ATD Summary 2017

seventieslord

Student Of The Game
Mar 16, 2006
36,080
7,132
Regina, SK
The PPGVsX score is to be used as a complement to the VsX score; the former should almost always be presented with the latter.VsX completely ignores per-game domination.By taking the best 7 years, it is highly likely that the score is a decent representation of a player's per-game productivity in his prime.I do not share your concerns (or not close to the same extent) about some players "cheating their way" to a high score because they had unsustainable pace that would have slowed down without injuries.This would have to happen in many years to truly give an unreasonable score.

If two players have a VsX score of 84, and player A has a PPGVsX score of 0.94 and player B has a PPGVsX score of 0.83, then you know player A is a more dominant player, but also more injury prone, than player B.It quickly reveals a picture, and this is the goal of the system.

VsX was never intended to be a perfect system, it was intended to be a quick system.Meaning, it's a quick way for our human brain to relativize the offense of scoring forwards.PPGVsX is the same; if presented side by side with the VsX score, it is intended to be a quick way to relativize the per-game dominance of scoring forwards.We could develop obscenely complicated systems, trying to take everything into account, but eventually it would be hard to use and most people wouldn't even understand the methodology of such systems, nevermind calculating the numbers themselves.

Although PPGVsX is a valiant effort to recognize per-game dominance, is there any reason to pursue it further when filling in the blanks with replacement level offense seems to be a nice all-in-one?
 

BenchBrawl

Registered User
Jul 26, 2010
30,844
13,628
Although PPGVsX is a valiant effort to recognize per-game dominance, is there any reason to pursue it further when filling in the blanks with replacement level offense seems to be a nice all-in-one?

If we use both scores side by side (VsX and PPGVsX), it is more informative.

But hey, it's not like we have much more to do to pursue it, the scores are all there (thanks to Hockey Outsider).

I like the other system too.The ranking is like a combination of the ranking of the two systems, so I agree it's a good all-in-one.
 
Last edited:

jarek

Registered User
Aug 15, 2009
10,004
238
The PPGVsX score is to be used as a complement to the VsX score; the former should almost always be presented with the latter.VsX completely ignores per-game domination.By taking the best 7 years, it is highly likely that the PPGVsX score is a decent representation of a player's per-game productivity in his prime.I do not share your concerns (or not close to the same extent) about some players "cheating their way" to a high score because they had unsustainable pace that would have slowed down without injuries.This would have to happen in many years to truly give an unreasonable score.

If two players have a VsX score of 84, and player A has a PPGVsX score of 0.94 and player B has a PPGVsX score of 0.83, then you know player A is a more dominant player, but also more injury prone, than player B.It quickly reveals a picture, and this is the goal of the system.

VsX was never intended to be a perfect system, it was intended to be a quick system.Meaning, it's a quick way for our human brain to relativize the offense of scoring forwards.PPGVsX is the same; if presented side by side with the VsX score, it is intended to be a quick way to relativize the per-game dominance of scoring forwards.We could develop obscenely complicated systems, trying to take everything into account, but eventually it would be hard to use and most people wouldn't even understand the methodology of such systems, nevermind calculating the numbers themselves.

Let me put it this way: if you want me to accept a player being more dominant on a per game level and using that to evaluate him, then I'm going to have to ask you exactly how many games he's going to play during an ATD season, and exactly how much offense will be injected into the lineup to compensate during those missed games.

You want to reward players for great per-game dominance. I can appreciate that. This cannot come without some counter balance however.
 

VanIslander

A 19-year ATDer on HfBoards
Sep 4, 2004
35,134
6,429
The numbers are more relevant when presented by era. For example, the early NHL:

18 Howie Morenz 0.99
19 Charlie Conacher 0.99
30 Busher Jackson 0.93
33 Bill Cook 0.92
36 Sweeney Schriner 0.92
38 Frank Boucher 0.91
43 Doug Bentley 0.91
49 Nels Stewart 0.89
54 Marty Barry 0.88
60 Roy Conacher 0.86
82 Paul Thompson 0.82
100 Herbie Lewis 0.79
105 Hooley Smith 0.79
106 Aurel Joliat 0.79
118 Joe Primeau 0.77
121 Cecil Dillon 0.77
124 Bryan Hextall 0.76
139 Cooney Weiland 0.75
145 Johnny Gottselig 0.74
163 Larry Aurie 0.73
164 Bun Cook 0.72
173 Eddie Wiseman 0.72
183 Bill Thoms 0.71
187 Dit Clapper 0.7
193 Jimmy Ward 0.7
216 Doc Romnes 0.69
218 Baldy Northcott 0.68

A lot of wingers there (Clapper was a winger back then); Boucher looks great among centers of his era.
 

Dreakmur

Registered User
Mar 25, 2008
18,552
6,730
Orillia, Ontario
And then at the end of the day, the trouble I have with this system is that there's no way of knowing exactly what would have happened.

On top of that, if we are going to actually use these numbers as legitimately as VsX, then injuries need to be taken much more seriously in the ATD. A player will need to be evaluated by (total games played / total number of games), and have their VsXPPG multiplied by that factor. Then, we would need to calculate an ACTUAL replacement VsXPPG, not just guess, adjust it by a factor of (100 - (total games played / total number of games)), and add it to the first player's VsXPPG. I suppose in the case of the ATD, you can use the VsX of the player replacing said player in the lineup. Although this adds even more wrinkles because then you would need to re-evaluate things like line chemistry.

Then you have to redo this for the playoffs.

And none of this accounts for the fact that we are only assuming that a player would have carried their per game numbers into those missed games. A player who was blazing hot and scored 10 points in the last 3 games before missing the rest of the season is not the same as a player who had scored 2 points in the last 10 games and then missed the rest of the season.

I just really don't see the point of all this when a player's injuries are already built into their legacy (their placement in the scoring race, specifically, as well as any potential all star and award voting). I don't really see what the problem is with just assuming a player will play 82 games and taking what they actually did at face value. It's so much simpler, and as far as I'm concerned, doesn't do said player any injustice at all. You don't even have to assume you're getting a "lesser" version of them if you just take what they did at face value.

Didn't you just bash Recchi's vs. X score because he benefited because of his health?
 

jarek

Registered User
Aug 15, 2009
10,004
238
Then I have no idea why you think the vs.x system favours him...

You don't think a system that cherry picks a player's best offensive years from his whole career, and those best years happen to be spread out quite a bit, isn't less impressive than a guy who has the exact same offensive resume but with those 7 best years in being consecutive?

I'm not saying that player B is substantially better than Recchi in this example, but I do think it's more impressive.
 

seventieslord

Student Of The Game
Mar 16, 2006
36,080
7,132
Regina, SK
You don't think a system that cherry picks a player's best offensive years from his whole career, and those best years happen to be spread out quite a bit, isn't less impressive than a guy who has the exact same offensive resume but with those 7 best years in being consecutive?

I'm not saying that player B is substantially better than Recchi in this example, but I do think it's more impressive.

I don't see how it matters.
 

BenchBrawl

Registered User
Jul 26, 2010
30,844
13,628
I have to agree with jarek on that one.It's not very important, but consecutiveness of prime years still have some weight (in my personal evaluation anyway).

As to why I give any weight to it at all, consecutiveness of prime years means the player is more consistent from one season to the next (when it matters the most; in his prime).This is especially important for franchise players and superstars, because GMs will usually build the team around them.When building, you hope that most of your core will peak at the same time and create a perfect storm (or a window of opportunity).It is a rock in the GM's and team's shoe if a top player is inconsistent from one season to the next.Another reason is that consecutiveness of prime years increases your star power.If you always have one great year followed by an average one and repeat this pattern often in your career, you consistently ruin the momentum of your star power (one step forward, one step backward, rinse, repeat).

Imagine if Lafleur's prime years were randomly distributed throughout his career.Maybe there is no dynasty.Maybe Montreal would win the same amount of cup, but maybe not.The point is in the late 70s you could count on Lafleur for being among the top offensive scorer in the NHL, both in the regular season and in the playoffs.The consecutiveness of Lafleur's prime is one (among many) reason why his star power got so high.
 
Last edited:

jarek

Registered User
Aug 15, 2009
10,004
238
I have to agree with jarek on that one.It's not very important, but consecutiveness of prime years still have some weight (in my personal evaluation anyway).

As to why I give any weight to it at all, consecutiveness of prime years means the player is more consistent from one season to the next (when it matters the most; in his prime).This is especially important for franchise players and superstars, because GMs will usually build the team around them.When building, you hope that most of your core will peak at the same time and create a perfect storm (or a window of opportunity).It is a rock in the GM's and team's shoe if a top player is inconsistent from one season to the next.Another reason is that consecutiveness of prime years increases your star power.If you always have one great year followed by an average one and repeat this pattern often in your career, you consistently ruin the momentum of your star power (one step forward, one step backward, rinse, repeat).

Imagine if Lafleur's prime years were randomly distributed throughout his career.Maybe there is no dynasty.Maybe Montreal would win the same amount of cup, but maybe not.The point is in the late 70s you could count on Lafleur for being among the top offensive scorer in the NHL, both in the regular season and in the playoffs.The consecutiveness of Lafleur's prime is one (among many) reason why his star power got so high.

Agree with everything you said here.

Recchi has the problem of you never know when he's going to have that big year. In an ATD season, this could translate to some games where he's really needed to step up but for whatever reason, he does not and the team loses. He's a great offensive player but not one I'd ever want to have to count on.
 

Dreakmur

Registered User
Mar 25, 2008
18,552
6,730
Orillia, Ontario
I dint see any advantage to a player having their best years clumped together.

Your argument about how he will score during an ATD season cuts both ways. If a guy who had his best seasons spread out is going to have his ATD scoring sores through the season, then guys who had their offends clumped together are going to score in bunches on the draft.
 

jarek

Registered User
Aug 15, 2009
10,004
238
I dint see any advantage to a player having their best years clumped together.

Your argument about how he will score during an ATD season cuts both ways. If a guy who had his best seasons spread out is going to have his ATD scoring sores through the season, then guys who had their offends clumped together are going to score in bunches on the draft.

Actually, I think they'll be more consistent throughout the year.
 

Hockey Outsider

Registered User
Jan 16, 2005
9,058
13,989
I agree there's some inherent value in focusing on a player's best consecutive seasons because a steady, consistent prime probably gives their team a better chance to win.

This chart compares players' best seven seasons, to best seven consecutive seasons using VsX. You'll see that this is brutally unforgiving to players who are injured during their prime.

Rank|Player|Consec|Best|Difference
1 | Wayne Gretzky | 152.7 | 155.6 | -2.9
2 | Phil Esposito | 130.4 | 130.4 | 0
3 | Gordie Howe | 122 | 125.5 | -3.5
4 | Bobby Orr | 114.8 | 114.8 | 0
5 | Jaromir Jagr | 110 | 114.2 | -4.1
6 | Stan Mikita | 107.4 | 107.8 | -0.5
7 | Guy Lafleur | 104.5 | 104.5 | 0
8 | Mario Lemieux | 103.7 | 119.8 | -16.1
9 | Jean Beliveau | 102.4 | 105.7 | -3.3
10 | Howie Morenz | 102.2 | 102.2 | 0
11 | Ted Lindsay | 101 | 104.4 | -3.4
12 | Bobby Hull | 100.4 | 108.3 | -7.9
13 | Marcel Dionne | 99.3 | 103.3 | -4.1
14 | Andy Bathgate | 99.1 | 101.1 | -1.9
15 | Maurice Richard | 95.4 | 102.4 | -7
16 | Bill Cook | 94.9 | 96 | -1.1
17 | Charlie Conacher | 93.6 | 96.2 | -2.6
18 | Mike Bossy | 93.3 | 94.8 | -1.4
19 | Joe Thornton | 92.8 | 95.6 | -2.8
20 | Alex Ovechkin | 92.5 | 98.4 | -5.9
21 | Steve Yzerman | 92.3 | 93.2 | -0.9
22 | Bill Cowley | 92.3 | 97 | -4.8
23 | Bryan Trottier | 91.9 | 93.7 | -1.8
24 | Bobby Clarke | 90.4 | 90.4 | 0
25 | Sweeney Schriner | 90.4 | 91.3 | -0.9
26 | Max Bentley | 90.4 | 90.4 | 0
27 | Frank Boucher | 90.3 | 95.1 | -4.8
28 | Joe Sakic | 89.2 | 97.7 | -8.5
29 | Teemu Selanne | 88.4 | 92.7 | -4.4
30 | Jari Kurri | 88.1 | 88.1 | 0
31 | Marty Barry | 88.1 | 89.6 | -1.4
32 | Sidney Crosby | 88 | 101.8 | -13.8
33 | Nels Stewart | 87.9 | 90.3 | -2.4
34 | Norm Ullman | 87.7 | 89.5 | -1.7
35 | Peter Stastny | 87.1 | 88.4 | -1.3
36 | Busher Jackson | 87 | 89.5 | -2.4
37 | Martin St. Louis | 86.9 | 92.4 | -5.5
38 | Gordie Drillon | 86.7 | 86.7 | 0
39 | Doug Bentley | 86.6 | 86.6 | 0
40 | Adam Oates | 86.6 | 90 | -3.4
41 | Syl Apps Sr | 86.5 | 92.4 | -5.8
42 | John Bucyk | 85.8 | 88.7 | -2.8
43 | Denis Savard | 85.5 | 85.5 | 0
44 | Jean Ratelle | 85.3 | 89.6 | -4.3
45 | Darryl Sittler | 84.6 | 85.7 | -1.1
46 | Patrick Kane | 84.5 | 84.5 | 0
47 | Dickie Moore | 83.6 | 85.4 | -1.8
48 | Brett Hull | 83.6 | 88 | -4.4
49 | Bernie Geoffrion | 83.6 | 89.9 | -6.3
50 | Peter Forsberg | 83.5 | 90.3 | -6.8
51 | Dale Hawerchuk | 83.4 | 86 | -2.6
52 | Steven Stamkos | 83 | 83 | 0
53 | Markus Naslund | 82.9 | 82.9 | 0
54 | Henrik Sedin | 82.5 | 84.5 | -2
55 | Eric Lindros | 82.4 | 85 | -2.6
56 | Toe Blake | 82.4 | 86.3 | -4
57 | Daniel Alfredsson | 82.3 | 82.3 | 0
58 | Ken Hodge | 82.1 | 83.2 | -1.1
59 | Jarome Iginla | 82 | 86.7 | -4.7
60 | Luc Robitaille | 81 | 84.2 | -3.2
61 | Ilya Kovalchuk | 80.9 | 83.9 | -3
62 | Paul Coffey | 80.9 | 88.1 | -7.2
63 | Ron Francis | 80.8 | 87.4 | -6.6
64 | Gilbert Perreault | 80.7 | 85.7 | -5
65 | Syd Howe | 80.6 | 83.9 | -3.3
66 | Evgeni Malkin | 80.5 | 89.2 | -8.7
67 | Paul Thompson | 80.5 | 82.6 | -2.1
68 | Frank Mahovlich | 80.4 | 87.8 | -7.4
69 | Mark Recchi | 80.2 | 88.4 | -8.2
70 | Mark Messier | 80.1 | 89.6 | -9.5
71 | Henri Richard | 80.1 | 85.2 | -5.1
72 | Bryan Hextall | 80.1 | 80.9 | -0.8
73 | Alex Delvecchio | 80 | 84.8 | -4.8
74 | Marian Hossa | 79.9 | 82.4 | -2.4
75 | Claude Giroux | 79.9 | 79.9 | 0
76 | Mats Sundin | 79.7 | 82.1 | -2.4
77 | Mike Modano | 79.7 | 81.5 | -1.8
78 | Milt Schmidt | 79.5 | 86.9 | -7.4
79 | Michel Goulet | 79.3 | 79.3 | 0
80 | Aurel Joliat | 79.2 | 82.6 | -3.5
81 | Bernie Nicholls | 78.7 | 80.6 | -1.9
82 | Nicklas Backstrom | 78.5 | 83.2 | -4.8
83 | Pierre Turgeon | 78.5 | 82.1 | -3.7
84 | Ryan Getzlaf | 78.4 | 82.1 | -3.7
85 | Rod Gilbert | 78.2 | 83.1 | -4.9
86 | Cecil Dillon | 78.1 | 78.1 | 0
87 | Lynn Patrick | 78.1 | 79.4 | -1.3
88 | John LeClair | 77.8 | 81.7 | -3.9
89 | Cooney Weiland | 77.5 | 78.9 | -1.4
90 | Theoren Fleury | 77.4 | 82 | -4.5
91 | Daniel Sedin | 77.2 | 79.9 | -2.8
92 | Elmer Lach | 77.2 | 86.1 | -8.9
93 | Pavel Datsyuk | 77.2 | 82.5 | -5.4
94 | Yvan Cournoyer | 77.1 | 77.1 | 0
95 | Anze Kopitar | 76.9 | 76.9 | 0
96 | John Tavares | 76.8 | 76.8 | 0
97 | Phil Watson | 76.7 | 76.7 | 0
98 | Ted Kennedy | 76.6 | 78.8 | -2.2
99 | Ziggy Palffy | 76.4 | 80.1 | -3.7
100 | Paul Kariya | 76.3 | 84.9 | -8.7
101 | Pat LaFontaine | 76.3 | 78.5 | -2.3
102 | Sid Abel | 76.2 | 87.3 | -11.1
103 | Bernie Federko | 76.1 | 77.6 | -1.4
104 | Bobby Rousseau | 75.5 | 75.5 | 0
105 | Hooley Smith | 75.5 | 78 | -2.4
106 | Vincent Lecavalier | 75.5 | 76.9 | -1.5
107 | Dany Heatley | 75.2 | 81 | -5.8
108 | Doug Gilmour | 75.1 | 82 | -6.9
109 | Joe Primeau | 74.8 | 74.8 | 0
110 | Bobby Bauer | 74.6 | 74.6 | 0
111 | Eric Staal | 74.5 | 78.8 | -4.3
112 | Corey Perry | 74.5 | 75 | -0.5
113 | Phil Kessel | 74.3 | 74.9 | -0.6
114 | Bun Cook | 74.3 | 76.3 | -2.1
115 | Sergei Fedorov | 74.3 | 80.8 | -6.5
116 | Bert Olmstead | 74.3 | 75.9 | -1.7
117 | Jacques Lemaire | 74.2 | 77.9 | -3.7
118 | Henrik Zetterberg | 73.3 | 78 | -4.7
119 | Herbie Lewis | 73.3 | 75 | -1.7
120 | Alexei Yashin | 73.2 | 77.1 | -3.9
121 | Rick Middleton | 73.2 | 74.1 | -1
122 | Pavol Demitra | 73 | 74.5 | -1.5
123 | Don McKenney | 73 | 73.3 | -0.3
124 | Jason Spezza | 73 | 79.1 | -6.1
125 | Roy Conacher | 72.9 | 85.2 | -12.3
126 | Pete Mahovlich | 72.9 | 73.8 | -0.9
127 | Jamie Benn | 72.7 | 72.7 | 0
128 | Brendan Shanahan | 72.6 | 79 | -6.4
129 | Keith Tkachuk | 72.6 | 79 | -6.5
130 | Milan Hejduk | 72 | 73.1 | -1.1
131 | Glenn Anderson | 72 | 72 | 0
132 | Rick Martin | 71.9 | 73.1 | -1.2
133 | Alexander Mogilny | 71.9 | 77.7 | -5.8
134 | Eddie Wiseman | 71.8 | 71.8 | 0
135 | Tony Amonte | 71.8 | 73 | -1.2
136 | Johnny Gottselig | 71.8 | 74.8 | -3
137 | Jonathan Toews | 71.7 | 72.3 | -0.6
138 | Dit Clapper | 71.7 | 73.7 | -2
139 | Brad Richards | 71.6 | 78 | -6.4
140 | Doug Weight | 71.6 | 78.3 | -6.8
141 | Jeremy Roenick | 71.4 | 81.2 | -9.8
142 | Larry Aurie | 71 | 71.6 | -0.5
143 | Rene Robert | 71 | 71 | 0
144 | Lanny McDonald | 70.9 | 74.1 | -3.2
145 | Lorne Carr | 70.9 | 74.4 | -3.5
146 | Woody Dumart | 70.8 | 72.7 | -1.9
147 | Patrik Elias | 70.8 | 78.9 | -8.1
148 | Red Kelly | 70.7 | 75.4 | -4.7
149 | Steve Shutt | 70.6 | 70.6 | 0
150 | Vincent Damphousse | 70.4 | 74 | -3.6
151 | Bill Mosienko | 70.2 | 75.2 | -5
152 | Alex Kovalev | 70.1 | 75.3 | -5.2
153 | Garry Unger | 70.1 | 70.1 | 0
154 | Joe Pavelski | 70.1 | 71.1 | -1
155 | Rick MacLeish | 70.1 | 73.3 | -3.2
156 | Phil Goyette | 70 | 75.2 | -5.2
157 | Denis Potvin | 69.8 | 74.9 | -5.1
158 | Alex Tanguay | 69.7 | 73.4 | -3.7
159 | Kenny Wharram | 69.4 | 69.8 | -0.4
160 | Steve Larmer | 69.1 | 71.9 | -2.8
161 | Syl Apps | 69.1 | 69.1 | 0
162 | Patrick Marleau | 68.8 | 73.2 | -4.4
163 | Dave Taylor | 68.7 | 70.2 | -1.5
164 | Bill Thoms | 68.7 | 72.7 | -4
165 | Baldy Northcott | 68.6 | 68.6 | 0
166 | Dave Keon | 68.5 | 74.3 | -5.7
167 | Joe Mullen | 68.5 | 71.6 | -3.1
168 | Miroslav Satan | 68.4 | 68.4 | 0
169 | Kent Nilsson | 68.4 | 71.3 | -3
170 | Raymond Bourque | 68.2 | 75.5 | -7.2
171 | Bill Barber | 68.2 | 70 | -1.8
172 | Gaye Stewart | 68.2 | 69.8 | -1.6
173 | Peter Bondra | 68.1 | 72 | -3.9
174 | Clint Smith | 67.8 | 75.1 | -7.3
175 | Mike Ribeiro | 67.8 | 74.1 | -6.3
176 | Fred Stanfield | 67.8 | 67.8 | -0.1
177 | Neil Colville | 67.8 | 72.1 | -4.4
178 | Paul Ronty | 67.8 | 67.8 | 0
179 | Marc Savard | 67.7 | 73.3 | -5.6
180 | Steve Sullivan | 67.6 | 67.6 | 0
181 | Pit Martin | 67.6 | 70.6 | -3
182 | Rod Brind'Amour | 67.4 | 72.6 | -5.2
183 | Dennis Maruk | 67.1 | 71.3 | -4.2
184 | Robert Lang | 67 | 67.8 | -0.8
185 | Red Berenson | 66.8 | 67.5 | -0.8
186 | Butch Goring | 66.8 | 67.8 | -1
187 | Ebbie Goodfellow | 66.7 | 69.6 | -2.8
188 | Ace Bailey | 66.6 | 66.6 | 0
189 | Pavel Bure | 66.5 | 85.2 | -18.7
190 | Brian Propp | 66.3 | 67.2 | -0.9
191 | Peter McNab | 66.1 | 66.1 | 0
192 | Al MacInnis | 65.8 | 70.5 | -4.6
193 | Ed Litzenberger | 65.7 | 67.5 | -1.7
194 | Tom Lysiak | 65.7 | 65.7 | 0
195 | Rick Nash | 65.6 | 69.2 | -3.6
196 | Scott Gomez | 65.5 | 67.6 | -2.1
197 | Billy Taylor | 65.5 | 65.5 | 0
198 | Olli Jokinen | 65.4 | 68.2 | -2.8
199 | Dennis Hull | 65.4 | 68.3 | -3
200 | Jason Allison | 65.2 | 65.9 | -0.6
201 | Brian Leetch | 65.1 | 73.7 | -8.6
202 | Blake Wheeler | 65 | 65.9 | -0.9
203 | John McKenzie | 65 | 65 | 0
204 | Thomas Vanek | 64.9 | 68.7 | -3.9
205 | Sid Smith | 64.8 | 68.2 | -3.4
206 | Dave Andreychuk | 64.7 | 69.5 | -4.8
207 | Kirk Muller | 64.7 | 64.7 | 0
208 | Craig Janney | 64.6 | 67.4 | -2.8
209 | Petr Nedved | 64.6 | 65.9 | -1.3
210 | Kevin Stevens | 64.6 | 67 | -2.4
211 | Mike Gartner | 64.6 | 68.8 | -4.2
212 | Vic Hadfield | 64.4 | 65.2 | -0.8
213 | Jakub Voracek | 64.4 | 64.4 | 0
214 | Nicklas Lidstrom | 64.3 | 69.4 | -5.1
215 | Vic Stasiuk | 64.3 | 64.3 | 0
216 | Todd Bertuzzi | 64.2 | 68.5 | -4.4
217 | Jason Pominville | 64.1 | 66 | -1.9
218 | Jean Pronovost | 64 | 65.3 | -1.2
219 | Bill Hay | 63.8 | 63.8 | 0
220 | Petr Sykora | 63.7 | 64.8 | -1.1
221 | Mats Naslund | 63.5 | 63.5 | 0
222 | Bobby Smith | 63.3 | 68.9 | -5.6
223 | Buddy O'Connor | 63.3 | 65.9 | -2.6
224 | Wilf Paiement | 63.2 | 65.3 | -2.1
225 | Ray Whitney | 63.2 | 72.9 | -9.6
226 | Tod Sloan | 63.2 | 70.9 | -7.7
227 | Bill Goldsworthy | 63.1 | 63.2 | -0.1
228 | Jim Pappin | 63.1 | 63.4 | -0.2
229 | Murray Oliver | 63 | 64.7 | -1.7
230 | Brad Park | 63 | 63.3 | -0.3
231 | Zach Parise | 63 | 70.9 | -7.9
232 | Bill Guerin | 63 | 63.9 | -0.9
233 | Dino Ciccarelli | 62.9 | 69.8 | -6.9
234 | Alexander Semin | 62.8 | 64.1 | -1.2
235 | Patrick Sharp | 62.8 | 66.2 | -3.3
236 | Shane Doan | 62.8 | 66.1 | -3.2
237 | Joe Nieuwendyk | 62.8 | 70.3 | -7.5
238 | Patrice Bergeron | 62.7 | 66.5 | -3.8
239 | Glen Murray | 62.5 | 65.5 | -3
240 | Danny Grant | 62.4 | 62.4 | 0
241 | Robert Reichel | 62.3 | 62.3 | 0
242 | Doc Romnes | 62.2 | 62.2 | 0
243 | Jimmy Ward | 62.2 | 68.5 | -6.3
244 | Wayne Cashman | 62.2 | 67.8 | -5.7
245 | Ryan Smyth | 62 | 62.1 | -0.1
246 | Johnny Peirson | 61.9 | 64.1 | -2.2
247 | Matt Duchene | 61.9 | 61.9 | 0
248 | Alex Zhamnov | 61.8 | 65.7 | -3.9
249 | Barry Pederson | 61.7 | 64.7 | -2.9
250 | Guy Chouinard | 61.7 | 61.7 | 0

Fifty of the top 250 players have the same score under both methods. That is, their best seven seasons (per VsX) happened to be consecutive. Not surprisingly this includes players like Orr and Lafleur. Another 26 players' scores changed by no more than 1.0 points (Yzerman, Mikita).

Not surprisingly, players who were most affected were often injured during their prime (Lemieux, Bure). Crosby has the 15th best VsX score through 2016, but if we only look at best consecutive seasons, he falls to 32nd - roughly on par with Kurri and Stastny.

Forsberg is another odd case. His 2003 campaign (Hart & Art Ross) gets excluded entirely. He played all of 278 games in the six seasons surrounding it, including just one with more than 60 games.

This also hurts players with late-career resurgences. For example, Recchi's 2000 season is excluded entirely. That drops him from 45th (Stastny/Kurri level) to 69th (Mahovlich/Delvecchio level).
 
Last edited:

tinyzombies

Registered User
Dec 24, 2002
16,827
2,339
Montreal, QC, Canada
I've been reading through this and it's fascinating. Is there a grand summary of the parameters or formulas you guys used? Is there a description of this new virtual forum/league where all players are leveled? Does it take place in 2017? Did you include the differences between the eras in the formulas for each season? An upgrade for the older guys for equipment/training? Do you adjust up for penalty minutes because of the style that used to be standard? Structural changes to the game? Etc. Or do you leave that to the draft and subjectivity?

I read the article one of you wrote about VSX. Hockeyreference has adjusted stats, and they adjusted the roster size and number of games, but I don't see where they would have accounted for a weaker schedule in the expanded league (weighted vs weaker divisions) compared to the original six (which was really the Original Three if you think about it).

Are there spreadsheets available to everyone, or are those proprietary? Obviously you have a long trail of proof. Formulas? Trying to get a handle on how to approach the next draft.
 
Last edited:

Johnny Engine

Moderator
Jul 29, 2009
4,972
2,352
I've been reading through this and it's fascinating. Is there a grand summary of the parameters or formulas you guys used?
Most of it is scattered around here and there, but the key to calculating VsX, for one, is found here:

Is there a description of this new virtual forum/league where all players are leveled? Does it take place in 2017? Did you include the differences between the eras in the formulas for each season? An upgrade for the older guys for equipment/training? Do you adjust up for penalty minutes because of the style that used to be standard? Structural changes to the game? Etc. Or do you leave that to the draft and subjectivity?
It's all imaginary. Myself, I imagine it looking like a Winter Classic, where everyone has access to decent equipment, but some people are wearing funny hats. Others might see it differently. When we get to the playoffs, certain posters will be tasked with recapping series based on the closeness of the votes, and which players were judged to be important. Anyone voting is encouraged to understand context when it comes to penalties - obviously, Sprague Cleghorn would be banned for life if he went around trying to swordfight people in 2017, but we understand that he accomplished what he accomplished in his own time, and the degree to which his lack of discipline hurt him, was what it was.
Some GMs occasionally declare that we need to start giving this or that more consideration in the voting - whether it's peak, playoff performance, whatever - but ultimately, that has the same effect as posting "people need to consider education/the military/whatever when they vote!" on your Facebook wall. The odd person might be convinced, but ultimately it's your prerogative to insert your own values into the conversation.

I read the article one of you wrote about VSX. Hockeyreference has adjusted stats, and they adjusted the roster size and number of games, but I don't see where they would have accounted for a weaker schedule in the expanded league (weighted vs weaker divisions) compared to the original six (which was really the Original Three if you think about it).
That should be baked into the formula. Offense spiked upward as the league got weaker on average during that time, and VsX scores reflect that. Norm Ullman's 70 point season in 1967 gives him a VsX score of 100 for that year, but Alex Delvecchio's 70 points the next year give him a score of just 80. It's fairly obvious why this happened, but the formula doesn't spell it out for you because numbers don't work that way. It's like how we know that Alex Ovechkin can score 50 goals because he's fast and has a great slap shot, but the number 50 doesn't tell you that. Almost every 50 goal scorer is fast and a big shooter, but then there's also Tim Kerr. It's the same with VsX - if you know the stories behind the numbers, you can see and quantify them, but they can't tell you the whole story.

Are there spreadsheets available to everyone, or are those proprietary? Obviously you have a long trail of proof. Formulas? Trying to get a handle on how to approach the next draft.
I don't personally keep a spreadsheet, though I really should to avoid redoing this work constantly. I'm sure others will send it to you if you ask.
 

Hockey Outsider

Registered User
Jan 16, 2005
9,058
13,989
Are there spreadsheets available to everyone, or are those proprietary? Obviously you have a long trail of proof. Formulas? Trying to get a handle on how to approach the next draft.

My spreadsheet is available to anyone who asks (send me a PM). That being said it might be better to hold off for a couple of weeks as I intend to update it once the 2017 regular season concludes.
 

King Forsberg

16 21 28 44 68 88 93
Jul 26, 2010
6,192
59
Biggest Steal(s) of the draft: Bobby Clarke #29
Biggest Reach(es) of the draft: Reijo Ruotsalainen
Smartest/best strategic pick in the draft: Keith Tkachuk
Biggest blunder selection of the draft: Graham Drinkwater
A Player finally getting respect in the draft: Vladimir Krutov, kinda
A player always taken too high, finally getting picked where he should in the draft: Dave Andreychuk
A player you've discovered in this draft: Konstantin Loktev
Most underrated player taken: Edgar Laprade
Most overrated player taken: Bob Gainey
Favorite scoring line of the draft: Leclair - Mikita - Bossy
Favorite checking line of the draft: Graves - Brind'Amour - Mullen
Best assembled line of the draft: Tkachuk - Gilmour - Lafleur
Worst assembled line of the draft: Mahavolich - Oates - Bondra
Favorite pairing of defensemen: Stevens - LaPointe
Most puzzling pairing of defensemen: N/A
Team in the other conference it'd be interesting to meet in the finals: Orilla Terriers
Team in the other conference you wouldn't want to meet in the finals: Montreal Canadiens
A funny/dramatic story (related to the ATD) you've learned about since the start of the draft:

---

Best selection: You cannot vote for players you own
1st round: Guy Lafleur
2nd round: Bobby Clarke, Patrick Roy
3rd round: Terry Sawchuk
4th round: Boris Mikhailov
5th round: Alexander Maltsev
6th round: Mark Recchi, Evgeni Malkin
7th round: Patrick Kane
8th round: Vladimir Krutov
9th round: Rod Gilbert
10th round: Billy Smith
11th round: Anze Kopitar
12nd round:Brian Propp
13th round: Tommy Ivan
14th round: Joe Mullen
15th round: Brent Burns
16th round: Dino Ciccarelli
17th round: Kimmo Timonen
18th round:Edgar Laprade
19th round: Dave Andreychuk
20th round:David Backes
21st round: Yuri Liapkin
22nd round: Bernie Nicholls
23th round: Jeff Carter
24th round: Brian Campbell
25th round: Kent Nilsson

'Worst' Selection
1st round:
2nd round: Jari Kurri
3rd round:
4th round:
5th round:
6th round:
7th round:
8th round:Bob Gainey
9th round:
10th round:
11th round:
12nd round:
13th round:
14th round:
15th round:
16th round:
17th round:
18th round:
19th round:
20th round:
21st round:
22nd round:
23th round:
24th round:
25th round:


Also, feel free to add any other comments or reflections about your experience/team this year.[/QUOTE]
 

Ad

Upcoming events

Ad

Ad

-->