HOH Top 40 Stanley Cup Playoff Performers of All Time

[bobholly39]

Speaking of Mario...I still think Lemieux should rank higher.

Assuming we don't penalize for missing the playoffs (and that was part of initial discussion thread - we weren't supposed to penalize someone like Lemieux for failing to qualify for the playoffs - only take into account the actual playoffs) - Lemieux has no real weaknesses.

As an offensive geared forward - he produced every single playoff run. His "worst" runs are comparable most player's best runs. His best runs are all-time great runs - arguably the best ever - and he has a few of them, even outside of the 2 smythes. We can talk about 1993 being a bad/disappointing run based on expectations (and mostly based on the fact that he's Mario) - but he had 18 points in 11 games. That's better than almost everyone's *best* run, or just about.

But I tend to value peak a lot. As such I would have wanted him in the top 10. Ironic part is he would have been voted top 10 if I wasn't MIA for vote#2, as I likely would have voted him high, and he finished just outside of top 5 in that round =/

 

[TheDevilMadeMe]

Speaking of Mario...I still think Lemieux should rank higher.

Assuming we don't penalize for missing the playoffs (and that was part of initial discussion thread - we weren't supposed to penalize someone like Lemieux for failing to qualify for the playoffs - only take into account the actual playoffs) - Lemieux has no real weaknesses.

As an offensive geared forward - he produced every single playoff run. His "worst" runs are comparable most player's best runs. His best runs are all-time great runs - arguably the best ever - and he has a few of them, even outside of the 2 smythes. We can talk about 1993 being a bad/disappointing run based on expectations (and mostly based on the fact that he's Mario) - but he had 18 points in 11 games. That's better than almost everyone's *best* run, or just about.

But I tend to value peak a lot. As such I would have wanted him in the top 10. Ironic part is he would have been voted top 10 if I wasn't MIA for vote#2, as I likely would have voted him high, and he finished just outside of top 5 in that round =/

I realize you're a big fan of Lemieux, but isn't Bobby Orr at #13, below Lemieux and Joe Sakic, more egregious?

I mean... Doug Harvey is way over Orr, and that one barely passes the smell test.

 

[MXD]

From 89 to 93, the Pens played 12 series against teams that averaged an 8.5 placing in league GAA and an average conference playoff seeding of 4.75. From 96 to 02, the AVs played 21 series against teams that averaged an 8.8 placing in league GAA and an average conference playoff seeding of 4.6 or so.

The Pens faced slightly better defensive teams in their era and slightly easier overall teams. I see no reason to consider the quality of the opposition as it appears to be quite similar on a relative basis. I am sure the same applies to Crosby.

Am I the only one who thinks one obvious issue with this?

Namely, that the Penguins never played a team that was worse than them defensively, other than in 93?

Not because of the luck of the draw or anything, but because they gave an awful lot of goals, significantly more than Cup-Level teams tended to do in the history of the game (and since then).

 

[MXD]

I realize you're a big fan of Lemieux, but isn't Bobby Orr at #13, below Lemieux and Joe Sakic, more egregious?

I mean... Doug Harvey is way over Orr, and that one barely passes the smell test.

...That's... far from the most outrageous thing involving Bobby Orr.

 

[quoipourquoi]

From 89 to 93, the Pens played 12 series against teams that averaged an 8.5 placing in league GAA and an average conference playoff seeding of 4.75. From 96 to 02, the AVs played 21 series against teams that averaged an 8.8 placing in league GAA and an average conference playoff seeding of 4.6 or so.

The Pens faced slightly better defensive teams in their era and slightly easier overall teams. I see no reason to consider the quality of the opposition as it appears to be quite similar on a relative basis. I am sure the same applies to Crosby.

Your insistence that team rank across leagues of varying size is more accurate than the actual statistical number is fun. In the first season with 30 teams, 2000-01, the 8th ranked Detroit Red Wings were just 18 GA removed from the Jennings (202 vs. 184). In Lemieux's first championship, the 8th ranked team was 53 GA removed from the Jennings (264 vs. 211).

Look at the range for 30 teams in 2016: 192-260 GA.
And look at the 21 team range in 1991: 211-354 GA.

Why throw out the real numbers if we have them? Why operate under the assumption that 8th equals 8th, 16th equals 16th, and 30th equals... well, there's no equivalent for that in a 21-team league, so I guess they must be super-bad.

 

[Michael Farkas]

So I tried to give priority to volume of impact playoff runs first...which is intentionally ambiguous, sure...Lemieux has three, one of which still ended in a major upset loss, notably, without Ron Francis (and looking deeper, perhaps a silly goaltending switch from Wregget back to Barrasso)...Orr has, what, three? I'm not aware of his performance in the 1969 playoffs, I'm not sure I've seen any footage from those playoffs personally...

How they are top 15 is more of a name drop than anything...I think we just felt obligated to put them there because they're greats...but were not more than 11 or 13 players that had four or five impact playoff runs in their careers...? They're both probably too high for my liking...

 

[MXD]

Your insistence that team rank across leagues of varying size is more accurate than the actual statistical number is fun. In the first season with 30 teams, 2000-01, the 8th ranked Detroit Red Wings were just 18 GA removed from the Jennings (202 vs. 184). In Lemieux's first championship, the 8th ranked team was 53 GA removed from the Jennings (264 vs. 211).

Look at the range for 30 teams in 2016: 192-260 GA.
And look at the 21 team range in 1991: 211-354 GA.

Why throw out the real numbers if we have them? Why operate under the assumption that 8th equals 8th, 16th equals 16th, and 30th equals... well, there's no equivalent for that in a 21-team league, so I guess they must be super-bad.

Not to mention, the Avs were sometimes better defensively than the team they're facing... One 8th in 30 might be 7th in 29 or 8th in 29, after all.

 

[quoipourquoi]

So I tried to give priority to volume of impact playoff runs first...which is intentionally ambiguous, sure...Lemieux has three, one of which still ended in a major upset loss, notably, without Ron Francis (and looking deeper, perhaps a silly goaltending switch from Wregget back to Barrasso)...Orr has, what, three? I'm not aware of his performance in the 1969 playoffs, I'm not sure I've seen any footage from those playoffs personally...

How they are top 15 is more of a name drop than anything...I think we just felt obligated to put them there because they're greats...but were not more than 11 or 13 players that had four or five impact playoff runs in their careers...? They're both probably too high for my liking...

I probably started out Round 1 more towards bobholly39/TDMM and ended Round 2 more towards your perspective. Orr vs. Harvey/Robinson... I kinda see the Harvey/Robinson argument.

 

[MXD]

I probably started out Round 1 more towards bobholly39/TDMM and ended Round 2 more towards your perspective. Orr vs. Harvey/Robinson... I kinda see the Harvey/Robinson argument.

I don't even understand why there's a discussion about this to be honest. The numbers were... quite clear : Robinson's career, and the impact he had on his teams, can litterally be wrapped up like this : Orr in the 70ies + Stevens beyond.

 

[quoipourquoi]

I don't even understand why there's a discussion about this to be honest. The numbers were... quite clear : Robinson's career, and the impact he had on his teams, can litterally be wrapped up like this : Orr in the 70ies + Stevens beyond.

To me, it's that idea of what Orr represents (similarly, I was Richard #1). It's hard to shake in a matter of 2-3 weeks in the one project that required me to challenge it. Lemieux, I could still kind of give a pass to, because single-playoff, top-5 playoffs - he's still comfortably between Gretzky and the next challengers of the four-round era. Messier narrowly ahead on quantity that Sakic/Forsberg lack (relative to Messier). But Orr? I could've used a re-do to re-assess relative to Harvey/Robinson. But hey, it's a tough project that required thinking differently, and that's why we did it.

It does make me more comfortable about Howe over Orr overall.

 

[daver]

Your insistence that team rank across leagues of varying size is more accurate than the actual statistical number is fun. In the first season with 30 teams, 2000-01, the 8th ranked Detroit Red Wings were just 18 GA removed from the Jennings (202 vs. 184). In Lemieux's first championship, the 8th ranked team was 53 GA removed from the Jennings (264 vs. 211).

Look at the range for 30 teams in 2016: 192-260 GA.
And look at the 21 team range in 1991: 211-354 GA.

Why throw out the real numbers if we have them? Why operate under the assumption that 8th equals 8th, 16th equals 16th, and 30th equals... well, there's no equivalent for that in a 21-team league, so I guess they must be super-bad.

Did not notice the 21 team vs. 30 team dynamic and you should be well aware that comparing raw goal totals from one era to another needs context, such as their relative placings in the league as we are discussing. I would adjust my numbers accordingly to move the Pens' number to an average of 9th or 10th place.

Not a huge difference, and more importantly, in no way captures a team's actual playoff defensive performance, one that, one would think, has propelled them to playoff success. The North Stars had a 2.75 GAA going into the '91 SCFs, by far the best thru 3 rounds and much better than their 10th place in the regular season.

I think you are introducing variables that are, at best, adding chaos to the stats rather than illuminating. The performance vs. peers number is a lot more real and believable.

 

[MXD]

To me, it's that idea of what Orr represents (similarly, I was Richard #1). It's hard to shake in a matter of 2-3 weeks in the one project that required me to challenge it. Lemieux, I could still kind of give a pass to, because single-playoff, top-5 playoffs - he's still comfortably between Gretzky and the next challengers of the four-round era. Messier narrowly ahead on quantity that Sakic/Forsberg lack (relative to Messier). But Orr? I could've used a re-do to re-assess relative to Harvey/Robinson. But hey, it's a tough project that required thinking differently, and that's why we did it.

It does make me more comfortable about Howe over Orr overall.

I do agree with you that a re-do of the project would bring about significantly more stream-lined results to a significantly lesser amount of blindspots.

 

[quoipourquoi]

The performance vs. peers number is a lot more real and believable.

2016
1st: 192 GA
2nd: 193 GA
3rd: 195 GA
...
12th: 215 GA
13th: 216 GA
14th: 217 GA


1992
1st: 207 GA
2nd: 236 GA
3rd: 244 GA
...
12th: 278 GA
13th: 283 GA
14th: 294 GA

Doesn't feel more real. Feels like rank numbers lose the sense of nuance that GA offers.

 

[bobholly39]

I realize you're a big fan of Lemieux, but isn't Bobby Orr at #13, below Lemieux and Joe Sakic, more egregious?

I mean... Doug Harvey is way over Orr, and that one barely passes the smell test.

You think Orr should be higher than Lemieux? Why?

And this isn't about me being a fan of Lemieux. I just value peak a lot, and as such value Lemieux very highly when it comes to playoff performances.

Lemieux has more longevity. He has less (or none) poor performances. More strong performances. And I think his smythes are stronger than Orr's.

Why would you have Orr above Lemieux?

 

[TheDevilMadeMe]

So I tried to give priority to volume of impact playoff runs first...which is intentionally ambiguous, sure...Lemieux has three, one of which still ended in a major upset loss, notably, without Ron Francis (and looking deeper, perhaps a silly goaltending switch from Wregget back to Barrasso)...Orr has, what, three? I'm not aware of his performance in the 1969 playoffs, I'm not sure I've seen any footage from those playoffs personally...

How they are top 15 is more of a name drop than anything...I think we just felt obligated to put them there because they're greats...but were not more than 11 or 13 players that had four or five impact playoff runs in their careers...? They're both probably too high for my liking...

Well, to explain more on what I said before - I think that if a poster has a problem with Lemieux being too low because of his per-game excellence in the playoffs, the same applies even moreso to Orr.

As for your way of looking at things - it seems awfully binary to me. A player gets full credit for playing great when his team goes far, and zero credit when his team doesn't go far. Now, I do think key players should get credit for their contribution to team success, but this strikes me as taking it a step too far. Not to mention that Orr and Lemieux seem much more important to their respective teams than Doug Harvey, not that Harvey wasn't also important.

I don't even understand why there's a discussion about this to be honest. The numbers were... quite clear : Robinson's career, and the impact he had on his teams, can litterally be wrapped up like this : Orr in the 70ies + Stevens beyond.

The numbers were clear that the Larry Robinson - Serge Savard pairing was at least as effective in a plus/minus sense as the Bobby Orr - Dallas Smith pairing.

I mean, I agree with you that the numbers were something of a revelation for Larry Robinson, and made me think he was probably the 2nd most important member of the dynasty in the playoffs (previously I had Dryden at that spot).

But the numbers don't show that opponents gameplanned around stopping Bobby Orr in a way that, well, they probably did around Guy Lafleur. And they also don't really separate Robinson's contribution from Savard's... now I do agree that Robinson was more important than Savard, but the gap was clearly less than between Orr and Dallas Smith, right?

 

[TheDevilMadeMe]

You think Orr should be higher than Lemieux? Why?

And this isn't about me being a fan of Lemieux. I just value peak a lot, and as such value Lemieux very highly when it comes to playoff performances.

Lemieux has more longevity. He has less (or none) poor performances. More strong performances. And I think his smythes are stronger than Orr's.

Why would you have Orr above Lemieux?

I think Orr's best was better than Lemieux's best, and his worst was also better than Lemieux's worst.

I mean, yes, Orr wasn't very good in his own end in 1971, but Lemieux was rarely good in his own end.

That is just my personal opinion though.

To put it another way, I think that one could reasonably prefer either Orr or Lemieux in the playoffs. Which is why I thought it was strange that your criticism was just about Lemieux being too low for reasons that also apply to Orr.

 

[daver]

2016
1st: 192 GA
2nd: 193 GA
3rd: 195 GA
...
12th: 215 GA
13th: 216 GA
14th: 217 GA


1992
1st: 207 GA
2nd: 236 GA
3rd: 244 GA
...
12th: 278 GA
13th: 283 GA
14th: 294 GA

Doesn't feel more real. Feels like rank numbers lose the sense of nuance that GA offers.

Incorporating stats from outside the parameters of this project feels more real than the actual playoff stats themselves?

Two questions come to mind:

What is different about the playoffs that a comparison of GAAs from different eras become relevant in player comparisons when the majority of the HOH, at best, treat the use of adjusted #'s for regular season stats with a ton of context?

What are the playoff GAs and why wouldn't those be even more real?

I really don't get this angle. When Mario put up his numbers in 91 and 92, was the narrative that he did it against bad defensive teams therefore his numbers were inflated? That doesn't seem reasonable given how close he was to Wayne at the time.

I am assuming Wayne's #'s numbers would take a similar tumble along with everyone else's from the era where the league GAA was at its historical highest. I really thought that use of league GAA to adjust #'s rather than look at performance vs. peers was frowned upon in the HOH. I guess not.

 

[MXD]

The numbers were clear that the Larry Robinson - Serge Savard pairing was at least as effective in a plus/minus sense as the Bobby Orr - Dallas Smith pairing.

I mean, I agree with you that the numbers were something of a revelation for Larry Robinson, and made me think he was probably the 2nd most important member of the dynasty in the playoffs (previously I had Dryden at that spot).

But the numbers don't show that opponents gameplanned around stopping Bobby Orr in a way that, well, they probably did around Guy Lafleur. And they also really separate Robinson's contribution from Savard's... now I do agree that Robinson was more important than Savard, but the gap was clearly less than between Orr and Dallas Smith, right?

Underlined : I don't remember the exact set of numbers that leads you to this conclusion, but I don't know if I should reply that it's a little tougher to plan around a D-Men who is deployed 30mins+/game than it is to plan around a forward who is deployed 20mins+/game, or that we shouldn't assume NHL coaches to be total morons who completely disregarded the fact that facing the Boston Bruins involved facing Bobby Orr, and that adjustments had to be made.

Bold : Yeah, but :
- Robinson and Savard were not always deployed together and it appears to not have affected Robinson in the slightest (in fact, when not together, Robinson was better);
- The gap between Player X and Dallas Smith will always be bigger than the gap between Player Y and Serge Savard, provided that player X and Y are both better than Serge Savard. In other words, I don't think it's particularily relevant when assessing Robinson vs. Orr, especially in view of the fact that Robinson would've been good enough to make this list if you completely take out everything that happened before 1980 (and just give him a random set of games where he played some solid hockey)
- Yes, Savard indirectly helped Robinson more than Smith helped Orr (that's obvious, Savard is an all-time great and Smith was a solid journeyman), but there's also a very significant gap between Robinson and Orr in terms of team results, too.


And mostly, there's a huge longevity difference between both players...

EDIT : I'll go further : There's no way I'm putting Lafleur ahead of Robinson now. I was already seeing not much distinction during the project before the numbers were disclosed (see V3) and now I see a clear one... Favouring Robinson. Lafleur was probably the best player of the dynasty (Robinson can absolutely be argued though), but Robinson added 10 or 11 years of excellent playoff performance afterwards.

 

[daver]

Am I the only one who thinks one obvious issue with this?

Namely, that the Penguins never played a team that was worse than them defensively, other than in 93?

Not because of the luck of the draw or anything, but because they gave an awful lot of goals, significantly more than Cup-Level teams tended to do in the history of the game (and since then).

What would be the significance of this?

 

[MXD]

What would be the significance of this?

None if you go with QPQ's preferred method and key on GA's, or way or another, instead of ranking.

The 8th best defensive team (out of 21) might actually be the 8th best team (out of 20) for the purposes of assessing the opponent defense of the Penguins.

Take, per example, the Hawks in '97, that the Avalanche faced.
They were 7th out of 26. But one of these teams were the Avalanche themselves, so 6th out of 25 seems a slightly more appropriate descriptor, considering Sakic could obviously not face his own team.

Basically, it was nitpicking. And a good reason why GA's-as-a-ratio-of-league-average seems a preferable method of assessing opponents' defense.

 

[daver]

None if you go with QPQ's preferred method and key on GA's, or way or another, instead of ranking.

The 8th best defensive team (out of 21) might actually be the 8th best team (out of 20) for the purposes of assessing the opponent defense of the Penguins.

Take, per example, the Hawks in '97, that the Avalanche faced.
They were 7th out of 26. But one of these teams were the Avalanche themselves, so 6th out of 25 seems a slightly more appropriate descriptor, considering Sakic could obviously not face his own team.

Basically, it was nitpicking. And a good reason why GA's-as-a-ratio-of-league-average seems a preferable method of assessing opponents' defense.

Besides not capturing a team's actual playoff defensive performance, there is also the neglect to capture the tightening up of scoring that happens in the playoffs. The league-wide GAA is usually clearly less than the regular season GAA.

I guess I don't see this method as being something other than a marginal consideration; certainly not one that should be used to significantly rearrange numbers based on performance vs. peers.

Mario's PPG from 89 to 93 would roughly translate to 1.50 to 1.55 if he played in Sakic/Forsberg's era (they were 1.12 to 1.18). Jagr's was 1.23ish at that same time.
Or vice versa, Sakic and Forsberg would be putting up a 1.50ish PPG in Mario's era which would be third best just behind Wayne.

Does that not sound like a more reasonable scenario than Mario being just ahead of Sakic?

 

[daver]

These numbers do not pass the eye test, IMO. A performance vs. peers analysis shows a much bigger gap between Mario and Sakic/Forsberg; a gap that one would expect given the gap in their regular season performances and when one thinks about how dominant Mario was in his two Cups runs.

From 89 - 93, which comprises 4 of Mario's 5 top runs as noted above, his PPG of 1.92 is 73% better than the average PPG of the next best 14 scorers (less Wayne) of that time period.

http://www.nhl.com/stats/player?agg...mesPlayed,gte,1&sort=points,goals,gamesPlayed

From 96 - 01, which comprises 4 of Sakic's 5 top runs as noted above, his PPG of 1.16 is 34% better than the average PPG of the next best 14 scorers of that time period.

http://www.nhl.com/stats/player?agg...mesPlayed,gte,1&sort=points,goals,gamesPlayed

From 96 - 02, which comprises 4 of Forsberg's 5 top runs as noted above, his PPG of 1.18 is 38% better than the average PPG of the next best 14 scorers of that time period.

The gap between Mario and Sakic and Forsberg using this method is 29% (173/1.34)and 26% respectively vs. 10% (1.37/1.25) and 23% respectively from the method used above. I think using regular season team GAAs makes zero sense. Look no further than this year when the Preds went from being a slightly below average defensive team in the regular season to the best playoff defensive team by far thru 3 rounds before running into the Pens.

The numbers for the previous post are above.

 

[MXD]

(...)

I guess I don't see this method as being something other than a marginal consideration; certainly not one that should be used to significantly rearrange numbers based on performance vs. peers.

(...)

It doesn't rearrange numbers based on performance vs. peers : it actually gives an even clearer number than performance vs. peers (as far as the opponent defense is concerned).

 

[Dennis Bonvie]

You think Orr should be higher than Lemieux? Why?

And this isn't about me being a fan of Lemieux. I just value peak a lot, and as such value Lemieux very highly when it comes to playoff performances.

Lemieux has more longevity. He has less (or none) poor performances. More strong performances. And I think his smythes are stronger than Orr's.

Why would you have Orr above Lemieux?

Based on what, points scored?

Mario's 2 Smythe's:

23 16-28-44 +14
15 16-18-34 +6

Orr's 2 Smythe's:

14 9-11-20 +24
15 5-19-24 +20

 

[daver]

It doesn't rearrange numbers based on performance vs. peers : it actually gives an even clearer number than performance vs. peers (as far as the opponent defense is concerned).

I get it is supposed to potentially put more value for playing against an "on paper" better defensive opponent but the method is full of holes and faulty adjusting.

- it disregards an opponent's actual playoff defensive performance while presuming that the opponent has played up to the standard set in the regular season

- it uses a GA metric from different seasons with no context which is akin to comparing Yzerman's 155 point season straight up with any other forward's season or "adjusting" Bossy's 147 point season in 1981 (the 3rd highest non-Wayne/Mario point total of the 1980's) to 108 points, good for a tie for 6th place in 1999 with Yashin.

- using RS league GAs also ignores the fact that the playoffs usually sees a clear drop in league-wide GAs so use of RS GAs doesn't make much sense

If the goal is to "adjust" the numbers to represent the different scoring environments each one played in, performance vs. peers does exactly that. If the goal is to reward a player for putting up a similar offensive performance vs. their peers against clearly better defensive opposition (relative to their era), I could see that as being a tiebreaker for players with similar offensive resumes but that's not the case with Mario and Sakic/Forsberg. IMO, I would need a lot more evidence than simply league GAs to establish a clear difference in defensive opposition.

I again would look at the opposition's placing in league GAAs with consideration for the amount of total teams in the league. I would presume that the team that finished 8th in 1991 would finish at the same position in 2001 regardless of how many raw goals they were from the mean average or the #1 defensive team. That is making way too many assumptions for a method that is making a bunch of assumptions already. In our example, the Pens faced slightly higher seeded teams and slightly higher placed defensive teams.

Look at this past playoffs to see the flaws in this method. The Pens faced the # 1 and #2 RS defensive teams in Round 1 and 2. They put up 4.2 goals a game against the #2 team and almost 3 a game against the #1 team. The #15 RS defensive team, the Preds, came into the SCF with a GAA of 1.81. The Pens put up just over 3 a game against them. The #3 RS defensive team, the Ducks, had the worst GAA of the eight QF teams. This shows no correlation between RS performance and playoff performance.