Top-100 Hockey Players of All-Time - Round 2, Vote 5

[Canadiens1958]

Except the undercounting of shots is something that is demonstrated by looking at the total shots counted in devils home games compared to away games.

Not the point.. Undercounting should be constant if it exists. Rebound control varies from goalie to goalie. Would show if you have the best and worst rebound goalie facing each other in the same NJ home game.

 

[Hockey Outsider]

@seventieslord - great summary of some of the arguments for and against Brodeur (previous page). I may expand on that next round, time permitting.

It's interesting how Brodeur is simultaneously underrated and overrated by save percentage, for different reasons. My own research (especially when looking at "goals versus threshold" - that is, how much value a netminder provided relative to the theoretical borderline NHLer) has firmly convinced me of Brodeur's value, even if his peak performance was more on the Luongo/Belfour level, rather than the Hasek/Roy level. He probably should have won the Smythe in 1995 as well.

As to whether shots were under-counted in New Jersey - I'm open to seeing any evidence about that, but as far as I know, that's never been proven.

 

[seventieslord]

I think he's trying to say that a goalie that has great rebound controal can't pad his SV% numbers from stopping rebounds.

Of course, I'm not saying that goalies give rebounds for that sole purpose. And I don't think second, third and fourth shots are statistically easier to stop that first shots, especially since we're talking here about uncontrolled rebounds (and a controlled rebound presumably doesn't result in a shot).

You're right, they're not. It would be impossible for a goalie to pad his stats, deliberately or otherwise, by giving up more rebounds.

 

[seventieslord]

What are the numbers there? Is it statistically significant (like 1+ shot a game), or de minimis?

And what affect would that have over save %? Does it move the needle in a meaningful way, or would it be like a slight move at the thousandth place?

I believe it works out to about 1 shot per game, and yes, that makes a big difference in save percentage. It could be as many as 3 points in the long run, which, if you look at hockey Outsider's tables, moves a goalie significantly up or down within that cluster.

Not the point.. Undercounting should be constant if it exists. Rebound control varies from goalie to goalie. Would show if you have the best and worst rebound goalie facing each other in the same NJ home game.

I'm pretty sure Devils fans have pointed out that the undercounting does exist and it is constant.

This can't be the first time you've heard of this....?

@seventieslord - great summary of some of the arguments for and against Brodeur (previous page). I may expand on that next round, time permitting.

It's interesting how Brodeur is simultaneously underrated and overrated by save percentage, for different reasons. My own research (especially when looking at "goals versus threshold" - that is, how much value a netminder provided relative to the theoretical borderline NHLer) has firmly convinced me of Brodeur's value, even if his peak performance was more on the Luongo/Belfour level, rather than the Hasek/Roy level. He probably should have won the Smythe in 1995 as well.

As to whether shots were under-counted in New Jersey - I'm open to seeing any evidence about that, but as far as I know, that's never been proven.

I thought it has been. BM67 has the data, I'm sure.

 

[ChiTownPhilly]

Then step-up with the name of a Russian interchangeable part defenceman who was available.

Coffey, Lidstrom, Murphy?

Oh, my dear Hockey Gods, you just cannot make this up!

Three listed were not Russian. Bowman's Russian 5 required Russians.

Axiomatic, isn't it?

You know- in my entire HF Boards posting-stint, there's no poster with whom I've had more disagreements than 'Canadiens1958.' However, I will certainly join with him in agreeing to the freaking self-evident(!) Russian-trained 5-man unit motion, puck-handling, positioning, rotation, recovery could only be done by those trained in its ways. This was practiced by the Russians until it worked its way into their muscle-memory. It was raised to an even higher degree with players like Fetisov and Larionov-- who shared more than a decade of experience as direct teammates before their Detroit campaigns. It wasn't refined by Bowman, great though he was/is. He couldn't have coached anyone up to do it. The annals are pretty clear on this matter. He sent them over the boards and said "do what you do."

 

[Michael Farkas]

For what it's worth...I just did a quick and dirty count of Slava Fetisov's ice time in game 4 of the 1997 Stanley Cup Final vs. Philadelphia. Some context: Fetisov can only play LD. Lidstrom played ahead of him. Konstantinov could also play LD at times. Larry Murphy played on the right, usually with Lidstrom. Fetisov played 19:14 of the game. He was just about perfect defensively, poised with the puck in his own end, a major part of their breakout effectiveness despite not being as fast as he was in his prime of course...

Just by eyeball, I'd say Lidstrom probably played like 25, 27, 28 minutes in that game...he was out there constantly. As we would come to expect. But, yeah, the 39 year old Soviet defenseman played 19:14 in an elimination game on a team that swept the Stanley Cup Final and was a noticeable positive in the game. Just information, not making a point really...

 

[seventieslord]

For what it's worth...I just did a quick and dirty count of Slava Fetisov's ice time in game 4 of the 1997 Stanley Cup Final vs. Philadelphia. Some context: Fetisov can only play LD. Lidstrom played ahead of him. Konstantinov could also play LD at times. Larry Murphy played on the right, usually with Lidstrom. Fetisov played 19:14 of the game. He was just about perfect defensively, poised with the puck in his own end, a major part of their breakout effectiveness despite not being as fast as he was in his prime of course...

Just by eyeball, I'd say Lidstrom probably played like 25, 27, 28 minutes in that game...he was out there constantly. As we would come to expect. But, yeah, the 39 year old Soviet defenseman played 19:14 in an elimination game on a team that swept the Stanley Cup Final and was a noticeable positive in the game. Just information, not making a point really...

Sounds about right.

TOI estimates for that season have him at 20.2 per game, the clear #3.

Lidstrom averaged 28, Konstantinov 23. Murphy estimated at 18 in the regular season after arriving from Toronto, though that's a small sample to estimate from, and he almost surely played more than that (and more than Fetisov) in the playoffs.

Being the #4 on a dominant cup winner at 38 is not insignificant. I guess, in comparison to everything else he did, it is, but ya know...

 

[BM67]

The shots, goals, and shooting% for both teams in Devils home and road games over 20 seasons. More shots in Devils road games in 17 of 20. Higher S% in Devils home games in 17 of 20 seasons. The only one of the 3 seasons after 1996-97 with higher S% in road games was 2000-01 where the Devils shot an incredible 11.52% on the road. Over the 20 seasons, the average shots are +2.85 shots in Devils road games, and +3.42 shots from 1997-98 on.

Year	GP	HSOG	HSOG/GP	HG	HS%	RSOG	RSOG/GP	RG	RS%
1993-94	42	2497	59.45	259	10.37	2596	61.81	267	10.29
1994-95	24	1367	56.96	128	9.36	1295	53.96	129	9.96
1995-96	41	2351	57.34	214	9.1	2459	59.98	203	8.26
1996-97	41	2416	58.93	204	8.44	2337	57	212	9.07
1997-98	41	2199	53.63	217	9.87	2145	52.32	174	8.11
1998-99	41	2256	55.02	225	9.97	2335	56.95	219	9.38
1999-00	41	2368	57.76	225	9.5	2450	59.76	229	9.35
2000-01	41	2244	54.73	237	10.56	2370	57.8	253	10.68
2001-02	41	2023	49.34	200	9.89	2322	56.63	192	8.27
2002-03	41	2205	53.78	206	9.34	2329	56.8	176	7.56
2003-04	41	2101	51.24	194	9.23	2333	56.9	183	7.84
2005-06	41	2278	55.56	225	9.88	2520	61.46	233	9.25
2006-07	41	2196	53.56	194	8.83	2490	60.73	205	8.23
2007-08	41	2268	55.32	197	8.69	2351	57.34	194	8.25
2008-09	41	2476	60.39	231	9.33	2637	64.32	214	8.12
2009-10	41	2205	53.78	210	9.52	2467	60.17	192	7.78
2010-11	41	2215	54.02	192	8.67	2281	55.63	186	8.15
2011-12	41	2208	53.85	213	9.65	2244	54.73	208	9.27
2012-13	24	1233	51.38	122	9.89	1239	51.62	110	8.88
2013-14	41	2070	50.49	191	9.23	2218	54.1	201	9.06
Total	787	43176	54.86	4084	9.46	45418	57.71	3980	8.76

[TBODY] [/TBODY]

 

[vadim sharifijanov]

On the flip side, he caught bad breaks in other areas - maybe the 2nd or 3rd best individual playoff run from a Finalist between 2002-2008 but all three happened in the same year, so no Conn Smythe.

interesting, i’d never thought of it that way before. but you’re probably right—btw roy/sakic amd malkin, the only performances i would put in a conversation with those three are 2006 pronger and 2008 zetts and crosby. otoh, 2006 might be a clear #2 for me after giggy.

 

[seventieslord]

Ok, so if shots were undercounted in NJ for years what the heck are we doing trying to paint Luongo a better goalie in 06-07?

Who's "we"?

I did say he was statistically superior, and he was. Then I moved on to some contexts that we should view the statistical record through.

All that does is make Marty's case stronger. Right?

The shot undercounting part does. The powerplay part doesn't. You can't have it both ways on that. Careful not to become a Brodeur ideologue. Seek the truth.

Is it Brodeur's fault the Devil's took less penalties?

Well actually, if it was literally Brodeur's fault that the Devils took fewer penalties, then we're talking about a top-5 player of all-time here, because it saved them countless high-quality shots against and goals against over the years.

In reality, Brodeur had little to no control on the team's discipline, and he benefited from that. It has to be accounted for. Just like contexts that make his case look better need to be accounted for.

Was Luongo giving up more rebounds?

you tell me.

Did Luongo face more rubber, especially factoring in the undercounting of shots?

I'm waiting for the shot undercounting data to be presented. I know it's out there but I've had a Devil of a time finding it on this migrated site. BM67, where are you??

As I've said repeatedly. The biggest knock on Brodeur is his peak is generally considered lower than the goalies off the board and up for vote, along with Dryden and maybe a few more. There's some truth to that, at least in regards to some of the other top goalies. Or that he just benefitted from the trap system, which is horse****.

Again, don't be an ideologue. Of course limiting total shots against, and limiting quality shots against, helps a goalie's ability to post great personal numbers and win games. We are here to determine just how much that should really be considered a factor. Calling it horseshit is not helpful or productive.

But straight out of HO's study we see that his peak save percentage is exactly that if Hall's. Barely behind Plante. Other metrics he's right there as well.

Which brings me to a slight criticism of the project. these numbers are all adjusted to "league average". Look who are all in the top-10: Dryden, Esposito, Resch, Parent. There was a serious difference in the 70s between the haves and the have nots. It allowed the more dominant players (probably at all positions) to post numbers that were more eye-catching. I think a mental adjustment has to be made (or some sort of study that normalizes what constitutes "average" in a weaker league) to properly place players like this.

The reason I am bringing this up, is because the same applies to O6 goalies like Hall, just in the opposite way. If Brodeur is just as much above the league average established by goalies like, I dunno.... Manny Fernandez and Byron Dafoe, as Hall is above the league average established by Sawchuk, Worsley and Bower, then we absolutely have to keep that in mind when quoting their league average adjusted stats.

 

[seventieslord]

The shots, goals, and shooting% for both teams in Devils home and road games over 20 seasons. More shots in Devils road games in 17 of 20. Higher S% in Devils home games in 17 of 20 seasons. The only one of the 3 seasons after 1996-97 with higher S% in road games was 2000-01 where the Devils shot an incredible 11.52% on the road. Over the 20 seasons, the average shots are +2.85 shots in Devils road games, and +3.42 shots from 1997-98 on.

Year	GP	HSOG	HSOG/GP	HG	HS%	RSOG	RSOG/GP	RG	RS%
1993-94	42	2497	59.45	259	10.37	2596	61.81	267	10.29
1994-95	24	1367	56.96	128	9.36	1295	53.96	129	9.96
1995-96	41	2351	57.34	214	9.1	2459	59.98	203	8.26
1996-97	41	2416	58.93	204	8.44	2337	57	212	9.07
1997-98	41	2199	53.63	217	9.87	2145	52.32	174	8.11
1998-99	41	2256	55.02	225	9.97	2335	56.95	219	9.38
1999-00	41	2368	57.76	225	9.5	2450	59.76	229	9.35
2000-01	41	2244	54.73	237	10.56	2370	57.8	253	10.68
2001-02	41	2023	49.34	200	9.89	2322	56.63	192	8.27
2002-03	41	2205	53.78	206	9.34	2329	56.8	176	7.56
2003-04	41	2101	51.24	194	9.23	2333	56.9	183	7.84
2005-06	41	2278	55.56	225	9.88	2520	61.46	233	9.25
2006-07	41	2196	53.56	194	8.83	2490	60.73	205	8.23
2007-08	41	2268	55.32	197	8.69	2351	57.34	194	8.25
2008-09	41	2476	60.39	231	9.33	2637	64.32	214	8.12
2009-10	41	2205	53.78	210	9.52	2467	60.17	192	7.78
2010-11	41	2215	54.02	192	8.67	2281	55.63	186	8.15
2011-12	41	2208	53.85	213	9.65	2244	54.73	208	9.27
2012-13	24	1233	51.38	122	9.89	1239	51.62	110	8.88
2013-14	41	2070	50.49	191	9.23	2218	54.1	201	9.06
Total	787	43176	54.86	4084	9.46	45418	57.71	3980	8.76

[TBODY] [/TBODY]

Thanks so much.

 

[vadim sharifijanov]

my two cents on brodeur—it feels weird to have him available so soon (presumably) before stevens. i understand why, but btw 1994 and 2003, would anybody consider brodeur the more important player?

is vezina-to-end of career brodeur that much better than the first decade of stevens’ career that they would be so far apart on an all-time list? or (as i suspect) might we want to put some pressure on how we prioritize historically significant achievements? not to say that i would have stevens ahead of marty, just that that cognitive dissonance is a decently clear way to illustrate an interesting methodological question.

 

[Nick Hansen]

It'll be interesting to see which ones out of Mikita, Ovechkin, Esposito and Lafleur will see themselves distanced by Jagr by ~ten spots after this vote is done.

 

[seventieslord]

So to put it in really simple terms, a fair way to account for undercounting would be to take 2.85, divide by 2 because there are two teams in a game, then divide by 2 again, because only half of Brodeur's games were at home. Giving us 0.71 shots per game we should add to Brodeur's totals as saves.

Stats guys, am I doing this right? I don't need it to be surgical, just reasonable and close.

 

[TheDevilMadeMe]

Teams take penalties when they are stuck in their own zone. Brodeur's puck handling absolutely contributed to fewer team penalties.

That doesn't mean it was the only or even most important reason in why the Devils took far fewer penalties than other teams of the era, but it definitely was a factor.

 

[TheDevilMadeMe]

my two cents on brodeur—it feels weird to have him available so soon (presumably) before stevens. i understand why, but btw 1994 and 2003, would anybody consider brodeur the more important player?

.

Brodeur was more important in the regular season. Year after year voted Team MVP by his teammates.

Stevens was more important than Brodeur to the team's long playoff runs, which are probably what a fan of a non divisional rival remembers.

 

[BM67]

my two cents on brodeur—it feels weird to have him available so soon (presumably) before stevens. i understand why, but btw 1994 and 2003, would anybody consider brodeur the more important player?

is vezina-to-end of career brodeur that much better than the first decade of stevens’ career that they would be so far apart on an all-time list? or (as i suspect) might we want to put some pressure on how we prioritize historically significant achievements? not to say that i would have stevens ahead of marty, just that that cognitive dissonance is a decently clear way to illustrate an interesting methodological question.

The Devils players themselves would it seems. He was voted team MVP 8 times from 96-03. List of New Jersey Devils award winners - Wikipedia

 

[VanIslander]

It'll be interesting to see which ones out of Mikita, Ovechkin, Esposito and Lafleur will see themselves distanced by Jagr by ~ten spots after this vote is done.

When you put it that way...

Ovechkin and Lafleur just don't stack up.

So,...

Espo & Stan... >... =... or close but <... JJ > OV & GL.

Some clarity. Thanks.

(Jagr acts here like a standard candle in Astrophysics.)

Here's hoping the 2nd fiddle in each of Chicago and Boston get inducted this round.

 

[Michael Farkas]

but btw 1994 and 2003, would anybody consider brodeur the more important player?

Me. By a considerable margin.

 

[Michael Farkas]

For some reason, I thought the centers project and my own list ( ) had Esposito over Bobby Clarke and I was gonna come in here and kick some garbage cans over (why do we have so many?) and proclaim that Clarke was the better player...then I realized that both the centers project and my own list had Clarke over Espo this whole time...

Well then...let me just quietly straighten up these garbage cans...

 

[seventieslord]

We better not be seriously on the verge of ranking Phil f***ing Esposito above either Stan Mikita or Bobby Clarke.

 

[Michael Farkas]

Yeah, I've now turned my attention to Esposito vs Mikita. I had Mikita waaayyyy higher than Espo on my initial list...but have soured on him in the last couple votes...off to the tape...

In the meantime, I'd like to hear how we feel like Mikita stacks up here...

 

[Canadiens1958]

I believe it works out to about 1 shot per game, and yes, that makes a big difference in save percentage. It could be as many as 3 points in the long run, which, if you look at hockey Outsider's tables, moves a goalie significantly up or down within that cluster.



I'm pretty sure Devils fans have pointed out that the undercounting does exist and it is constant.

This can't be the first time you've heard of this....?



I thought it has been. BM67 has the data, I'm sure.

Heard about it but not a believer in such theories. Giving up two fewer rebounds per game would have the same effect over a season.

 

[MXD]

Brodeur was more important in the regular season. Year after year voted Team MVP by his teammates.

Stevens was more important than Brodeur to the team's long playoff runs, which are probably what a fan of a non divisional rival remembers.

Non divisionnal here...

It's odd because I'm reconciling with the fact that my perception of Brodeur is probably biased by how he played vs. my team. Basically, had Brodeur played like he did against the Canadiens for his whole career, he'd be the best player of all-time, bar none.

Montreal Canadiens

70

45

19

6

128

1818

1690

.930

1.83

9

2

4188:56

63

22

3

[TBODY] [/TBODY]

For once, I'll be letting the Montreal Canadiens perspective prevail over the overall picture.

 

[MXD]

Heard about it but not a believer in such theories. Giving up two fewer rebounds per game would have the same effect over a season.

The issue with his is, of course, that there's no way to prove that Brodeur (or any other goalie) has given rebounds that directly led to two more shots a game. There's also no way to prove that each rebound not given by Brodeur (or any other goalie) didn't result in his team losing the ensuing faceoff and allowing the opposite team one or more shots.