Top-100 Hockey Players of All-Time - Preliminary Discussion Thread (Revenge of Michael Myers)

[Hockey Outsider]

Georges Boucher was ranked in the Top 60 Defensemen project.

Seems like I was "right" with my pick of Frank Fredrickson, whom I prefer over the rest, as much as one can be "right" about it which is subjective.

Good catch - I removed Boucher from my post. Larry Murphy now rounds out the top 3.

I think I'd agree, subjectively, that Fredrickson is the best player not named on the Round 1 aggregate list.

 

[Captain Bowie]

Without looking how many of the all time great F's do you think needed the power play more than Ovechkin at 39%?

What do you mean by needed?

If we're only judging players based on ES production, he still leads all current players in ES goals by a wide margin. In fact you'd have to expand the range all the way back to 98/99 to find someone have more ES goals than him, and it's Iginla with 14 more goals in almost 400 more games.

 

[seventieslord]

I'm probably one of the lowest guys on Terry Sawchuck here, at least when it comes to where he belongs on the pecking order of netminders. I have him out of what's commonly called The Big 7, because I don't think a Top-7 exists in the first place : There's a big-6, and little 4 (of which Sawchuk is part of). He's at best (to me, again) the 3rd most useful member of the dynasty, past Howe and Kelly, and I could probably be convinced to drop him, should a good case to this effect be presented.

But the underlined needs to be corrected (before others chime in). We don't have the numbers for Sawchuk's first five seasons, which were, roughly, his best seasons, and he went AS1 X 3 + AS2 X 2. Those numbers were obtained with good teams, I know. In which case, we must not forget that he played on less-than-optimal teams for the reminder of his prime.

Also, while Sawchuck's numbers don't look that great post-dynasty, just everyone (who saw him play) clearly tought of him as a top netminder (or the top netminder), and I don't think we can quite ignore what people who saw him play (in his prime) had to say (especially in the absence of numbers).

A good tool we have to evaluate Sawchuk is the fact that, post-dynasty, he almost never played a full season. We have many instances where we can look at his personal numbers and compare to the other goaltender(s) on his team. Looking at the quality of the other goaltender and how much he outperformed them or vice versa, we can see if he was really a difference maker after the dynasty. On first glance, he certainly looks like he was "just an NHL goaltender" as soon as he left Detroit the first time, but even if that's true, there were those first five seasons.

I had him 43rd (and actually broke from my tradition of putting Brodeur 6th, at his expense), which I can see wasn't the absolute lowest he was ranked, and the main reasons was that his known numbers are just pretty mediocre.

 

[Canadiens1958]

Give up one goal you should still win. Give up two goals you really should still win or get a point. Give up three... then it gets iffy.

...and at the end of the day it's still largely a team stat.

Don't get me wrong, I've always been a fan of the shutout (getting added points in my hockey pools for a shutout has also increased its popularity for me throughout the years). I just think that giving up just one or two goals a game throughout a large sample of games is better than getting a half dozen shutouts and then letting in large chunks of goals in other games.

Abstraction. Do you have evidence of such strings or goaltenders?

 

[DannyGallivan]

Abstraction. Do you have evidence of such strings or goaltenders?

Here's another way of looking at it... how much importance to we place on shutouts if they are more than a flashy stat? Should we give added weight to career shutouts? If so, Brodeur and Sawchuk move to the front of the class.

 

[Canadiens1958]

I guess we can wonder forever whether or not it's era-induced scoring constipation as well...he moved on from Montreal right around the time the league's sphincters tightened up...

That said, he was just pretty damn good for the loosey goosey time...

Playoff shutouts from 1980 to 1993:
1. Roy - 5
2. Everyone else - less



Really: Barrasso, Hanlon, Penney (!), Sauve, Smith 4
Fuhr, Lindbergh, Moog, Ranford 3

Penney's 4 in 27 GP in Montreal speaks to their defensive conscience I suppose...Lindbergh's 3 in 23 (which really all came in the 18 game run to the '85 Final) are also interesting, who knows what might have happened there, of course...

Does it.

Steve Penney

Steve Penney Stats | Hockey-Reference.com

Brian Hayward

Brian Hayward Stats | Hockey-Reference.com

Hayward, 0 playoff shutouts on a better Canadiens team.

Roy 5 shutouts spanning both to a large degree.

Do not have the time to run the numbers but the best playoff correlation is the length of time from the start of a game to giving up the first goal to the opposition and wins.

Shutouts in this context have significance.

 

[DannyGallivan]

Abstraction. Do you have evidence of such strings or goaltenders?

Roy is widely considered top two in goaltenders among what I've gleaned on these forums, yet he's 15th in career shutouts. He's my example that shutouts are an aesthetic stat.

Note: Roy is second to Brodeur in career playoff shutouts, but he played the most playoff games by a country mile.

 

[Canadiens1958]

Bobby Hull scored 27.7% of his points on the PP. Much lower than I expected to be honest.

Bossy would be interesting given the era and as a counterpoint to Lafleur.

 

[Canadiens1958]

A good tool we have to evaluate Sawchuk is the fact that, post-dynasty, he almost never played a full season. We have many instances where we can look at his personal numbers and compare to the other goaltender(s) on his team. Looking at the quality of the other goaltender and how much he outperformed them or vice versa, we can see if he was really a difference maker after the dynasty. On first glance, he certainly looks like he was "just an NHL goaltender" as soon as he left Detroit the first time, but even if that's true, there were those first five seasons.

I had him 43rd (and actually broke from my tradition of putting Brodeur 6th, at his expense), which I can see wasn't the absolute lowest he was ranked, and the main reasons was that his known numbers are just pretty mediocre.

Other than Glenn Hall very few full seasons were played by O6 goalies once the schedule went to 70 games.

 

[K Fleur]

Bossy would be interesting given the era and as a counterpoint to Lafleur.

33.2% points on the powerplay for Bossy

30.8% points on the powerplay for Lafleur

 

[Canadiens1958]

Here's another way of looking at it... how much importance to we place on shutouts if they are more than a flashy stat? Should we give added weight to career shutouts? If so, Brodeur and Sawchuk move to the front of the class.

Not the question. Short answer is no.

 

[Captain Bowie]

Roy is widely considered top two in goaltenders among what I've gleaned on these forums, yet he's 15th in career shutouts. He's my example that shutouts are an aesthetic stat.

Note: Roy is second to Brodeur in career playoff shutouts, but he played the most playoff games by a country mile.

Roy's also not close to the top for GAA either, or sv%.

 

[Canadiens1958]

Roy's also not close to the top for GAA either, or sv%.

Averaging stats.

 

[DannyGallivan]

Roy's also not close to the top for GAA either, or sv%.

Further proof that more than any other position, it's a slippery slope to judge goalies on their numbers alone.

 

[Captain Bowie]

Further proof that more than any other position, it's a slippery slope to judge goalies on their numbers alone.

Whose saying by the numbers alone? The numbers are useful, given the proper context. Don't see how that's any different for goalies than skaters.

 

[BenchBrawl]

Oh, I'm not at all, just to be clear. I'm bored and wanted to do some data crunching. If anything I'd love to see a study done on special teams specifically as it pertains to individual production, especially now that nhl.com is releasing so much new information.

Most of the big guns in today's game are above 34-35% whereas somebody like Guy Lafleur was at 30.8%....hence why I'm interested in diving further into the fine print in regards to special teams.

Lafleur was better at ES, when the play is less structured.

 

[DannyGallivan]

Whose saying by the numbers alone? The numbers are useful, given the proper context. Don't see how that's any different for goalies than skaters.

I'm not disagreeing.

The main point of my argument was over the relevancy of the shutout stat.

 

[Captain Bowie]

I'm not disagreeing.

The main point of my argument was over the relevancy of the shutout stat.

But it seems you are arguing against a point that no one made.

 

[VanIslander]

So, Sawchuk had some great playoffs. If he would have won two Conn Smythes, that puts his career in Bernie Parent territory. If three or four then Roy type playoffs but without Roy type regular seasons. So, HOW GREAT was Sawchuk to the Wings dynasty?

Given Turk Broda was important to the just prior Leafs dynasty, Sawchuk must have been either head and shoulders more dominant than the skaters in front of him (hard to believe given at least three all time greats from that team will be ranked in the top50 of this project) or else he was the difference maker more often (arguably four Conn Smythes?).

In the late 1970's many older people I knew loved to debate Plante vs. Hall vs. Sawchuk and the word was the greatest goalie ever was one of those three ( in my life no one has broached the idea of any earlier goalies like Vezina, though one non-NHLer (Tretiak) had supporters. Anyways, I just wanna say Sawchuk's argument hinged back then on player accolades of his contributions to Detroit success. The question now is how much MORE was it compared to other alltime greats? Is it enough to make up differences in regular season performances?

 

[seventieslord]

If Ovechkin wins the Maurice Richard Trophy one more time, I'll have no choice but to call him the greatest goal scorer in history. Right now, he's tied with the Golden Jet for seven league titles in that area.

So you don't think Ovechkin winning another Rocket (breaking the tie with Hull for the most all time), scoring more than 900 goals (becoming the undisputed #1 goalscorer of all time), or having another Conn Smythe-worthy performance adds anything to his resume? With any of these things coming to fruition, I don't see how he cannot be in the Top 10. Surely he would overtake Hull and Richard for starters.

I think these posts are based on two flawed ways of thinking:

1. Leading the league in something >>>>> coming second, or third, or fourth. Let's call it the Ricky Bobby mindset.
2. Goals >>>> points.

What Sentinel is saying to me (and what I believe DG is implying), is that if a scenario unfolded this season where Ovechkin was, say, 45-23-68, leading the league in goals while 23rd in the league in points, and Crosby was 32-46-82 and 9th in the league in points this season (both rather realistic projections), that might be enough to propel Ovechkin (who we all apparently agree is behind Crosby currently) ahead of Crosby. Or, maybe it would take such a thing happening two, or three more times. I don't know. Not really important. The problems I see with this are:

1. Goals aren't scored in a vaccuum. They're not independent of everything else. That's why I'll never say anyone but Lemieux and Gretzky are the greatest goal scorers in history. They could lead the league with over a goal per game while also scoring even 23 to 62 MORE assists! What if they only focused on goals? Bobby Hull, Alexander Ovechkin, Maurice Richard, they're not like those guys. They might have more instances leading the league in goals, making them "the most statistically impressive goal scorers" but they're not "the best goal scorers", not in a universe where those two existed. If you said to Ovechkin, "look Alex, the Russian mob placed a massive bet on you scoring 60 assists this season. And if you disappoint them you're dead.", guess what? He'd go out and do it. He's talented enough that he can, even if he never has before. But you better believe he wouldn't be leading the league in goals. He might get 25. On the other hand, we know exactly what happens when Crosby, primarily a playmaking center, decides for whatever reason that he's going to lead the league in goals. He succeeds, but in doing so he posts the worst and 3rd-worst assists-per-game averages of his career. There's a certain level of offense that these guys are able to generate as elite players, and they also have some control over how that offense gets directed from them to the net, but there's a limit to the total . The fact that one is typically balanced about it and the other isn't, should not be seen as a way to pump up the imbalanced player.

2. To carry it further, in an example exactly like, or similar to, the one described above, Crosby would have had the clearly superior season, yet, Ovechkin would earn some sort of mental checkmark in a box you're holding open for him. Should Ovechkin move above Crosby despite not having a better season than him? Does the fact that it was only better in one singular way actually help?

3. Suppose they buck their recent trends and Crosby declines to 68 points while Ovechkin posts 82. Ovechkin clearly had the better season. Are they so close that the fact that a 33-year old Ovechkin having a better season than 31-year old Crosby tips the scales? What if Ovechkin does it with a 36-46-82 statline and is just 5th in goals? Do you care whether he narrows the gap by actually having a better season, or just whether he scores the most goals in the NHL?

4. Crosby is not likely to outdo any of his nine best complete regular seasons this year, and Ovechkin is unlikely to outdo any of his best nine. Let's say the best case scenario plays out and they both post their tenth best seasons. Irrespective of whether Ovechkin leads the league in scoring, or which of the two has the better season, does what they each did in their respective tenth best seasons really matter that much compared to what we know about them as players and how their full careers have played out to this point? Or are you still reading this asking, "well, it depends. Does Ovechkin lead the league in goals???"

don't take a single stat that has no correlation to anything that wins games (saves...nothing to do with wins) and go "oh, look, he's great!" Save pct. is tied to shot volume with a very high correlation. Which tells us, virtually nothing...

I'm not sure about this. I took a look at the 58 goalies from last season and checked the correlation between their save percentage and their shots against per 60. It's -.024. The negative on this is almost meaningless; the main takeaway is that it's practically zero.

If you compare a goalie's save percentage to their win% last year, on the other hand, the correlation is +.604. I'm not a stats major, but I think that counts as somewhat strong.

That's on the aggregate, though. What about game by game?

I took a look at 2,303 instances of a goaltender playing at least 57 minutes in a game last season. I simply checked their shots against and their save percentage. The correlation between the two arrays of data: 0.25. This indicates a positive correlation, but it's quite weak.

Is there some other way to check this that I'm not considering? Because so far I'm not seeing this high correlation you're citing.

 

[seventieslord]

Other than Glenn Hall very few full seasons were played by O6 goalies once the schedule went to 70 games.

I realize that. I'm talking about seasons with fewer than 60 games.

If we're going to look at how Sawchuk or any other goalie performed compared to his teammates, it's not very helpful statistically to look at seasons like '56 or '59, when the other goalies only played as many games as you can count on one hand.

 

[seventieslord]

Roy's also not close to the top for GAA either, or sv%.

When adjusted to the league average, he sure is.

 

[Canadiens1958]

I realize that. I'm talking about seasons with fewer than 60 games.

If we're going to look at how Sawchuk or any other goalie performed compared to his teammates, it's not very helpful statistically to look at seasons like '56 or '59, when the other goalies only played as many games as you can count on one hand.

Now we are getting to the core of the issue.

O6 goalies like Sawchuk,Bower, who were developed traditionally without goalie specific training and the early goalie coached goalies out of Quebec. Plante, the elite of the Mauricie group, Worsley, Gelineau, Johnston etc out of Montreal.

 

[Michael Farkas]

I'm not sure about this. I took a look at the 58 goalies from last season and checked the correlation between their save percentage and their shots against per 60. It's -.024. The negative on this is almost meaningless; the main takeaway is that it's practically zero.

If you compare a goalie's save percentage to their win% last year, on the other hand, the correlation is +.604. I'm not a stats major, but I think that counts as somewhat strong.

That's on the aggregate, though. What about game by game?

I took a look at 2,303 instances of a goaltender playing at least 57 minutes in a game last season. I simply checked their shots against and their save percentage. The correlation between the two arrays of data: 0.25. This indicates a positive correlation, but it's quite weak.

Is there some other way to check this that I'm not considering? Because so far I'm not seeing this high correlation you're citing.

Line is over/under average shots per game. Let's use 30, as it's about right and a nice even number...

30 shots or more
Hasek: .933
Brodeur: .930
Roy: .924
Belfour: .918

29 shots or fewer
Hasek: .916
Brodeur: .910
Roy: .908
Belfour: .906

Part of the 2016-17 season courtesy of Deathstroke
19 or FEWER SHOTS: .859
20 to 29 SHOTS: .909
30 to 39 SHOTS: .929
40 or MORE SHOTS: .935

@Doctor No (noted goalie and statistician and goalie statistician) had some backing evidence of this with his data here: https://hfboards.mandatory.com/threads/nhl-goalies-better-vs-high-shot-volumes.1632645/page-3 but it was lost in the board migration...I'm not sure if he still has it. But I chimed in with my chest out, so it must have backed my notion haha

Whether the magic number really is 30 or if that's just convenient, I have no idea. But I do know that the 20 all-time* best playoff save pct. men in a single playoffs (as far back as new save pct. data goes, I think that's 1953 or thereabouts...obviously the pre-forward pass guys will have a bone to pick...), they don't fair as one might expect...

The 20 best single playoff save percentages (min. 5 games), those guys go a combined 27-15 in playoff series...the 20 best ever. Their per-game winning percentage is even a bit lower than the series win pct.

Even the 20 best GAA playoffs of all time (min. 5 games) tell a vastly different story...those goalies go a combined 41-9. It's just no contest...you can load a wheelbarro full of saves and take it to the scorer's table, it's just not a worth a damn...

 

[ImporterExporter]

What do you mean by needed?

If we're only judging players based on ES production, he still leads all current players in ES goals by a wide margin. In fact you'd have to expand the range all the way back to 98/99 to find someone have more ES goals than him, and it's Iginla with 14 more goals in almost 400 more games.

Needed wasn't the proper word. I'll expand on how I view goal scoring, player to player, later this evening after my kiddo falls asleep.