. Comparing GAA and save percentage, without taking the league context into account, is useless.

Adjusted for era, Patrick Roy has the highest playoff save percentage of any goalie (1984-2018) who has played anywhere close to the same number of games as him. Link.

 

[Canadiens1958]

On average, Roy spent his career playing in a league that had a much higher scoring environment than Brodeur. Comparing GAA and save percentage, without taking the league context into account, is useless.

Adjusted for era, Patrick Roy has the highest playoff save percentage of any goalie (1984-2018) who has played anywhere close to the same number of games as him. Link.

Curious to see the splits and adjustments for Patrick Roy with the Canadiens and with the Avalanche.

 

[LeHab]

How about these were two great goalies in their own way? What was amazing with Roy is just how much impact he had back in Quebec. Suddenly every kid wanted to be the next Roy which spawned over a decade of elite Quebec goaltenders. This even if Lemieux and Gretzky were performing at a high level.

 

[Hockey Outsider]

Curious to see the splits and adjustments for Patrick Roy with the Canadiens and with the Avalanche.

His adjusted playoff save percentage was 92.7% in Montreal versus 91.4% in Colorado. The main reasons are three of his five big runs were as a Hab, and he was less consistent in Denver (his last two playoffs drag down his average).

 

[Orange Dragon]

Hasek by a margin bigger than in any other position. Also, Roy is not second.

So who is?

 

[Canadiens1958]

His adjusted playoff save percentage was 92.7% in Montreal versus 91.4% in Colorado. The main reasons are three of his five big runs were as a Hab, and he was less consistent in Denver (his last two playoffs drag down his average).

How big of a difference,if any, in SOGs per game? Thank you.

 

[Canadiens1958]

Stats monkey?!

I am a regular monkey, thank you very much!

I get what you’re saying in that we have such a brilliant collection of top goaltenders - I think I have 7 in my top-35 players - that having a conversation between two of them might be too narrowly focused, because it’s like debating Crosby/Ovechkin without Malkin or Gretzky/Lemieux without Messier (don’t @ me).

But of the top goaltenders, I (and probably many others) feel that some of the seven answers you hear regarding best goaltender are more right than others.

I think there’s a reason you’ll hear Roy and Plante in the context of the Beliveau/Harvey/Hull/Richard block of career value or Sawchuk and Hasek in the context of the otherworldly five-year prime discussions, but Brodeur is almost entirely relegated to historical discussions exclusively with other goaltenders. I’d say Glenn Hall and Ken Dryden are in that same boat.

Assuming Martin Brodeur is your selection for greatest goaltender of all-time, how many skaters are we saying are better than literally every goaltender? For me, Martin Brodeur is about as good as Nicklas Lidstrom. Maybe it’s because they played at the same time, but I just see too many parallels between the two.

Now do you see Martin Brodeur at a higher level than this (Lidstrom)? Do you see him as a reasonable comparison point to Jean Beliveau and Maurice Richard the way many do of Patrick Roy? Or do you think the high-water mark for goaltending is more #15-25 range than #5-10?

Ken Dryden, path worth exploring. Only two RS seasons with 60+ games but a playoff beast. Short career.

Ken Dryden Stats | Hockey-Reference.com

 

[Bear of Bad News]

Hasek by a margin bigger than in any other position. Also, Roy is not second.

Is this another "fact"?

 

[Troubadour]

But they did have Forsberg in the 2002 series and were still heavily outplayed. I’ve given you the series shot differential twice now. Even if Forsberg makes them more than a 99-point regular season team (they had 95, 98, 96, and 118 leading into 2002 and just lost two top-4 defensemen from the previous season), they’re not adding a second Forsberg so they’re still getting outshot heavily. And you’re putting it all on the goaltender to make up for this ~55 shot disparity and saying it matters that it’s a 4-3 series loss despite the shot differential. Why?

It's more like why are you so hung up on this? I tried to explain what I meant--to no success. Thus I'm about to drop this.

And how many series have been won by cumulative save percentage?

Roy gave his teams a great chance to win 6 games out of 8, as did Hasek. If you want to argue that Roy gave his teams less of a chance to win in his two worst games than Hasek did in his two worst games of the eight, no one’s going to dispute that either. The problem is assigning this much retroactive value to 8 games in which the shots are lopsided as though it’s the only head-to-head matchup that either cared about in their careers.

But if you want to assert that it has such meaning, then I also want to hear if you believe Roy would play worse than he did against Buffalo in 1993 if Hasek was in net instead of Fuhr, or if Hasek would play better against Canada in 1991 if Roy was in net instead of Ranford.

Are you so confident in the meaning of this 3-4-1 head-to-head record in the playoffs/Olympics that you think the mere presence of the other would substantially alter their performances in 1991 and 1993?

Here it gets tricky. Because we have to go Big in Japan again, as you proposed the Nagano semis as an example of Roy topping Hasek SV%, ignoring that it went to a SO where Hasek stole the game and made Roy--as Royal as he was in that shootout himself--look rather human and ordinary.

Which begs a question. How many games did a better per-game SV% itself win? I have no idea. All I know is that, in the Roy Hasek H2H context, Roy having a slightly better SV% in the Olympic semis did you-know-what all, and Hasek having a better cumulative SV% correlated with him taking the semis AND the 2002 SF as well.

So in the context of what some of us appear to talk about, the cumulative stat does seem a bit more relevant and way less misleading than say... pointing out Roy's better SV% in Nagano semis, wouldn't you say?

Also, their H2H has a regular season history as well. Why leave that out?

I apologize for the rest, I don't have time as of now. Could get back to it later. Cheer up and cheers.

 

[billingtons ghost]

On average, Roy spent his career playing in a league that had a much higher scoring environment than Brodeur. Comparing GAA and save percentage, without taking the league context into account, is useless.

Adjusted for era, Patrick Roy has the highest playoff save percentage of any goalie (1984-2018) who has played anywhere close to the same number of games as him. Link.

While I am all for putting stats in context, (and I commend you for that) and I completely agree with your point in general - I don't think you can definitively say Roy outplayed or out-statted Brodeur over the period of overlap. Brodeur had 7 of 9 years with sub 2.00 playoff GAA, Roy had 2.
Marty had a couple of clunkers in '98-99 and 2000 when it comes to SV%, but even with those - .918 avg versus .921 is hardly something to base an argument upon given the other evidence.

1993–94	New Jersey Devils	NHL	17	8	9	1171	38	1	1.95	.928
1994–95	New Jersey Devils	NHL	20	16	4	1222	34	3	1.67	.927
1996–97	New Jersey Devils	NHL	10	5	5	659	19	2	1.73	.929
1997–98	New Jersey Devils	NHL	6	2	4	366	12	0	1.97	.927
1998–99	New Jersey Devils	NHL	7	3	4	425	20	0	2.83	.856
1999–00	New Jersey Devils	NHL	23	16	7	1450	39	2	1.61	.927
2000–01	New Jersey Devils	NHL	25	15	10	1505	52	4	2.07	.897
2001–02	New Jersey Devils	NHL	6	2	4	381	9	1	1.42	.938
2002–03	New Jersey Devils	NHL	24	16	8	1491	41	7	1.65	.934

[TBODY] [/TBODY]

1993–94	Montreal Canadiens	NHL	6	3	3	374	16	0	2.56	.930
1995–96	Colorado Avalanche	NHL	22	16	6	1453	51	3	2.10	.921
1996–97	Colorado Avalanche	NHL	17	10	7	1033	38	3	2.21	.932
1997–98	Colorado Avalanche	NHL	7	3	4	429	18	0	2.51	.906
1998–99	Colorado Avalanche	NHL	19	11	8	1173	52	1	2.66	.920
1999–00	Colorado Avalanche	NHL	17	11	6	1039	31	3	1.79	.928
2000–01	Colorado Avalanche	NHL	23	16	7	1450	41	4	1.70	.934
2001–02	Colorado Avalanche	NHL	21	11	10	1241	52	3	2.51	.909
2002–03	Colorado Avalanche	NHL	7	3	4	423	16	1	2.27	.910

[TBODY] [/TBODY]

 

[Doctor No]

Just to be clear, your argument is that Brodeur's statistics during his prime years were about the same as Roy's statistics during his post-prime years?

 

[billingtons ghost]

Just to be clear, your argument is that Brodeur's statistics during his prime years were about the same as Roy's statistics during his post-prime years?

Are you saying that Roy's prime years were significantly shorter in comparison that they can't be justly compared to another goalie who was still playing 70 games a season at the age of 38 and for a cup at the age of 40 - setting the NHL all-time wins record in a season at the age of 36? Not sure what this 'prime' is.

Are we looking at continued and prolonged excellence - or do Arturs Irbe, Jim Carey and Jose Theodore also top your list of best playoff goalies ever?

That's obvious silly hyperbole, but my underlying point is valid -

Just to be clear - I'm comparing overlapping years because the question arrived about 'the state of the league during that time' and I do not possess a time machine.

I could just as easily blather and cherry-pick - Perhaps Roy's SV% is inflated because of 'league style of play' during his early years, and if Brodeur played during those years he would have had a .940 SV% every year. It's just can't be evenly compared.

 

[Doctor No]

You're being obtuse, and given your allegiances, it's deliberately so.

 

[Canadiens1958]

Are you saying that Roy's prime years were significantly shorter in comparison that they can't be justly compared to another goalie who was still playing 70 games a season at the age of 38 and for a cup at the age of 40 - setting the NHL all-time wins record in a season at the age of 36? Not sure what this 'prime' is.

Are we looking at continued and prolonged excellence - or do Arturs Irbe, Jim Carey and Jose Theodore also top your list of best playoff goalies ever?

That's obvious silly hyperbole, but my underlying point is valid -

Just to be clear - I'm comparing overlapping years because the question arrived about 'the state of the league during that time' and I do not possess a time machine.

I could just as easily blather and cherry-pick - Perhaps Roy's SV% is inflated because of 'league style of play' during his early years, and if Brodeur played during those years he would have had a .940 SV% every year. It's just can't be evenly compared.

Very provocative series of posts raising valid points.

Basically you suggest reviewing how goalies are ranked compared to skaters given the paradoxical situation that playing more games benefits skaters while playing fewer games benefits goalies. Complicated by the fact that goalies tend to managed in various back-to-back, three games in four nights and sitting at home waiting situations.

The situation with Glenn Hall comes to mind. Played 502 consecutive RS games in the one goalie era, yet tended to weakenin the playoffs.

The idea of overlapping or two consecutive seasons deserves further exploration as a counterbalance to prime especially as we get deeper into the Salary Cap era and roster stability becomes more and more fragile.

 

[GuineaPig]

I've always argued that it is misleading to look at individual seasons for goalies because of the role of variance in sv %. 60-70 games really isn't all that large a sample size (which in my mind makes all the head-to-head comparisons especially silly). Plenty of goalies go red hot for 50 games. Does it make more sense to think of Jose Theodore as a solid starting goalie in the early '00s (0.917 from 2000-01 to 2002-03), or as a goalie who was average, then, world class, and then average again over three consecutive groupings of 60 games?

 

[quoipourquoi]

Which begs a question. How many games did a better per-game SV% itself win?

That doesn’t independently win a game either. And that’s the point. Patrick Roy and Dominik Hasek do not substantially control the shots taken by the other team. They can have the same error rate in a game but if Patrick Roy faces 42 shots and Dominik Hasek faces 21 shots, Roy’s team loses 2-1.

That is not a hypothetical; this happened in Game 3.

But what it does tell us - looking at the individual games as opposed to defining a performance by its cumulative - is how often a goaltender played well enough to give a team a chance to win. Because it’s about winning four individual games.

If you know Patrick Roy stops just 76 of 86 cumulative shots in his losses in Games 3, 6, and 7, you can do the quick math and say, yes, .884 is not a good number.

But if you know that the breakdown was 40/42 (.952), 26/28 (.929), and 10/16 (.625), the last one was totally on him, but he did his part in the first two.


So, two things:

1. Swap Dominik Hasek and Patrick Roy - do the Colorado Avalanche with Hasek then outshoot the Detroit Red Wings with Patrick Roy by a 223-168 margin? Is there evidence that having Patrick Roy results in substantially more shots against or that having Dominik Hasek results in substantially more shots for? If not, then I’m not going to take a whole lot of stock in head-to-head matchups based largely on this series that swung on a 42-21 shot game.

2. Do you believe that if Roy was in net for Canada in 1991 instead of Ranford, Dominik Hasek would have performed substantially better than he did (6 GA on 36 shots)? Do you believe that if Hasek was in net for Buffalo in 1993 instead of Fuhr, Patrick Roy would have performed substantially worse than he did? If not, then why put stock in the head-to-head record?

 

[quoipourquoi]

I could just as easily blather and cherry-pick - Perhaps Roy's SV% is inflated because of 'league style of play' during his early years, and if Brodeur played during those years he would have had a .940 SV% every year. It's just can't be evenly compared.

I’d like to think that if Martin Brodeur was capable of a .940 in 1991-92, the Devils would have played him more than they did that year. At any rate, I think it’s more likely that he would have seen fewer starts in his career due to the prevalence of goaltending tandems if he was born 6 years earlier.

 

[billingtons ghost]

You're being obtuse, and given your allegiances, it's deliberately so.

You're certainly correct with respect to the hyperbolic Jose Theodore and the 'what is prime' but I think my underlying points at least deserve consideration instead of one-line zingers from the hip from you...

 

[billingtons ghost]

Some really excellent stuff here...

That doesn’t independently win a game either. And that’s the point. Patrick Roy and Dominik Hasek do not substantially control the shots taken by the other team.

I mean for me - we are stuck with really poor metrics to gauge goaltender performance until those guys implement Valiquette's golden road shot qualification. I think at the end of the day someone like Fuhr or Barasso or Quick or (god forbid me saying this) Richter might end up the 'best' goaltender at stopping unstoppable shots. Hasek might be there too.

I've always argued that it is misleading to look at individual seasons for goalies because of the role of variance in sv %. 60-70 games really isn't all that large a sample size (which in my mind makes all the head-to-head comparisons especially silly). Plenty of goalies go red hot for 50 games. Does it make more sense to think of Jose Theodore as a solid starting goalie in the early '00s (0.917 from 2000-01 to 2002-03), or as a goalie who was average, then, world class, and then average again over three consecutive groupings of 60 games?

Absolutely spot on - we just saw it this season on numerous teams where goalies got out to hot starts and then fell apart.

 

[billingtons ghost]

Very provocative series of posts raising valid points.

Basically you suggest reviewing how goalies are ranked compared to skaters given the paradoxical situation that playing more games benefits skaters while playing fewer games benefits goalies. Complicated by the fact that goalies tend to managed in various back-to-back, three games in four nights and sitting at home waiting situations.

The situation with Glenn Hall comes to mind. Played 502 consecutive RS games in the one goalie era, yet tended to weakenin the playoffs.

The idea of overlapping or two consecutive seasons deserves further exploration as a counterbalance to prime especially as we get deeper into the Salary Cap era and roster stability becomes more and more fragile.

Thanks for adding some credence to what I was trying to say. Glen Hall is a very good example of what level of grind goalies go through.

 

[quoipourquoi]

Marty had a couple of clunkers in '98-99 and 2000 when it comes to SV%, but even with those - .918 avg versus .921 is hardly something to base an argument upon given the other evidence

I’d like to highlight something here, and that is the use of cumulative save percentage. Even though you’re using a 1994-2003 overlap for both players, running a cumulative save percentage report does not work, as it will treat 18 saves on 20 shots in 1994 the same as 18 saves on 20 shots in 2003 - essentially meaning one who has to perform better in 1994 to get equal credit for something more commonplace in 2003.

Cumulative reports across a series of seasons that have seen substantial changes to what is considered normal are essentially broken.

Here’s a look at their error rate vs. expectation in the years in which their teams went three rounds deep in the overlap.

Martin Brodeur, 1994 - 66.9% on 531 shots
Martin Brodeur, 1995 - 64.9% on 463 shots
Martin Brodeur, 2000 - 70.8% on 537 shots
Martin Brodeur, 2001 - 96.2% on 507 shots
Martin Brodeur, 2003 - 68.8% on 622 shots

Patrick Roy, 1996 - 66.9% on 649 shots
Patrick Roy, 1997 - 72.5% on 559 shots
Patrick Roy, 1999 - 82.5% on 650 shots
Patrick Roy, 2000 - 69.6% on 431 shots
Patrick Roy, 2001 - 62.6% on 622 shots
Patrick Roy, 2002 - 87.9% on 572 shots

Now, that’s essentially the best of Martin Brodeur’s playoff career - all of his championships and all but one of his finals runs (97.2% on 629 shots) - but we’re missing three finals runs from Roy. Even if we argue about whether they’re equal in the playoffs or if one is marginally ahead from 1994-2003, we can’t exactly put 1986-1993 from Roy and 2004-2012 from Brodeur on the back burner as though they do not help explain the disparity presented by Hockey Outsider.

1986-1993 includes 108 of Roy’s 247 playoff games where he often played better than he did in the above years. 2004-2014 includes 66 of Brodeur’s 204 playoff games where he often played worse than he did in the above years.

 

[Hockey Outsider]

How big of a difference,if any, in SOGs per game? Thank you.

He faced a bit more rubber in Colorado - 29.1 shots per 60 minutes compared to 27.7 in Montreal.

 

[Canadiens1958]

He faced a bit more rubber in Colorado - 29.1 shots per 60 minutes compared to 27.7 in Montreal.

Thank you. Interesting given that in Montreal he played during a higher scoring era.