Post-consolidation VsX Benchmarks

[Hockey Outsider]

I've done something slightly different. Here's a table showing the top 250 players using the original benchmarks:

1. Ten year VsX score (unweighted), using original seasonal benchmarks
2. Ten year VsX score (unweighted), using original revised benchmarks
3. The different (in "units", rather than percentages) between the two methods

Results:
- 175 players were unaffected
- 38 players' scores were affected by less than 1.0 units
- 17 players' scores were affected by between 1.0 and 2.0 units
- 16 players' scores were affected by 2.0 and 3.0 units: Berenson, Hadfield, Martin, Gilbert, D. Hull, J. Pronovost, Stanfield, Martin, Unger, Lemaire, Cashman, Park, F. Mahovlich, P. Mahovlich, Cournoyer and Bucyk
- 4 players were impacted by more than that: Clarke (3.0), Hodge (3.1), Orr (4.7) and Esposito (5.3)

Player | Old | New | Difference
Wayne Gretzky | 144 | 144.7 | 0.7
Gordie Howe | 118.1 | 118.1 | 0
Phil Esposito | 112.4 | 117.6 | 5.3
Mario Lemieux | 112 | 112 | 0
Jaromir Jagr | 105.9 | 105.9 | 0
Stan Mikita | 101.7 | 102.5 | 0.8
Bobby Hull | 100.2 | 101.8 | 1.7
Jean Beliveau | 99.8 | 100 | 0.2
Maurice Richard | 97.9 | 97.9 | 0
Marcel Dionne | 96.3 | 97.5 | 1.1
Ted Lindsay | 95.8 | 95.8 | 0
Andy Bathgate | 95.6 | 95.6 | 0
Joe Sakic | 94 | 94 | 0
Bobby Orr | 87.4 | 92.1 | 4.7
Guy Lafleur | 89.4 | 91 | 1.5
Mike Bossy | 89 | 89.4 | 0.4
Joe Thornton | 89.2 | 89.2 | 0
Howie Morenz | 88.9 | 88.9 | 0
Frank Boucher | 88.8 | 88.8 | 0
Steve Yzerman | 88.8 | 88.8 | 0
Teemu Selanne | 88.7 | 88.7 | 0
Bryan Trottier | 86.9 | 87.2 | 0.4
Nels Stewart | 87.1 | 87.1 | 0
Bill Cowley | 86.7 | 86.7 | 0
Martin St. Louis | 86.3 | 86.3 | 0
Adam Oates | 86.2 | 86.2 | 0
Jean Ratelle | 83.4 | 85.3 | 1.9
Mark Messier | 84.7 | 85 | 0.3
Norm Ullman | 83.3 | 84.7 | 1.3
Bill Cook | 84.4 | 84.4 | 0
Mark Recchi | 84.3 | 84.3 | 0
Ron Francis | 84.2 | 84.2 | 0
Frank Mahovlich | 81.4 | 84.1 | 2.7
John Bucyk | 81 | 83.9 | 2.9
Syl Apps Sr | 83.9 | 83.9 | 0
Brett Hull | 83.6 | 83.6 | 0
Bobby Clarke | 80.4 | 83.5 | 3
Bernie Geoffrion | 83.4 | 83.4 | 0
Alex Delvecchio | 82.9 | 82.9 | 0
Alex Ovechkin | 82.8 | 82.8 | 0
Charlie Conacher | 82.5 | 82.5 | 0
Peter Forsberg | 82.1 | 82.1 | 0
Paul Coffey | 81.6 | 82 | 0.4
Marty Barry | 81.8 | 81.8 | 0
Dale Hawerchuk | 81.4 | 81.7 | 0.3
Gilbert Perreault | 79.6 | 81.6 | 1.9
Toe Blake | 81.4 | 81.4 | 0
Jarome Iginla | 81.3 | 81.3 | 0
Jari Kurri | 80.8 | 81.2 | 0.4
Luc Robitaille | 80.9 | 80.9 | 0
Peter Stastny | 80.4 | 80.8 | 0.4
Darryl Sittler | 78.8 | 80.4 | 1.6
Busher Jackson | 80.2 | 80.2 | 0
Max Bentley | 80 | 80 | 0
Mats Sundin | 79.9 | 79.9 | 0
Rod Gilbert | 77.6 | 79.7 | 2.2
Sweeney Schriner | 79.7 | 79.7 | 0
Elmer Lach | 79.2 | 79.2 | 0
Pierre Turgeon | 78.9 | 78.9 | 0
Henri Richard | 78.9 | 78.9 | 0
Milt Schmidt | 78.8 | 78.8 | 0
Paul Kariya | 78.6 | 78.6 | 0
Sidney Crosby | 78.6 | 78.6 | 0
Denis Savard | 77.7 | 78 | 0.3
Theoren Fleury | 77.9 | 77.9 | 0
Mike Modano | 77.7 | 77.7 | 0
Syd Howe | 77.7 | 77.7 | 0
Daniel Alfredsson | 77.4 | 77.4 | 0
Marian Hossa | 77.2 | 77.2 | 0
Sergei Fedorov | 77.1 | 77.1 | 0
Doug Gilmour | 77.1 | 77.1 | 0
Ilya Kovalchuk | 76.9 | 76.9 | 0
Doug Bentley | 76.7 | 76.7 | 0
Jeremy Roenick | 76.1 | 76.1 | 0
Eric Lindros | 76 | 76 | 0
Sid Abel | 75.8 | 75.8 | 0
Henrik Sedin | 75.4 | 75.4 | 0
Aurel Joliat | 75.1 | 75.1 | 0
Brendan Shanahan | 74.4 | 74.4 | 0
Pavel Datsyuk | 74.4 | 74.4 | 0
Dickie Moore | 74.3 | 74.3 | 0
Roy Conacher | 74.1 | 74.1 | 0
Bernie Federko | 73.6 | 74 | 0.4
Ken Hodge | 70.9 | 74 | 3.1
Keith Tkachuk | 73.9 | 73.9 | 0
Ted Kennedy | 73.8 | 73.8 | 0
Bernie Nicholls | 73.4 | 73.7 | 0.3
Patrik Elias | 73.6 | 73.6 | 0
Brad Richards | 73.2 | 73.2 | 0
Markus Naslund | 73 | 73 | 0
Alexander Mogilny | 72.9 | 72.9 | 0
Pavel Bure | 72.7 | 72.7 | 0
Dany Heatley | 72.3 | 72.3 | 0
Daniel Sedin | 72.3 | 72.3 | 0
Doug Weight | 72.2 | 72.2 | 0
Michel Goulet | 71.5 | 71.8 | 0.4
Pat LaFontaine | 71.8 | 71.8 | 0
Jacques Lemaire | 69.1 | 71.7 | 2.5
Raymond Bourque | 71.2 | 71.5 | 0.3
Yvan Cournoyer | 68.7 | 71.5 | 2.8
Bryan Hextall | 70.8 | 71.3 | 0.5
Eric Staal | 71.3 | 71.3 | 0
John LeClair | 71.2 | 71.2 | 0
Vincent Damphousse | 71.1 | 71.1 | 0
Vincent Lecavalier | 71 | 71 | 0
Alexei Yashin | 70.8 | 70.8 | 0
Dave Keon | 69.8 | 70.7 | 1
Red Kelly | 70.5 | 70.5 | 0
Henrik Zetterberg | 69.8 | 69.8 | 0
Bill Mosienko | 69.8 | 69.8 | 0
Hooley Smith | 69.5 | 69.5 | 0
Ziggy Palffy | 69.4 | 69.4 | 0
Cooney Weiland | 69.3 | 69.3 | 0
Bert Olmstead | 69.2 | 69.2 | 0
Bobby Rousseau | 67.9 | 69.1 | 1.2
Lorne Carr | 69 | 69 | 0
Dit Clapper | 69 | 69 | 0
Paul Thompson | 68.9 | 68.9 | 0
Brian Leetch | 68.8 | 68.8 | 0
Alex Kovalev | 68.7 | 68.7 | 0
Phil Watson | 68.7 | 68.7 | 0
Bun Cook | 68.7 | 68.7 | 0
Rod Brind'Amour | 68.6 | 68.6 | 0
Patrick Marleau | 68.5 | 68.5 | 0
Ray Whitney | 68.5 | 68.5 | 0
Johnny Gottselig | 68.5 | 68.5 | 0
Denis Potvin | 67 | 68.4 | 1.4
Steve Larmer | 67.9 | 68.2 | 0.2
Lanny McDonald | 67.8 | 68 | 0.2
Lynn Patrick | 67.9 | 67.9 | 0
Pavol Demitra | 67.7 | 67.7 | 0
Phil Goyette | 66.9 | 67.6 | 0.7
Woody Dumart | 67.5 | 67.5 | 0
Pete Mahovlich | 64.8 | 67.5 | 2.7
Tony Amonte | 67.5 | 67.5 | 0
Rick Middleton | 66.9 | 67.2 | 0.3
Cecil Dillon | 67.1 | 67.1 | 0
Jason Spezza | 67 | 67 | 0
Dino Ciccarelli | 66.7 | 66.9 | 0.2
Clint Smith | 66.8 | 66.8 | 0
Don McKenney | 66.8 | 66.8 | 0
Joe Mullen | 66.3 | 66.6 | 0.3
Peter Bondra | 66.6 | 66.6 | 0
Milan Hejduk | 66.5 | 66.5 | 0
Bobby Smith | 66 | 66.3 | 0.2
Joe Nieuwendyk | 66.3 | 66.3 | 0
Alex Tanguay | 66.3 | 66.3 | 0
Dave Andreychuk | 66 | 66.2 | 0.3
Bill Barber | 64.8 | 66.2 | 1.4
Al MacInnis | 66.1 | 66.1 | 0
Mike Gartner | 65.8 | 66 | 0.3
Herbie Lewis | 66 | 66 | 0
Glenn Anderson | 65.4 | 65.7 | 0.3
Nicklas Lidstrom | 65.6 | 65.6 | 0
Mike Ribeiro | 64.9 | 64.9 | 0
Bill Thoms | 64.8 | 64.8 | 0
Brian Propp | 64.4 | 64.7 | 0.3
Pit Martin | 62.3 | 64.6 | 2.3
Garry Unger | 61.7 | 64.1 | 2.4
Dave Taylor | 63.8 | 64 | 0.2
Marc Savard | 64 | 64 | 0
Dean Prentice | 63.9 | 63.9 | 0
Evgeni Malkin | 63.8 | 63.8 | 0
Ebbie Goodfellow | 63.7 | 63.7 | 0
Dennis Maruk | 63.3 | 63.5 | 0.2
Rick MacLeish | 61.8 | 63.3 | 1.5
Ryan Getzlaf | 63.3 | 63.3 | 0
Rick Martin | 61 | 63.2 | 2.2
Scott Gomez | 63.2 | 63.2 | 0
Marian Gaborik | 62.7 | 62.7 | 0
Shane Doan | 62.5 | 62.5 | 0
Olli Jokinen | 62.4 | 62.4 | 0
Phil Housley | 62.2 | 62.4 | 0.3
Eddie Wiseman | 62.3 | 62.3 | 0
Tom Lysiak | 60.8 | 62.1 | 1.3
Owen Nolan | 62.1 | 62.1 | 0
Tod Sloan | 62 | 62 | 0
Miroslav Satan | 62 | 62 | 0
Steve Shutt | 60.8 | 61.6 | 0.8
Larry Aurie | 61.5 | 61.5 | 0
Butch Goring | 59.9 | 61.4 | 1.5
Alex Zhamnov | 61.4 | 61.4 | 0
Brian Bellows | 61 | 61.3 | 0.3
Herb Cain | 61.3 | 61.3 | 0
Kenny Wharram | 61.1 | 61.1 | 0
Rick Nash | 61.1 | 61.1 | 0
Wayne Cashman | 58.2 | 60.8 | 2.6
Gordie Drillon | 60.7 | 60.7 | 0
Steve Sullivan | 60.7 | 60.7 | 0
Dennis Hull | 58.2 | 60.4 | 2.2
Craig Janney | 60.3 | 60.3 | 0
Pat Verbeek | 60.3 | 60.3 | 0
Rene Robert | 58.5 | 60 | 1.5
Vic Hadfield | 57.8 | 59.9 | 2.1
Todd Bertuzzi | 59.9 | 59.9 | 0
Murray Oliver | 59.5 | 59.9 | 0.4
Pierre Larouche | 59.3 | 59.8 | 0.5
Petr Sykora | 59.7 | 59.7 | 0
Robert Lang | 59.7 | 59.7 | 0
Jimmy Ward | 59.7 | 59.7 | 0
Jean Pronovost | 57.4 | 59.6 | 2.2
Bob Nevin | 57.8 | 59.4 | 1.6
Ryan Smyth | 59.3 | 59.3 | 0
Claude Provost | 59.3 | 59.3 | 0
Buddy O'Connor | 59.2 | 59.2 | 0
Martin Straka | 59.1 | 59.1 | 0
Camille Henry | 58.9 | 58.9 | 0
Red Berenson | 56.7 | 58.8 | 2.1
Kirk Muller | 58.8 | 58.8 | 0
Ivan Boldirev | 57.4 | 58.8 | 1.4
Johnny Gagnon | 58.7 | 58.7 | 0
Brad Park | 56 | 58.7 | 2.6
Jason Arnott | 58.6 | 58.6 | 0
Petr Nedved | 58.6 | 58.6 | 0
Vyacheslav Kozlov | 58.6 | 58.6 | 0
Fred Stanfield | 56.3 | 58.5 | 2.3
Cliff Ronning | 58.5 | 58.5 | 0
Cory Stillman | 58.5 | 58.5 | 0
Andrew Brunette | 58.3 | 58.3 | 0
Neal Broten | 57.9 | 58.2 | 0.3
Sergei Gonchar | 58.2 | 58.2 | 0
Larry Murphy | 58.1 | 58.1 | 0
Peter McNab | 57.9 | 58.1 | 0.2
Wilf Paiement | 57.7 | 57.9 | 0.2
Bill Guerin | 57.8 | 57.8 | 0
Gary Roberts | 57.7 | 57.7 | 0
George Armstrong | 57.6 | 57.6 | 0
Jozef Stumpel | 57.6 | 57.6 | 0
Eddie Shore | 57.5 | 57.5 | 0
Steve Thomas | 57.4 | 57.4 | 0
Thomas Vanek | 57.3 | 57.3 | 0
Chris Drury | 57.2 | 57.2 | 0
Kevin Stevens | 57.2 | 57.2 | 0
Rick Tocchet | 57 | 57 | 0
Tomas Sandstrom | 56.7 | 56.7 | 0
Mike Ridley | 56.7 | 56.7 | 0
Anze Kopitar | 56.7 | 56.7 | 0
Saku Koivu | 56.7 | 56.7 | 0
Corey Perry | 56.7 | 56.7 | 0
Andrew Cassels | 56.5 | 56.5 | 0
Michael Nylander | 56.3 | 56.3 | 0
Bob Pulford | 55.8 | 56.2 | 0.4
Gus Bodnar | 56.2 | 56.2 | 0
Sergei Zubov | 56.1 | 56.1 | 0
Hec Kilrea | 56.1 | 56.1 | 0
Dave Gagner | 56 | 56 | 0
Robert Reichel | 55.9 | 55.9 | 0
Greg Adams | 55.9 | 55.9 | 0
Ray Ferraro | 55.7 | 55.7 | 0
Brian Rolston | 55.6 | 55.6 | 0

 

[TheDevilMadeMe]

That crafty Ken Hodge. Gets his stats inflated by playing with the Orr/Esposito duo, then gets a favorable benchmark on top of it for helping to break the system!

 

[Sturminator]

Thanks for this work, HO. The positive adjustments for non-Bruins affected look really quite good to me.

The Bruins, themselves...sigh...I just don't know. I sort of think we should use a straight Vs2 system for Bruins-internal scoring during this period. If all those other pluggers in yellow and black were benefitting so much from playing with Orr and Esposito, why not just compare them to Orr and Esposito? But this may be too harsh.

Asterisks all around!

 

[Hockey Outsider]

It's a tricky question, and I don't have a good answer. The 1970s Bruins seem to break the system (as it currently exists), no matter how you look at it.

One of the main appeals of the system is its simplicity. It seems like we need to make a trade-off between accuracy and simplicity. (In other words, do we increase the accuracy for the Bruins by adding more rules, which detracts from the system's simplicity and appeal, or do we keep it simple and make a mental, qualitative adjustment for the Bruins?) I'll need to give it further thought.

 

[Sturminator]

I think you're right about the problem of simplicity vs. accuracy, HO.

I joked about asterisks, but my feeling on the matter is that the proposed "Orr Rule" is enough of a change to handle the problem. The issue of accuracy in this era regarding the entire player universe outside of Boston is clearly one worth addressing, but whether or not we have a precise accounting of guys like Hodge and Cashman seems vastly less important.

I think the single rule change is the way to go. It's really all for naught if the system becomes an incomprehensible black box.

 

[ted2019]

That crafty Ken Hodge. Gets his stats inflated by playing with the Orr/Esposito duo, then gets a favorable benchmark on top of it for helping to break the system!

That's why I drafted him in the ATD.

 

[Sturminator]

After much deliberation, I have decided to incorporate the "Orr Rule" into the VsX system without any further tinkering for Bruins-internal scoring. I have updated to the OP to reflect this.

 

[steve141]

After much deliberation, I have decided to incorporate the "Orr Rule" into the VsX system without any further tinkering for Bruins-internal scoring. I have updated to the OP to reflect this.

I've read through the discussion it but don't quite understand. Could you clarify, how does the Orr rule work, and what are we hoping to accomplish by using it?

 

[jigglysquishy]

Is the benchmark for 2014-15 going to be Tavare's 86 points or Crosby's 84?

 

[Sturminator]

I've read through the discussion it but don't quite understand. Could you clarify, how does the Orr rule work, and what are we hoping to accomplish by using it?

The Orr Rule works by setting an exception in years when a defenseman is equal to or greater than the benchmark. The defenseman's scoring is removed for the purposes of setting the benchmark, and the scoring of all forwards on his team above the benchmark is normalized to that of a roughly average #1 scorer (that is, the benchmark +5%).

The reason we need this rule is to smooth out a few seasons in the early 1970's when the post-expansion Bruins broke the system, and made the benchmark too high. Basically, the idea is that peak Bobby Orr was such an outlier in the history of hockey that the benchmarks set by his teams need an adjustment. The Orr Rule appears to set benchmarks which are much fairer to the rest of the league, though it will inflate the VsX scores of Bruins players in this period even more than they already were (and thus the VsX scores of guys like Hodge and Cashman will have to carry something of an asterisk).

The rule also catches the single season when Paul Coffey was the #2 scorer in the league, and sets the benchmark to Goulet at #3 rather than Coffey at #2 (where the previous benchmark had been). Coffey's scoring in this season is also a historical outlier, so the rule sort of kills two birds with one stone in this sense.

 

[Sturminator]

Is the benchmark for 2014-15 going to be Tavare's 86 points or Crosby's 84?

Tavares' 86 points.

 

[Hockey Outsider]

All-Time Results Through 2015 - best ten years

Rank | Player | Result
1 | Wayne Gretzky | 144.7
2 | Gordie Howe | 118.1
3 | Phil Esposito | 117.6
4 | Mario Lemieux | 112
5 | Jaromir Jagr | 105.9
6 | Stan Mikita | 102.5
7 | Bobby Hull | 101.8
8 | Jean Beliveau | 100
9 | Maurice Richard | 97.9
10 | Marcel Dionne | 97.5
11 | Ted Lindsay | 95.8
12 | Andy Bathgate | 95.6
13 | Joe Sakic | 94
14 | Alex Ovechkin | 92.2
15 | Bobby Orr | 92.1
16 | Guy Lafleur | 91
17 | Mike Bossy | 89.4
18 | Joe Thornton | 89.2
19 | Howie Morenz | 88.9
20 | Frank Boucher | 88.8
21 | Steve Yzerman | 88.8
22 | Teemu Selanne | 88.7
23 | Sidney Crosby | 88.3
24 | Bryan Trottier | 87.2
25 | Nels Stewart | 87.1
26 | Bill Cowley | 86.7
27 | Martin St. Louis | 86.3
28 | Adam Oates | 86.2
29 | Jean Ratelle | 85.3
30 | Mark Messier | 85
31 | Norm Ullman | 84.7
32 | Bill Cook | 84.4
33 | Mark Recchi | 84.3
34 | Ron Francis | 84.2
35 | Frank Mahovlich | 84.1
36 | John Bucyk | 83.9
37 | Syl Apps Sr | 83.9
38 | Brett Hull | 83.6
39 | Bobby Clarke | 83.5
40 | Bernie Geoffrion | 83.4
41 | Alex Delvecchio | 82.9
42 | Charlie Conacher | 82.5
43 | Peter Forsberg | 82.1
44 | Paul Coffey | 82
45 | Marty Barry | 81.8
46 | Dale Hawerchuk | 81.7
47 | Gilbert Perreault | 81.6
48 | Jarome Iginla | 81.5
49 | Toe Blake | 81.4
50 | Jari Kurri | 81.2
51 | Luc Robitaille | 80.9
52 | Peter Stastny | 80.8
53 | Darryl Sittler | 80.4
54 | Busher Jackson | 80.2
55 | Max Bentley | 80
56 | Mats Sundin | 79.9
57 | Rod Gilbert | 79.7
58 | Sweeney Schriner | 79.7
59 | Elmer Lach | 79.2
60 | Henrik Sedin | 79.1
61 | Pierre Turgeon | 78.9
62 | Henri Richard | 78.9
63 | Milt Schmidt | 78.8
64 | Paul Kariya | 78.6
65 | Marian Hossa | 78.1
66 | Denis Savard | 78
67 | Theoren Fleury | 77.9
68 | Mike Modano | 77.7
69 | Syd Howe | 77.7
70 | Daniel Alfredsson | 77.4
71 | Sergei Fedorov | 77.1
72 | Doug Gilmour | 77.1
73 | Pavel Datsyuk | 77.1
74 | Ilya Kovalchuk | 76.9
75 | Doug Bentley | 76.7
76 | Jeremy Roenick | 76.1
77 | Eric Lindros | 76
78 | Daniel Sedin | 75.8
79 | Sid Abel | 75.8
80 | Aurel Joliat | 75.1
81 | Brendan Shanahan | 74.4
82 | Dickie Moore | 74.3
83 | Roy Conacher | 74.1
84 | Eric Staal | 74
85 | Bernie Federko | 74
86 | Ken Hodge | 74
87 | Keith Tkachuk | 73.9
88 | Ted Kennedy | 73.8
89 | Bernie Nicholls | 73.7
90 | Patrik Elias | 73.6
91 | Brad Richards | 73.2
92 | Markus Naslund | 73
93 | Alexander Mogilny | 72.9
94 | Pavel Bure | 72.7
95 | Henrik Zetterberg | 72.5
96 | Dany Heatley | 72.3
97 | Jason Spezza | 72.2
98 | Doug Weight | 72.2
99 | Evgeni Malkin | 72
100 | Michel Goulet | 71.8
101 | Pat LaFontaine | 71.8
102 | Jacques Lemaire | 71.7
103 | Raymond Bourque | 71.5
104 | Yvan Cournoyer | 71.5
105 | Ryan Getzlaf | 71.4
106 | Bryan Hextall | 71.3
107 | John LeClair | 71.2
108 | Vincent Damphousse | 71.1
109 | Vincent Lecavalier | 71
110 | Alexei Yashin | 70.8
111 | Dave Keon | 70.7
112 | Red Kelly | 70.5
113 | Bill Mosienko | 69.8
114 | Patrick Marleau | 69.7
115 | Hooley Smith | 69.5
116 | Ziggy Palffy | 69.4
117 | Cooney Weiland | 69.3
118 | Bert Olmstead | 69.2
119 | Bobby Rousseau | 69.1
120 | Lorne Carr | 69
121 | Dit Clapper | 69
122 | Paul Thompson | 68.9
123 | Brian Leetch | 68.8
124 | Alex Kovalev | 68.7
125 | Phil Watson | 68.7
126 | Bun Cook | 68.7
127 | Rod Brind'Amour | 68.6
128 | Ray Whitney | 68.5
129 | Johnny Gottselig | 68.5
130 | Denis Potvin | 68.4
131 | Steve Larmer | 68.2
132 | Lanny McDonald | 68
133 | Lynn Patrick | 67.9
134 | Pavol Demitra | 67.7
135 | Phil Goyette | 67.6
136 | Alex Tanguay | 67.6
137 | Woody Dumart | 67.5
138 | Pete Mahovlich | 67.5
139 | Tony Amonte | 67.5
140 | Mike Ribeiro | 67.3
141 | Rick Middleton | 67.2
142 | Cecil Dillon | 67.1
143 | Dino Ciccarelli | 66.9
144 | Clint Smith | 66.8
145 | Don McKenney | 66.8
146 | Joe Mullen | 66.6
147 | Peter Bondra | 66.6
148 | Milan Hejduk | 66.5
149 | Bobby Smith | 66.3
150 | Joe Nieuwendyk | 66.3
151 | Dave Andreychuk | 66.2
152 | Bill Barber | 66.2
153 | Al MacInnis | 66.1
154 | Mike Gartner | 66
155 | Herbie Lewis | 66
156 | Glenn Anderson | 65.7
157 | Nicklas Lidstrom | 65.6
158 | Bill Thoms | 64.8
159 | Brian Propp | 64.7
160 | Rick Nash | 64.6
161 | Pit Martin | 64.6
162 | Anze Kopitar | 64.1
163 | Garry Unger | 64.1
164 | Dave Taylor | 64
165 | Marc Savard | 64
166 | Dean Prentice | 63.9
167 | Ebbie Goodfellow | 63.7
168 | Dennis Maruk | 63.5
169 | Marian Gaborik | 63.5
170 | Thomas Vanek | 63.3
171 | Rick MacLeish | 63.3
172 | Rick Martin | 63.2
173 | Scott Gomez | 63.2
174 | Corey Perry | 63.1
175 | Shane Doan | 62.5
176 | Olli Jokinen | 62.4
177 | Phil Housley | 62.4
178 | Eddie Wiseman | 62.3
179 | Tom Lysiak | 62.1
180 | Owen Nolan | 62.1
181 | Tod Sloan | 62
182 | Miroslav Satan | 62
183 | Steve Shutt | 61.6
184 | Larry Aurie | 61.5
185 | Nicklas Backstrom | 61.4
186 | Butch Goring | 61.4
187 | Alex Zhamnov | 61.4
188 | Brian Bellows | 61.3
189 | Herb Cain | 61.3
190 | Kenny Wharram | 61.1
191 | Wayne Cashman | 60.8
192 | Gordie Drillon | 60.7
193 | Steve Sullivan | 60.7
194 | Dennis Hull | 60.4
195 | Patrick Kane | 60.4
196 | Craig Janney | 60.3
197 | Pat Verbeek | 60.3
198 | Rene Robert | 60
199 | Jason Pominville | 59.9
200 | Vic Hadfield | 59.9
201 | Todd Bertuzzi | 59.9
202 | Murray Oliver | 59.9
203 | Pierre Larouche | 59.8
204 | Petr Sykora | 59.7
205 | Robert Lang | 59.7
206 | Jimmy Ward | 59.7
207 | Jean Pronovost | 59.6
208 | Bob Nevin | 59.4
209 | Ryan Smyth | 59.3
210 | Claude Provost | 59.3
211 | Buddy O'Connor | 59.2
212 | Martin Straka | 59.1
213 | Camille Henry | 58.9
214 | Red Berenson | 58.8
215 | Kirk Muller | 58.8
216 | Ivan Boldirev | 58.8
217 | Johnny Gagnon | 58.7
218 | Zach Parise | 58.7
219 | Brad Park | 58.7
220 | Jason Arnott | 58.6
221 | Petr Nedved | 58.6
222 | Vyacheslav Kozlov | 58.6
223 | Fred Stanfield | 58.5
224 | Cliff Ronning | 58.5
225 | Cory Stillman | 58.5
226 | Andrew Brunette | 58.3
227 | Neal Broten | 58.2
228 | Sergei Gonchar | 58.2
229 | Larry Murphy | 58.1
230 | Peter McNab | 58.1
231 | Wilf Paiement | 57.9
232 | Bill Guerin | 57.8
233 | Gary Roberts | 57.7
234 | George Armstrong | 57.6
235 | Jozef Stumpel | 57.6
236 | Walt Tkaczuk | 57.5
237 | Eddie Shore | 57.5
238 | Steve Thomas | 57.4
239 | Patrice Bergeron | 57.3
240 | Chris Drury | 57.2
241 | Kevin Stevens | 57.2
242 | Rick Tocchet | 57
243 | Phil Kessel | 56.9
244 | Tomas Sandstrom | 56.7
245 | Mike Ridley | 56.7
246 | Saku Koivu | 56.7
247 | Andrew Cassels | 56.5
248 | Syl Apps | 56.4
249 | Michael Nylander | 56.3
250 | Bob Pulford | 56.2

I'm using the "Orr rule" as per Sturm's revised post. The benchmark for 2015 was 86 points. The results are unweighted (ie each season is worth the same).

A few players made significant jumps:

- Alexander Ovechkin. Moves from 82.8 to 92.2 - into 14th place all-time.

- Sidney Crosby. Call this year a disappointment if you want, but it was effectively a full season of Art Ross calibre offense. He jumps from 78.6 to 88.3.

- Evgeni Malkin climbs into the top 100, jumping from 63.8 to 72.0.

Other players in the top 100 that moved up include Thornton (by a fraction of a percent), Iginla (+0.2), H. Sedin (+3.7), Hossa (+0.9), Datsyuk (+2.7), D. Sedin (+3.4), Stall (+2.7), Zetterberg (+2.7) and Spezza (+5.2).

 

[Hockey Outsider]

All-Time Results Through 2015 - best seven years

Rank | Player | Result
1 | Wayne Gretzky | 155.6
2 | Phil Esposito | 130.4
3 | Gordie Howe | 125.5
4 | Mario Lemieux | 119.8
5 | Bobby Orr | 114.8
6 | Jaromir Jagr | 114.2
7 | Bobby Hull | 108.3
8 | Stan Mikita | 107.8
9 | Jean Beliveau | 105.7
10 | Guy Lafleur | 104.5
11 | Ted Lindsay | 104.4
12 | Marcel Dionne | 103.3
13 | Maurice Richard | 102.4
14 | Howie Morenz | 102.2
15 | Sidney Crosby | 101.5
16 | Andy Bathgate | 101.1
17 | Alex Ovechkin | 98.4
18 | Joe Sakic | 97.7
19 | Bill Cowley | 97
20 | Charlie Conacher | 96.2
21 | Bill Cook | 96
22 | Frank Boucher | 95.1
23 | Mike Bossy | 94.8
24 | Joe Thornton | 94.1
25 | Bryan Trottier | 93.7
26 | Steve Yzerman | 93.2
27 | Teemu Selanne | 92.7
28 | Martin St. Louis | 92.4
29 | Syl Apps Sr | 92.4
30 | Sweeney Schriner | 91.3
31 | Bobby Clarke | 90.4
32 | Max Bentley | 90.4
33 | Peter Forsberg | 90.3
34 | Nels Stewart | 90.3
35 | Adam Oates | 90
36 | Bernie Geoffrion | 89.9
37 | Mark Messier | 89.6
38 | Jean Ratelle | 89.6
39 | Marty Barry | 89.6
40 | Norm Ullman | 89.5
41 | Busher Jackson | 89.5
42 | Evgeni Malkin | 89.2
43 | John Bucyk | 88.7
44 | Peter Stastny | 88.4
45 | Mark Recchi | 88.4
46 | Jari Kurri | 88.1
47 | Paul Coffey | 88.1
48 | Brett Hull | 88
49 | Frank Mahovlich | 87.8
50 | Ron Francis | 87.4
51 | Sid Abel | 87.3
52 | Milt Schmidt | 86.9
53 | Jarome Iginla | 86.7
54 | Gordie Drillon | 86.7
55 | Doug Bentley | 86.6
56 | Toe Blake | 86.3
57 | Elmer Lach | 86.1
58 | Dale Hawerchuk | 86
59 | Gilbert Perreault | 85.7
60 | Darryl Sittler | 85.7
61 | Denis Savard | 85.5
62 | Dickie Moore | 85.4
63 | Roy Conacher | 85.2
64 | Henri Richard | 85.2
65 | Pavel Bure | 85.2
66 | Eric Lindros | 85
67 | Paul Kariya | 84.9
68 | Alex Delvecchio | 84.8
69 | Henrik Sedin | 84.5
70 | Luc Robitaille | 84.2
71 | Ilya Kovalchuk | 83.9
72 | Syd Howe | 83.9
73 | Ken Hodge | 83.2
74 | Rod Gilbert | 83.1
75 | Markus Naslund | 82.9
76 | Aurel Joliat | 82.6
77 | Paul Thompson | 82.6
78 | Pavel Datsyuk | 82.5
79 | Marian Hossa | 82.4
80 | Daniel Alfredsson | 82.3
81 | Pierre Turgeon | 82.1
82 | Mats Sundin | 82.1
83 | Doug Gilmour | 82
84 | Theoren Fleury | 82
85 | John LeClair | 81.7
86 | Mike Modano | 81.5
87 | Nicklas Backstrom | 81.3
88 | Jeremy Roenick | 81.2
89 | Ryan Getzlaf | 81.1
90 | Dany Heatley | 81
91 | Bryan Hextall | 80.9
92 | Sergei Fedorov | 80.8
93 | Bernie Nicholls | 80.6
94 | Ziggy Palffy | 80.1
95 | Daniel Sedin | 79.9
96 | Lynn Patrick | 79.4
97 | Michel Goulet | 79.3
98 | Keith Tkachuk | 79
99 | Brendan Shanahan | 79
100 | Cooney Weiland | 78.9
101 | Patrik Elias | 78.9
102 | Eric Staal | 78.8
103 | Ted Kennedy | 78.8
104 | Steven Stamkos | 78.7
105 | Pat LaFontaine | 78.5
106 | Jason Spezza | 78.4
107 | Doug Weight | 78.3
108 | Cecil Dillon | 78.1
109 | Henrik Zetterberg | 78
110 | Brad Richards | 78
111 | Hooley Smith | 78
112 | Jacques Lemaire | 77.9
113 | Alexander Mogilny | 77.7
114 | Bernie Federko | 77.6
115 | Patrick Kane | 77.2
116 | Alexei Yashin | 77.1
117 | Yvan Cournoyer | 77.1
118 | Vincent Lecavalier | 76.9
119 | Phil Watson | 76.7
120 | Bun Cook | 76.3
121 | Bert Olmstead | 75.9
122 | Bobby Rousseau | 75.5
123 | Raymond Bourque | 75.5
124 | Red Kelly | 75.4
125 | Anze Kopitar | 75.4
126 | Alex Kovalev | 75.3
127 | Bill Mosienko | 75.2
128 | Phil Goyette | 75.2
129 | Clint Smith | 75.1
130 | Herbie Lewis | 75
131 | Denis Potvin | 74.9
132 | Johnny Gottselig | 74.8
133 | Joe Primeau | 74.8
134 | Bobby Bauer | 74.6
135 | Pavol Demitra | 74.5
136 | Lorne Carr | 74.4
137 | Dave Keon | 74.3
138 | Lanny McDonald | 74.1
139 | Rick Middleton | 74.1
140 | Mike Ribeiro | 74.1
141 | Vincent Damphousse | 74
142 | Corey Perry | 73.9
143 | Pete Mahovlich | 73.8
144 | Dit Clapper | 73.7
145 | Brian Leetch | 73.7
146 | Alex Tanguay | 73.4
147 | Marc Savard | 73.3
148 | Don McKenney | 73.3
149 | Rick MacLeish | 73.3
150 | Patrick Marleau | 73.2
151 | Milan Hejduk | 73.1
152 | Rick Martin | 73.1
153 | Tony Amonte | 73
154 | Ray Whitney | 72.9
155 | Bill Thoms | 72.7
156 | Woody Dumart | 72.7
157 | Phil Kessel | 72.6
158 | Claude Giroux | 72.6
159 | Rod Brind'Amour | 72.6
160 | Neil Colville | 72.1
161 | Glenn Anderson | 72
162 | Peter Bondra | 72
163 | Steve Larmer | 71.9
164 | Eddie Wiseman | 71.8
165 | Joe Mullen | 71.6
166 | Larry Aurie | 71.6
167 | Jonathan Toews | 71.4
168 | Kent Nilsson | 71.3
169 | Dennis Maruk | 71.3
170 | Rene Robert | 71
171 | Zach Parise | 70.9
172 | Tod Sloan | 70.9
173 | Steve Shutt | 70.6
174 | Pit Martin | 70.6
175 | Al MacInnis | 70.5
176 | Johnny Gagnon | 70.4
177 | Joe Nieuwendyk | 70.3
178 | Dave Taylor | 70.2
179 | Garry Unger | 70.1
180 | Bill Barber | 70
181 | Marian Gaborik | 69.9
182 | Dino Ciccarelli | 69.8
183 | Kenny Wharram | 69.8
184 | Gaye Stewart | 69.8
185 | Ebbie Goodfellow | 69.6
186 | Dave Andreychuk | 69.5
187 | Nicklas Lidstrom | 69.4
188 | Rick Nash | 69.2
189 | Syl Apps | 69.1
190 | Bobby Smith | 68.9
191 | Mike Gartner | 68.8
192 | Thomas Vanek | 68.7
193 | Baldy Northcott | 68.6
194 | Todd Bertuzzi | 68.5
195 | Jimmy Ward | 68.5
196 | Miroslav Satan | 68.4
197 | Dennis Hull | 68.3
198 | Sid Smith | 68.2
199 | Olli Jokinen | 68.2
200 | Wayne Cashman | 67.8
201 | Fred Stanfield | 67.8
202 | Robert Lang | 67.8
203 | Butch Goring | 67.8
204 | Paul Ronty | 67.8
205 | Walt Tkaczuk | 67.8
206 | Scott Gomez | 67.6
207 | Steve Sullivan | 67.6
208 | Red Berenson | 67.5
209 | Ed Litzenberger | 67.5
210 | Martin Straka | 67.4
211 | Craig Janney | 67.4
212 | Brian Propp | 67.2
213 | Herb Cain | 67.1
214 | Camille Henry | 67
215 | Dean Prentice | 67
216 | Kevin Stevens | 67
217 | Pierre Larouche | 66.6
218 | Owen Nolan | 66.6
219 | Ace Bailey | 66.6
220 | Claude Provost | 66.4
221 | Peter McNab | 66.1
222 | Shane Doan | 66.1
223 | Jason Pominville | 66
224 | Jason Allison | 65.9
225 | Buddy O'Connor | 65.9
226 | Petr Nedved | 65.9
227 | Alex Zhamnov | 65.7
228 | Tom Lysiak | 65.7
229 | John Tavares | 65.6
230 | Glen Murray | 65.5
231 | Billy Taylor | 65.5
232 | Wilf Paiement | 65.3
233 | Jean Pronovost | 65.3
234 | Vic Hadfield | 65.2
235 | Joe Pavelski | 65.2
236 | Charlie Simmer | 65.2
237 | Art Chapman | 65.2
238 | Brian Bellows | 65.1
239 | Phil Housley | 65.1
240 | John McKenzie | 65
241 | Petr Sykora | 64.8
242 | Murray Oliver | 64.7
243 | Kirk Muller | 64.7
244 | Barry Pederson | 64.7
245 | Neal Broten | 64.5
246 | Cory Stillman | 64.4
247 | Vic Stasiuk | 64.3
248 | Johnny Peirson | 64.1
249 | Alexander Semin | 64.1
250 | Vaclav Prospal | 64.1

A few players made significant jumps:

- Crosby jumped from 97.3 to 101.5, good for 15th all-time. He ranks higher than Ovechkin on the seven year list, but not the ten year list.

- Ovechkin moved slightly from 97.2 to 98.4 - not a large jump, as his 2015 campaign edges out his fairly strong 2011 season.

- He didn`t have a great year, but it was enough to move Malkin moves into the top fifty.

The other gainers in the top hundrer were H. Sedin (+2.0), Datsyuk (+0.9), Backstrom (+6.5), Getzlaf (+3.2) and D. Sedin (+2.8).

Stamkos soared from 66.7 to 78.7, even if this season was a bit dissapointing. He`s just outside the top hundred (currently 104th) despite only playing seven seasons, one of which was an unremarkable rookie season, and the other of which was a write-off due to injuries. He`s set to soar on this list in the near future.

 

[steve141]

The Orr Rule works by setting an exception in years when a defenseman is equal to or greater than the benchmark. The defenseman's scoring is removed for the purposes of setting the benchmark, and the scoring of all forwards on his team above the benchmark is normalized to that of a roughly average #1 scorer (that is, the benchmark +5%).

The reason we need this rule is to smooth out a few seasons in the early 1970's when the post-expansion Bruins broke the system, and made the benchmark too high. Basically, the idea is that peak Bobby Orr was such an outlier in the history of hockey that the benchmarks set by his teams need an adjustment. The Orr Rule appears to set benchmarks which are much fairer to the rest of the league, though it will inflate the VsX scores of Bruins players in this period even more than they already were (and thus the VsX scores of guys like Hodge and Cashman will have to carry something of an asterisk).

The rule also catches the single season when Paul Coffey was the #2 scorer in the league, and sets the benchmark to Goulet at #3 rather than Coffey at #2 (where the previous benchmark had been). Coffey's scoring in this season is also a historical outlier, so the rule sort of kills to birds with one stone in this sense.

Let me start by saying that I have immense respect for the thought that's been put into this, and that I think the result is very informative. Thanks a lot!

With regards to the Orr rule though, I guess I don't understand why it matters that he's a defenceman? Would it have mattered anything to Hawerchuk if Kurri was the one who got 126 points in 83-84 and Coffey 135 in 84-85?

Isn't the real problem the "breakage" done by several of the teams which have multiple offensive stars on one team, not that some of them were defencemen? I'm thinking of cases like Detroit in 49-50, or Montreal in 44-45 or 54-55. Basically anytime you have an offensive freak that inflates the numbers of his teammates greatly (I'm looking at you Kevin Stevens!) it's hard to know what the best benchmark is.

Some general teammate-fudge seems more natural to me than a defenceman-specific one.


In the specific instance of the Bruins I am also still a bit confused about what exactly we are trying to achieve. I might be misunderstanding the methodology, but my interpretation is that the benchmark in 70-71 was lowered from 100 to 90 because of the Orr rule? This means that Hull with 96 points goes from being 4 points under the benchmark to 6 over, even though he finished 5th in scoring in his 8th best adjusted season. That seems too generous to me. I can understand that Orr and Esposito were outliers, but a peak Hull would have been able to beat Ken Hodge anyway. Or am I misunderstanding something?

 

[Czech Your Math]

I think a simpler, more objective system would be better.

Probably the simplest one would be Vs. 3rd place.

 

[Sturminator]

I think a simpler, more objective system would be better.

Probably the simplest one would be Vs. 3rd place.

I'm not quite sure what you mean by "objective" in this context, but any benchmark is going to be arbitrary.

Trust me, if there was a simpler system that I thought would yield results as reasonable, I would use it. Using the #3 scorer is, in fact, the first "fix" I introduced to the old Vs2 system when it became clear that there were multiple outliers in some seasons.

Sadly, Vs3 is not enough to generate reasonable benchmarks. If you just used Vs3, the war years, the Bruins years, a few seasons in the 50's (1957-58, for example, is just a mess), a few seasons in the late 70's (like 1979-80), and a couple of the Gretzky/Lemieux years (1988-89 is particularly bad) would still end up very distorted, with the benchmark almost certainly too high.

I would be thrilled if I could make the system less complex than it currently is, but I don't think straight Vs3 is a viable answer.

 

[jigglysquishy]

A look at Crosby and Ovechkin's VsX. They're the two biggest stars of the past decade and their VsX scores finally dropped non-elite seasons (i.e. this season was their 7th season with a really high VsX score).

Year | Crosby | Ovechkin
2005-06 | 96.2 | 100.0
2006-07 | 105.3 | 80.7
2007-08 | 67.9 | 105.6
2008-09 | 93.6 | 100.0
2009-10 | 100.0 | 100.0
2010-11 | 66.7 | 85.9
2011-12 | 38.1 | 67.0
2012-13 | 98.2 | 98.2
2013-14 | 119.5 | 90.8
2014-15 | 97.7 | 94.2
Unweighted Average | 101.5 | 98.4

If the benchmark stays around the 85 point mark next season it would require Crosby to have 80+ points and Ovi to have 77+ points to change their scores. Both are likely.

Also goes to show just how dominant Gretzky was. He averages 155.3 over his best seven years. A 155.3 in 2014-15 would be 133 points. His best VsX score (1.69 in 86-87) would be 145 points today.

 

[TheDevilMadeMe]

I'm not quite sure what you mean by "objective" in this context, but any benchmark is going to be arbitrary.

Trust me, if there was a simpler system that I thought would yield results as reasonable, I would use it. Using the #3 scorer is, in fact, the first "fix" I introduced to the old Vs2 system when it became clear that there were multiple outliers in some seasons.

Sadly, Vs3 is not enough to generate reasonable benchmarks. If you just used Vs3, the war years, the Bruins years, a few seasons in the 50's (1957-58, for example, is just a mess), a few seasons in the late 70's (like 1979-80), and a couple of the Gretzky/Lemieux years (1988-89 is particularly bad) would still end up very distorted, with the benchmark almost certainly too high.

I would be thrilled if I could make the system less complex than it currently is, but I don't think straight Vs3 is a viable answer.

I think Vs5 (using the 5th leading scorer as a benchmark) works for post-expansion players at least but during weaker periods of talent (1940s/early 50s), it's going to give you wonky results.

I do fear that VsX is becoming too complicated in the quest for superior results.

 

[BraveCanadian]

I think Vs5 (using the 5th leading scorer as a benchmark) works for post-expansion players at least but during weaker periods of talent (1940s/early 50s), it's going to give you wonky results.

I do fear that VsX is becoming too complicated in the quest for superior results.

Maybe an average of 2-5 to smooth out some of the stupid seasons (1989) by including a few more players would work

 

[Czech Your Math]

I'm not quite sure what you mean by "objective" in this context, but any benchmark is going to be arbitrary.

Trust me, if there was a simpler system that I thought would yield results as reasonable, I would use it. Using the #3 scorer is, in fact, the first "fix" I introduced to the old Vs2 system when it became clear that there were multiple outliers in some seasons.

Sadly, Vs3 is not enough to generate reasonable benchmarks. If you just used Vs3, the war years, the Bruins years, a few seasons in the 50's (1957-58, for example, is just a mess), a few seasons in the late 70's (like 1979-80), and a couple of the Gretzky/Lemieux years (1988-89 is particularly bad) would still end up very distorted, with the benchmark almost certainly too high.

I would be thrilled if I could make the system less complex than it currently is, but I don't think straight Vs3 is a viable answer.

Technically, I guess it's objective... but then so is the federal tax code, supposedly.

I understand compromising simplicity for accuracy, and I know it's not an easy line to walk. I think some years/eras are always going to be "outliers" and nearly impossible to objectively determine a fair benchmark, regardless of method.

Maybe an average of 2-5 to smooth out some of the stupid seasons (1989) by including a few more players would work

An average of the 2nd-5th Canadians seemed probably the best/fairest method when I looked at various methods some months ago.

 

[BraveCanadian]

Technically, I guess it's objective... but then so is the federal tax code, supposedly.

I understand compromising simplicity for accuracy, and I know it's not an easy line to walk. I think some years/eras are always going to be "outliers" and nearly impossible to objectively determine a fair benchmark, regardless of method.

I agree. For me the simplicity or at least the consistency is very important. If you end up with a "system" that has a bunch of exceptions and asterisks on it then it isn't a "system" any longer.

An average of the 2nd-5th Canadians seemed probably the best/fairest method when I looked at various methods some months ago.

Intuitively it makes sense to me that something like this would be the best compromise of accuracy and simplicity.

 

[Sturminator]

Technically, I guess it's objective... but then so is the federal tax code, supposedly.

I understand compromising simplicity for accuracy, and I know it's not an easy line to walk. I think some years/eras are always going to be "outliers" and nearly impossible to objectively determine a fair benchmark, regardless of method.

The original benchmarks are still in the OP, and the War Years Fudge and Orr Rule discussions are provided as links-to-discussions for exactly this reason. It is not necessary to use the solutions offered for these two periods in order to use the system. The War Years Fudge, especially, is extremely case-specific and it very much deserves the name "fudge". I get that, but if what you're saying is that you like the "old VsX" system with its War/Bruins years asterisks better, ignoring the bolded numbers in the OP is not exactly difficult.

An average of the 2nd-5th Canadians seemed probably the best/fairest method when I looked at various methods some months ago.

We're probably at the high-water mark for the complexity and accuracy of the VsX system. The next big improvement that we could make to a percentages-based system would be on the simplicity side. If you have any better suggestions than the above (which suffers from a host of problems in various periods) for a simpler system, I'm all ears.

 

[Czech Your Math]

We're probably at the high-water mark for the complexity and accuracy of the VsX system. The next big improvement that we could make to a percentages-based system would be on the simplicity side. If you have any better suggestions than the above (which suffers from a host of problems in various periods) for a simpler system, I'm all ears.

I appreciate the work that you put into the system. I know it's easier to criticize or nitpick than come up with the original system or study.

I don't think the problems inherent in various eras are going to be easily overcome by any system, even with exceptions, modifications, asterisks, etc.

Since you asked, I believe the following systems are worth some study:

Vs3 (third place) This allows the removal of a lot of outliers: Howe & Richard... Howe & linemate... Beliveau & teammate... Howe & Hull/Mikita... Hull & Mikita... Espo & Orr... Lafleur & Dionne... Gretzky & Dionne... Gretzky & 80s superstar... Gretzky & linemate... Gretzky & Lemieux... Lemieux & linemate... Lemieux & Jagr... Jagr & career year by DPE superstar... whichever 2 of Crosby/Ovechkin/Malkin are healthy

VsN (avg. of top N scorers, where N = number of teams) This allows the number of scorers to remain constant in proportion to the opportunity (i.e., TOI, first line minutes, power play time, etc.).

VsN Canadians (same as above, but only Canadians used as reference)

Vs2-5C or Vs3-5C (avg. of the 2nd-5th or 3rd-5th Canadian scorers) The theory here is that it allows for the removal of an outlier or two, uses a more consistent talent pool (Canadians only), and also uses an average of 3-4 players for smoothing purposes.

 

[Hockey Outsider]

I've designed my spreadsheets in such a way that it's trivial to update the benchmark each year. I can do a comparison between the current system and the new system. If someone does the work to calculate the benchmark each year I can provide the results & changes compared to the current system pretty quickly.

 

[Hockey Outsider]

For curiosity's sake, I compared the current system (with Sturm's yearly benchmarks) to the simple "Vs3" system. That is, the third place scorer each year is used as the benchmark, with no exceptions or modifications.

I'm too busy to post all the tables but hopefully you can take my word on the results.

Of the 88 seasons we're analyzing (1927-2015), the benchmark changed by 5% or less in 62 of those campagins (70%). The benchmarked changed by 10% or less in 81 of those seasons.

The "Vx3" benchmark is significantly higher in 1979 (129 vs 116), 1989 (155 vs 139), 1974 (105 vs 91) and 1971 (116 vs 90). Two of those years are from the early seventies (massive imbalance in league parity + Orr's Bruins) and the other two years have a clear separation among the league's best players (in 1979 each of Trottier, Dionne, Lafleur and Bossy had very strong seasons, and I don't think I need to tell anyone reading this what Gretkzy, Lemieux and Yzerman did in 1989).

The "Vx3" benchmark is significantly lower in three years: 1945 (67 vs 78), 1946 (50 vs 60) and 1944 (74 vs 95). All, obviously, are seasons heavily impacted by the WWII talent depletion.

Under the Vx3 system, there are 180 players with a seven year score of 70 or more (vs 175 under Sturm's system). 138 players' scores change by no more than three points (77%). 165 players' scores change by no more than five points (92%). At a first glance, this might support the Vs3 system - most players don't change significantly, and the system is obviously simpler and more intuitive.

The players whose scores increased the most are Elmer Lach (+9.8), Bill Mosienko (+7.7), Clinth Smith (+6.5), Doug Bentley (+6.5), Bill Cowley (+5.6) and Toe Blake (+5.1).

The players whose scores decreased the most are Bryan Hextall (-9.7), Ken Hodge (-8.1), John Bucyk (-8.1), Phil Esposito (-7.3), Bobby Orr (-6.8), Yvan Cournoyer (-6.7), Dave Keon (-6.4), Brett Hull (-6.0) and Norm Ullman (-5.5).

Based on the changes in their rankings, and the new relative rankings, the Vs3 method seems to systematically overrate WWII era players and underrate post-expansion era players (which is exactly what we'd expect based on the changes in benchmarks I described earlier). Bill Cowley is ranked ahead of Beliveau, Lafleur and Dionne. Elmer Lach is ranked ahead of Sakic, Yzerman and Selanne. Yvan Cournoyer is ranked on par with Olli Jokinen and Dino Ciccarelli.

Conclusion: although there's a cost of added complexity, the VsX system, using Sturm's metrics, is clearly superior to the Vs3 system, which is systematically biased for/against players from certain eras.

====

If I had to guess, the "VsN" system, where N is the number of teams, would probably give the best results. We'd be looking so deep into the scoring race that the impact of outliers is virtually eliminated. The issues are 1) lots of work to calculate the benchmark each year (I don't have time to try this) and 2) still going to have wonky results for WWII, which we always will unless we mark a reasonable attempt to quantify it, as Sturm has done.