Hockey Outsider
Registered User
- Jan 16, 2005
- 9,497
- 15,814
Year-end all-forward all-star teams
- Thread starter Hockey Outsider
- Start date
Hockey Outsider
Registered User
- Jan 16, 2005
- 9,497
- 15,814
Comparison of positional vs all-forward (first-team only)
From 1967-68 to 2020-21, 69 forwards have earned 159 spots on the NHL's official year-end all-star team (this is a span of 53 seasons, and 53 * 3 = 159). 69 forwards have earned 162 spots in the top three in Hart voting for forwards (there are a few more players in this category due to ties - I haven't attempted to break any ties that exist). In total, we have 98 players who appear in either or both categories.
Before someone says this analysis is a waste of time, consider this - of those 98 players, just 12 of them have the same all-star record under both frameworks (top three forward vs best at their position). This includes players like Guy Lafleur (6x first-team RW vs. 6x top-three in Hart voting among forwards), Jarome Iginla (3x in both categories) and Sergei Fedorov (1x in both categories).
Players most helped by looking at top three forwards
There are 15 players who have at least two extra years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. It shouldn't surprise anyone that almost all of these players are centers (the only exceptions are Fleury and Giroux, who split positions). This highlights how tough it can be to make the first all-star team as a centre. Gilmour and MacKinnon were both top three forwards three times, but were never named the best centre in the leauge.
Crosby, with eight years as a hypothetical "first-team forward" looks really good based on this metric. He's one of only three forwards (along with Gretzky and Lemieux) to have more than six such seasons. (On the other hand, McDavid gets screwed in 2018. He was clearly the best player in the NHL that season but, as they usually do, the Hart voters penalized him for missing the playoffs. So McDavid was the first-team centre, but only finished 5th among forwards in Hart voting, so he only gets a spot on the second all-forward team - which isn't reflected above).
Players most helped by looking at each forward position separately
There are 17 players who have at least two fewer years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. There are only two centres here (Sakic and Forsberg - I've argued that both of them have been underrated by the Hart trophy voting, picking up surprisingly few votes outside of the years they won). The other fifteen players are wingers.
Not to keep picking on Luc Robitaille, but he has the largest difference by far. Five times he was named the best LW in the league, but he never placed in the top three in Hart trophy voting. (Yzerman still doesn't look great by this metric - that's the downside of peaking at the same time as Gretzky and Lemieux - but at least he's up 2-0).
Goulet also takes a big hit, losing all three of his first-team all-star spots. Ovechkin also loses three spots, but he still has five years as a first-team forward, which is tied for the 8th best result here.
From 1967-68 to 2020-21, 69 forwards have earned 159 spots on the NHL's official year-end all-star team (this is a span of 53 seasons, and 53 * 3 = 159). 69 forwards have earned 162 spots in the top three in Hart voting for forwards (there are a few more players in this category due to ties - I haven't attempted to break any ties that exist). In total, we have 98 players who appear in either or both categories.
Before someone says this analysis is a waste of time, consider this - of those 98 players, just 12 of them have the same all-star record under both frameworks (top three forward vs best at their position). This includes players like Guy Lafleur (6x first-team RW vs. 6x top-three in Hart voting among forwards), Jarome Iginla (3x in both categories) and Sergei Fedorov (1x in both categories).
Players most helped by looking at top three forwards
| Player | First position | First forward | Difference |
| Wayne Gretzky | 8 | 13 | 5 |
| Mario Lemieux | 5 | 9 | 4 |
| Sidney Crosby | 4 | 8 | 4 |
| Bobby Clarke | 2 | 6 | 4 |
| Marcel Dionne | 2 | 5 | 3 |
| Doug Gilmour | 0 | 3 | 3 |
| Nathan MacKinnon | 0 | 3 | 3 |
| Bryan Trottier | 2 | 4 | 2 |
| Joe Thornton | 1 | 3 | 2 |
| Stan Mikita | 1 | 3 | 2 |
| Claude Giroux | 0 | 2 | 2 |
| Jean Beliveau | 0 | 2 | 2 |
| Jean Ratelle | 0 | 2 | 2 |
| Red Berenson | 0 | 2 | 2 |
| Theo Fleury | 0 | 2 | 2 |
There are 15 players who have at least two extra years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. It shouldn't surprise anyone that almost all of these players are centers (the only exceptions are Fleury and Giroux, who split positions). This highlights how tough it can be to make the first all-star team as a centre. Gilmour and MacKinnon were both top three forwards three times, but were never named the best centre in the leauge.
Crosby, with eight years as a hypothetical "first-team forward" looks really good based on this metric. He's one of only three forwards (along with Gretzky and Lemieux) to have more than six such seasons. (On the other hand, McDavid gets screwed in 2018. He was clearly the best player in the NHL that season but, as they usually do, the Hart voters penalized him for missing the playoffs. So McDavid was the first-team centre, but only finished 5th among forwards in Hart voting, so he only gets a spot on the second all-forward team - which isn't reflected above).
Players most helped by looking at each forward position separately
| Player | First position | First forward | Difference |
| Luc Robitaille | 5 | 0 | -5 |
| Alex Ovechkin | 8 | 5 | -3 |
| Michel Goulet | 3 | 0 | -3 |
| Mike Bossy | 5 | 3 | -2 |
| Bobby Hull | 4 | 2 | -2 |
| Joe Sakic | 3 | 1 | -2 |
| Paul Kariya | 3 | 1 | -2 |
| Peter Forsberg | 3 | 1 | -2 |
| Gordie Howe | 3 | 1 | -2 |
| Brendan Shanahan | 2 | 0 | -2 |
| Clark Gillies | 2 | 0 | -2 |
| John LeClair | 2 | 0 | -2 |
| Brad Marchand | 2 | 0 | -2 |
| Charlie Simmer | 2 | 0 | -2 |
| Jari Kurri | 2 | 0 | -2 |
| Ken Hodge | 2 | 0 | -2 |
| Richard Martin | 2 | 0 | -2 |
There are 17 players who have at least two fewer years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. There are only two centres here (Sakic and Forsberg - I've argued that both of them have been underrated by the Hart trophy voting, picking up surprisingly few votes outside of the years they won). The other fifteen players are wingers.
Not to keep picking on Luc Robitaille, but he has the largest difference by far. Five times he was named the best LW in the league, but he never placed in the top three in Hart trophy voting. (Yzerman still doesn't look great by this metric - that's the downside of peaking at the same time as Gretzky and Lemieux - but at least he's up 2-0).
Goulet also takes a big hit, losing all three of his first-team all-star spots. Ovechkin also loses three spots, but he still has five years as a first-team forward, which is tied for the 8th best result here.
Last edited:
Hockey Outsider
Registered User
- Jan 16, 2005
- 9,497
- 15,814
Comparison of positional vs all-forward (first & second team)
From 1967-68 to 2020-21, 129 forwards have earned 318 spots on the NHL's official year-end all-star team (this is a span of 53 seasons, and 53 * 6 = 318). 131 forwards have earned 324 spots in the top three in Hart voting for forwards (there are a few more players in this category due to ties - I haven't attempted to break any ties that exist). In total, we have 170 players who appear in either or both categories.
Just 11 players, out of 170 in total, have the same voting outcome under both frameworks. An example is Pavel Bure - he was a first-team all-star at RW once, and a second-team all-star at RW twice. He also had one year where he finished in the top three in forwards among Hart voting, and two years where he finished between 4th and 6th (which I'm treating as the equivalent of the second all-forward team).
Players most helped by looking at top three forwards
There are 27 forwards who have at least two more years as an all-forward (compared to the official, positional selections). Nearly all of them are centres (the only exceptions are Kerr and Giroux). There are eight players with three more years as an all-forward, and all of them are centres.
Bobby Clarke looks great based on this metric. Officially he only had four years as a first- or second-team all-star, with an identical record to Brad Marchand, Rick Martin, and Teemu Selanne. But he was a top-six forward ten times over the course of his career, which has only been topped by Gretzky post-expansion.
A lot of the top forwards from the eighties and nineties (Yzerman, Stastny, Gilmour, Hawerchuk, ec) also fare much better here. They were at a particular disadvantage because Gretzky took up one of the two spots available to centres 11 years in a row (1980 to 1991), and Lemieux took up four more of those spots (1986 to 1989). Thus you have situations like Steve Yzerman having an all-time great season (1989), and being clearly the best forward in the NHL after the two freaks of nature. But he didn't get an all-star spot that season, but his linemate Gerrard Gallant did (Yzerman outscored him by 62 points).
Players most helped by looking at each forward position separately
There are 24 players who have at least two fewer years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. Surprisingly, Wayne Gretzky is the only centre in this category; everybody else is a winger.
Luc Robitaille has eight year-end all-star selections at LW (five on the first team and three on the second team). He doesn't have a single selection under the "all forwards" framework, as he never placed in the top six in Hart voting. This doesn't mean that Robitaille is a bad player, or that he's overrated. It just means that it's misleading to compare his all-star record directly to players from other positions, because he had comparatively weak competition at LW.
From 1967-68 to 2020-21, 129 forwards have earned 318 spots on the NHL's official year-end all-star team (this is a span of 53 seasons, and 53 * 6 = 318). 131 forwards have earned 324 spots in the top three in Hart voting for forwards (there are a few more players in this category due to ties - I haven't attempted to break any ties that exist). In total, we have 170 players who appear in either or both categories.
Just 11 players, out of 170 in total, have the same voting outcome under both frameworks. An example is Pavel Bure - he was a first-team all-star at RW once, and a second-team all-star at RW twice. He also had one year where he finished in the top three in forwards among Hart voting, and two years where he finished between 4th and 6th (which I'm treating as the equivalent of the second all-forward team).
Players most helped by looking at top three forwards
| Player | 1st position | 2nd position | Position total | 1st forward | 2nd forward | Forward total | Difference |
| Bobby Clarke | 2 | 2 | 4 | 6 | 4 | 10 | 6 |
| Steve Yzerman | 1 | 1 | 2 | 4 | 6 | 5 | |
| Darryl Sittler | 1 | 1 | 1 | 5 | 6 | 5 | |
| Peter Stastny | 4 | 4 | 4 | ||||
| Doug Gilmour | 3 | 3 | 3 | ||||
| Jean Ratelle | 1 | 1 | 2 | 2 | 4 | 3 | |
| Dale Hawerchuk | 1 | 1 | 1 | 3 | 4 | 3 | |
| Joe Sakic | 3 | 3 | 1 | 5 | 6 | 3 | |
| Red Berenson | 2 | 2 | 2 | ||||
| Brian Sutter | 1 | 1 | 2 | 2 | |||
| Jeremy Roenick | 1 | 1 | 2 | 2 | |||
| Ron Francis | 1 | 1 | 2 | 2 | |||
| Claude Giroux | 1 | 1 | 2 | 1 | 3 | 2 | |
| Stan Mikita | 1 | 1 | 2 | 3 | 1 | 4 | 2 |
| Nathan MacKinnon | 2 | 2 | 3 | 1 | 4 | 2 | |
| Marcel Dionne | 2 | 2 | 4 | 5 | 1 | 6 | 2 |
| Nicklas Backstrom | 2 | 2 | 2 | ||||
| Pete Mahovlich | 2 | 2 | 2 | ||||
| Ryan Getzlaf | 1 | 1 | 1 | 2 | 3 | 2 | |
| Sergei Fedorov | 1 | 1 | 1 | 2 | 3 | 2 | |
| Eric Lindros | 1 | 1 | 2 | 2 | 2 | 4 | 2 |
| Sidney Crosby | 4 | 4 | 8 | 8 | 2 | 10 | 2 |
| Jonathan Toews | 1 | 1 | 3 | 3 | 2 | ||
| Mike Modano | 1 | 1 | 3 | 3 | 2 | ||
| Tim Kerr | 1 | 1 | 3 | 3 | 2 | ||
| Gilbert Perreault | 2 | 2 | 1 | 3 | 4 | 2 | |
| Joe Thornton | 1 | 3 | 4 | 3 | 3 | 6 | 2 |
There are 27 forwards who have at least two more years as an all-forward (compared to the official, positional selections). Nearly all of them are centres (the only exceptions are Kerr and Giroux). There are eight players with three more years as an all-forward, and all of them are centres.
Bobby Clarke looks great based on this metric. Officially he only had four years as a first- or second-team all-star, with an identical record to Brad Marchand, Rick Martin, and Teemu Selanne. But he was a top-six forward ten times over the course of his career, which has only been topped by Gretzky post-expansion.
A lot of the top forwards from the eighties and nineties (Yzerman, Stastny, Gilmour, Hawerchuk, ec) also fare much better here. They were at a particular disadvantage because Gretzky took up one of the two spots available to centres 11 years in a row (1980 to 1991), and Lemieux took up four more of those spots (1986 to 1989). Thus you have situations like Steve Yzerman having an all-time great season (1989), and being clearly the best forward in the NHL after the two freaks of nature. But he didn't get an all-star spot that season, but his linemate Gerrard Gallant did (Yzerman outscored him by 62 points).
Players most helped by looking at each forward position separately
| Player | First position | Second position | Position total | First forward | Second forward | Position total | Difference |
| Luc Robitaille | 5 | 3 | 8 | 0 | -8 | ||
| Jari Kurri | 2 | 3 | 5 | 0 | -5 | ||
| Richard Martin | 2 | 2 | 4 | 0 | -4 | ||
| Michel Goulet | 3 | 2 | 5 | 1 | 1 | -4 | |
| Alexander Ovechkin | 8 | 3 | 11 | 5 | 2 | 7 | -4 |
| Bill Barber | 1 | 2 | 3 | 0 | -3 | ||
| Frank Mahovlich | 1 | 2 | 3 | 0 | -3 | ||
| Steve Shutt | 1 | 2 | 3 | 0 | -3 | ||
| Martin St. Louis | 1 | 4 | 5 | 2 | 2 | -3 | |
| Cam Neely | 4 | 4 | 1 | 1 | -3 | ||
| Paul Kariya | 3 | 2 | 5 | 1 | 1 | 2 | -3 |
| John LeClair | 2 | 3 | 5 | 2 | 2 | -3 | |
| Alex Mogilny | 2 | 2 | 0 | -2 | |||
| Charlie Simmer | 2 | 2 | 0 | -2 | |||
| John Bucyk | 1 | 1 | 2 | 0 | -2 | ||
| Keith Tkachuk | 2 | 2 | 0 | -2 | |||
| Ken Hodge | 2 | 2 | 0 | -2 | |||
| Rod Gilbert | 1 | 1 | 2 | 0 | -2 | ||
| Vladimir Tarasenko | 2 | 2 | 0 | -2 | |||
| Jamie Benn | 2 | 1 | 3 | 1 | 1 | -2 | |
| Wayne Gretzky | 8 | 7 | 15 | 13 | 13 | -2 | |
| Brendan Shanahan | 2 | 1 | 3 | 1 | 1 | -2 | |
| Kevin Stevens | 1 | 2 | 3 | 1 | 1 | -2 | |
| Yvan Cournoyer | 4 | 4 | 2 | 2 | -2 |
There are 24 players who have at least two fewer years as a top three forward (based on Hart voting), rather than looking at things strictly on a positional basis. Surprisingly, Wayne Gretzky is the only centre in this category; everybody else is a winger.
Luc Robitaille has eight year-end all-star selections at LW (five on the first team and three on the second team). He doesn't have a single selection under the "all forwards" framework, as he never placed in the top six in Hart voting. This doesn't mean that Robitaille is a bad player, or that he's overrated. It just means that it's misleading to compare his all-star record directly to players from other positions, because he had comparatively weak competition at LW.
Last edited:
Hockey Outsider
Registered User
- Jan 16, 2005
- 9,497
- 15,814
Summary of players under the "top three forwards" framework
| Player | 1st team all-F | 2nd team all-F | Total |
| Wayne Gretzky | 13 | 13 | |
| Sidney Crosby | 8 | 2 | 10 |
| Bobby Clarke | 6 | 4 | 10 |
| Mario Lemieux | 9 | 9 | |
| Jaromir Jagr | 6 | 2 | 8 |
| Phil Esposito | 6 | 1 | 7 |
| Alexander Ovechkin | 5 | 2 | 7 |
| Mike Bossy | 3 | 4 | 7 |
| Guy Lafleur | 6 | 6 | |
| Marcel Dionne | 5 | 1 | 6 |
| Joe Thornton | 3 | 3 | 6 |
| Steve Yzerman | 2 | 4 | 6 |
| Darryl Sittler | 1 | 5 | 6 |
| Joe Sakic | 1 | 5 | 6 |
| Bryan Trottier | 4 | 1 | 5 |
| Connor McDavid | 3 | 2 | 5 |
| Mark Messier | 3 | 2 | 5 |
| Teemu Selanne | 3 | 2 | 5 |
| Brett Hull | 3 | 1 | 4 |
| Nathan MacKinnon | 3 | 1 | 4 |
| Stan Mikita | 3 | 1 | 4 |
| Bobby Hull | 2 | 2 | 4 |
| Eric Lindros | 2 | 2 | 4 |
| Jean Ratelle | 2 | 2 | 4 |
| Patrick Kane | 2 | 2 | 4 |
| Dale Hawerchuk | 1 | 3 | 4 |
| Gilbert Perreault | 1 | 3 | 4 |
| Peter Stastny | 4 | 4 | |
| Doug Gilmour | 3 | 3 | |
| Evgeni Malkin | 3 | 3 | |
| Jarome Iginla | 3 | 3 | |
| Markus Naslund | 3 | 3 | |
| Claude Giroux | 2 | 1 | 3 |
| Gordie Howe | 1 | 2 | 3 |
| Henrik Sedin | 1 | 2 | 3 |
| Nikita Kucherov | 1 | 2 | 3 |
| Pavel Bure | 1 | 2 | 3 |
| Peter Forsberg | 1 | 2 | 3 |
| Ryan Getzlaf | 1 | 2 | 3 |
| Sergei Fedorov | 1 | 2 | 3 |
| Brad Marchand | 3 | 3 | |
| Jonathan Toews | 3 | 3 | |
| Mike Modano | 3 | 3 | |
| Tim Kerr | 3 | 3 | |
| Jean Beliveau | 2 | 2 | |
| John Tavares | 2 | 2 | |
| Martin St. Louis | 2 | 2 | |
| Red Berenson | 2 | 2 | |
| Theoren Fleury | 2 | 2 | |
| Brian Sutter | 1 | 1 | 2 |
| Denis Savard | 1 | 1 | 2 |
| Jeremy Roenick | 1 | 1 | 2 |
| Pat LaFontaine | 1 | 1 | 2 |
| Paul Kariya | 1 | 1 | 2 |
| Pavel Datsyuk | 1 | 1 | 2 |
| Ron Francis | 1 | 1 | 2 |
| Steven Stamkos | 1 | 1 | 2 |
| Adam Oates | 2 | 2 | |
| Danny Gare | 2 | 2 | |
| Ilya Kovalchuk | 2 | 2 | |
| John LeClair | 2 | 2 | |
| Mark Recchi | 2 | 2 | |
| Mats Sundin | 2 | 2 | |
| Nicklas Backstrom | 2 | 2 | |
| Pete Mahovlich | 2 | 2 | |
| Yvan Cournoyer | 2 | 2 | |
| Adam Graves | 1 | 1 | |
| Alexei Yashin | 1 | 1 | |
| Anze Kopitar | 1 | 1 | |
| Artemi Panarin | 1 | 1 | |
| Auston Matthews | 1 | 1 | |
| Corey Perry | 1 | 1 | |
| Daniel Sedin | 1 | 1 | |
| Dave Keon | 1 | 1 | |
| Eric Staal | 1 | 1 | |
| Jamie Benn | 1 | 1 | |
| Leon Draisaitl | 1 | 1 | |
| Owen Nolan | 1 | 1 | |
| Rick Middleton | 1 | 1 | |
| Steve Larmer | 1 | 1 | |
| Taylor Hall | 1 | 1 | |
| Vic Hadfield | 1 | 1 | |
| Vincent Lecavalier | 1 | 1 | |
| Aleksander Barkov | 1 | 1 | |
| Alex Kovalev | 1 | 1 | |
| Bob Gainey | 1 | 1 | |
| Bob Nevin | 1 | 1 | |
| Brendan Shanahan | 1 | 1 | |
| Cam Neely | 1 | 1 | |
| Clark Gillies | 1 | 1 | |
| Daniel Alfredsson | 1 | 1 | |
| Dany Heatley | 1 | 1 | |
| David Pastrnak | 1 | 1 | |
| Denis Maruk | 1 | 1 | |
| Dennis Hextall | 1 | 1 | |
| Doug Weight | 1 | 1 | |
| Jack Eichel | 1 | 1 | |
| Jason Spezza | 1 | 1 | |
| Joe Mullen | 1 | 1 | |
| Joe Pavelski | 1 | 1 | |
| John MacLean | 1 | 1 | |
| John Tonelli | 1 | 1 | |
| Johnny Gaudreau | 1 | 1 | |
| Kent Nilsson | 1 | 1 | |
| Kevin Stevens | 1 | 1 | |
| Kirk Muller | 1 | 1 | |
| Lanny McDonald | 1 | 1 | |
| Marian Hossa | 1 | 1 | |
| Mark Pavelich | 1 | 1 | |
| Michel Goulet | 1 | 1 | |
| Mickey Redmond | 1 | 1 | |
| Neal Broten | 1 | 1 | |
| Norm Ullman | 1 | 1 | |
| Orland Kurtenbach | 1 | 1 | |
| Patrice Bergeron | 1 | 1 | |
| Patrick Marleau | 1 | 1 | |
| Patrik Elias | 1 | 1 | |
| Peter Bondra | 1 | 1 | |
| Phil Goyette | 1 | 1 | |
| Pierre Turgeon | 1 | 1 | |
| Rick Dudley | 1 | 1 | |
| Rick Nash | 1 | 1 | |
| Rick Tocchet | 1 | 1 | |
| Rod Brind'Amour | 1 | 1 | |
| Ross Lonsberry | 1 | 1 | |
| Ryan Kesler | 1 | 1 | |
| Syl Apps Jr | 1 | 1 | |
| Terry O'Reilly | 1 | 1 | |
| Todd Bertuzzi | 1 | 1 | |
| Tyler Seguin | 1 | 1 | |
| Zach Parise | 1 | 1 |
delete this table |
Last edited:
Hockey Outsider
Registered User
- Jan 16, 2005
- 9,497
- 15,814
Self-criticisms
I'll offer some self-criticisms of this approach. Someone can dismiss this analysis by saying that the Hart trophy voters don't know what they're doing. That may be true, but they're the same group that votes for the year-end all-star teams. So this objection is self-defeating; if you dismiss the Hart voters, you're dismissing the all-star voters too.
That being said - there are two ways in which the Hart trophy voters appear to be systematically biased (in a way that doesn't appear in all-star voting). The first is - in Hart trophy voting, there's a strong bias against players who miss the playoffs, and that doesn't seem to exist in all-star voting. For example, in 1994, Wayne Gretzky won the Art Ross trophy, but missed the playoffs. He placed 2nd in all-star voting at C, but didn't get a single vote for the Hart trophy. We see similar things happen in 2013 (St. Louis won the Art Ross on a non-playoff team and placed 2nd in all-star voting at RW, but only 8th among forwards in Hart voting), 2015 (Benn won the Art Ross on a non-playoff team and placed 2nd in all-star voting at LW, but only 8th among forwards in Hart voting) and 2018 (discussed previously). There are exceptions of course (Iginla won the Art Ross in 2002 and still finished 1st among forwards in Hart voting, as did Lemieux in 1988). Still, if a player consistently missed the playoffs, my approach (using Hart voting) is probably unfair to them.
The second difference is the Hart is supposed to go to the player "most valuable to his team", while the all-star spots are supposed to go to the best player at their position. Generally there's a big overlap between these two categories, but not always. Jari Kurri (who drops from five all-star spots in real life, to zero under my approach) appears to be hit hardest by this. He was a top-six forwards in the NHL at least twice in his career (1985 and 1986 - and probably a few more seasons too), but he got hardly any Hart votes because it's hard to be "most valuable" when his centre is the greatest player in hockey history. In 1986, for example, it was probably literally true that Neal Broten was more valuable to Minnesota than Kurri was to the Oilers - but Kurri was almost certainly the better player, and had the better season, even though the Hart voting doesn't reflect that. Michel Goulet, who spent most of his prime with Peter Stastny, also appears to be impacted in a similar way.
The final point relates to data quality. This isn't an issue from 1995-96 onwards (when the NHL allowed each voter to make five selections for the Hart). Prior to that, there were only three spots on each ballot. As a result, a player could sneak into the top six in voting with a miniscule number of votes. This was especially true in the 1980's, when Gretzky monopolized nearly all of the first-place votes. The most extreme example was in 1983, when four players finished tied for 6th in Hart voting among forwards (Bossy, Tonelli, Sutter, and Pavelich) on the strength of a single vote. This is statistical noise, but I haven't attempted to account for this. I've taken the Hart voting data as is, even if a player's ranking is based on one or two votes. As a result, there are a few questionable choices (like Mark Pavelich finished 6th among forwards that one year) - but these occurrences are rare, and pretty much non-existent from 1995-96 onwards.
As always, comments, questions, corrections, and personal attacks are welcome.
I'll offer some self-criticisms of this approach. Someone can dismiss this analysis by saying that the Hart trophy voters don't know what they're doing. That may be true, but they're the same group that votes for the year-end all-star teams. So this objection is self-defeating; if you dismiss the Hart voters, you're dismissing the all-star voters too.
That being said - there are two ways in which the Hart trophy voters appear to be systematically biased (in a way that doesn't appear in all-star voting). The first is - in Hart trophy voting, there's a strong bias against players who miss the playoffs, and that doesn't seem to exist in all-star voting. For example, in 1994, Wayne Gretzky won the Art Ross trophy, but missed the playoffs. He placed 2nd in all-star voting at C, but didn't get a single vote for the Hart trophy. We see similar things happen in 2013 (St. Louis won the Art Ross on a non-playoff team and placed 2nd in all-star voting at RW, but only 8th among forwards in Hart voting), 2015 (Benn won the Art Ross on a non-playoff team and placed 2nd in all-star voting at LW, but only 8th among forwards in Hart voting) and 2018 (discussed previously). There are exceptions of course (Iginla won the Art Ross in 2002 and still finished 1st among forwards in Hart voting, as did Lemieux in 1988). Still, if a player consistently missed the playoffs, my approach (using Hart voting) is probably unfair to them.
The second difference is the Hart is supposed to go to the player "most valuable to his team", while the all-star spots are supposed to go to the best player at their position. Generally there's a big overlap between these two categories, but not always. Jari Kurri (who drops from five all-star spots in real life, to zero under my approach) appears to be hit hardest by this. He was a top-six forwards in the NHL at least twice in his career (1985 and 1986 - and probably a few more seasons too), but he got hardly any Hart votes because it's hard to be "most valuable" when his centre is the greatest player in hockey history. In 1986, for example, it was probably literally true that Neal Broten was more valuable to Minnesota than Kurri was to the Oilers - but Kurri was almost certainly the better player, and had the better season, even though the Hart voting doesn't reflect that. Michel Goulet, who spent most of his prime with Peter Stastny, also appears to be impacted in a similar way.
The final point relates to data quality. This isn't an issue from 1995-96 onwards (when the NHL allowed each voter to make five selections for the Hart). Prior to that, there were only three spots on each ballot. As a result, a player could sneak into the top six in voting with a miniscule number of votes. This was especially true in the 1980's, when Gretzky monopolized nearly all of the first-place votes. The most extreme example was in 1983, when four players finished tied for 6th in Hart voting among forwards (Bossy, Tonelli, Sutter, and Pavelich) on the strength of a single vote. This is statistical noise, but I haven't attempted to account for this. I've taken the Hart voting data as is, even if a player's ranking is based on one or two votes. As a result, there are a few questionable choices (like Mark Pavelich finished 6th among forwards that one year) - but these occurrences are rare, and pretty much non-existent from 1995-96 onwards.
As always, comments, questions, corrections, and personal attacks are welcome.
JackSlater
Registered User
- Apr 27, 2010
- 19,288
- 14,634
Just from looking at the results and considering the basic logic, this adjustment seems to paint a more accurate picture of how players were perceived than do the actual results. We all know that certain positions (centre usually, RW in the 50s etc.) are more loaded when it comes to all star voting than others are and how that distorts all star results. The main critique that I can think of, which is that Hart voting and all star voting are ostensibly based on similar but different criteria, has already been noted. I'd add to that a specific example of Esposito finishing as first all star at centre and Clarke finishing as second all star at centre in 1973, when Clarke won the Hart. I also see that Ovechkin's 2013 second place RW finish is seemingly not counted, which I like.
Anyway, good work and an interesting read.
Anyway, good work and an interesting read.
I think it’s an interesting thing to look at, though I do think there’s enough discrepancies between Hart voting and player evaluation that I’m not sure if it’s something worth using too thoroughly comparisons, whereas I think an actual evaluation of top 3 and top 6 forwards would be. It would actually be an interesting to do polls on this for each season and see what the results end up being.
MadLuke
Registered User
- Jan 18, 2011
- 10,768
- 6,260
Could be interesting to confirm or refute global trend (like LW being the easiest to get in, center the hardest), etc....
But it could be that their is an Hart bias toward center (would be strange to have one for RW to LW, so that is weak)
To judge actual player, how much missing the playoff and how much having someone just a little bit has good has you on your team or has good having a giant Hart impact but no all star team impact if it is 2 different position (even being helped because you make the playoff more playing with a super stars) is large.
But it could be that their is an Hart bias toward center (would be strange to have one for RW to LW, so that is weak)
To judge actual player, how much missing the playoff and how much having someone just a little bit has good has you on your team or has good having a giant Hart impact but no all star team impact if it is 2 different position (even being helped because you make the playoff more playing with a super stars) is large.
VanIslander
20 years of All-Time Drafts on HfBoards
Interested to hear comments about a Hart bias towards Centres. I am presuming that this based on voting not being aligned with Art Ross scoring finishes.
I get the sense that a C will always get the nod over a W if their offense is similar given the perception of a C being the equivalent of a the QB/point guard among the forward group and their greater defensive responsibilities/impact. And when the concept of "value" gets added in, that can exaggerate that bias.
Off the top of my head, I cannot recall a C winning the Hart over a W that was really controversial or even Hart nominations. Any examples of the bias that seem obvious?
I get the sense that a C will always get the nod over a W if their offense is similar given the perception of a C being the equivalent of a the QB/point guard among the forward group and their greater defensive responsibilities/impact. And when the concept of "value" gets added in, that can exaggerate that bias.
Off the top of my head, I cannot recall a C winning the Hart over a W that was really controversial or even Hart nominations. Any examples of the bias that seem obvious?
Michael Farkas
Celebrate 68
Bias has a negative connotation. Generally speaking, the center is the driving force behind a line. The best players play center, it's the most ice to manage, the most responsibility...going way back to the O6 era, most of the time it was center carry situations - in fact, some teams, like the post-lockout Devils, also went down this road with Scott Gomez - a smaller, knock-kneed skater - being a center, three-line carrier.
Wingers skate the least and have the least responsibility in almost every situation. Obviously, there are some exceptions - like Patrick Kane - but most wingers are the end of the line for a play...
As opposed to bias, HO is further illustrating the known dearth of LWers in history. They get the same "value" out of post-season AST nods, but as a rule, they haven't been as impactful per AST nod as the other forward positions, so to speak...
Wingers skate the least and have the least responsibility in almost every situation. Obviously, there are some exceptions - like Patrick Kane - but most wingers are the end of the line for a play...
As opposed to bias, HO is further illustrating the known dearth of LWers in history. They get the same "value" out of post-season AST nods, but as a rule, they haven't been as impactful per AST nod as the other forward positions, so to speak...
The Panther
Registered User
It's interesting, but (as the OP notes) it's too discriminatory against elite players who played alongside super-elites. Jari Kurri was mentioned. Looking at the entire 1980s (1980-81 through 1989-90, which is Kurri's entire Edmonton career), he may have an argument as the 2nd-best forward of that ten-year period, if you consider the ten years cumulatively, including playoffs, two-way play, clutch factors, etc. So, to push him that far down the list is an indignity I will not bear!
(Nice job as always, though by @Hockey Outsider.)
(Nice job as always, though by @Hockey Outsider.)
There aren't many controversial hart wins period - few and far between. So specifically about C over W is also going to be rare. Here's a few possible ones though:
1. 2010 Sedin over Ovechkin. Ovechkin was "better" - but played a few less games. Also, on a powerhouse team. Was Sedin > Ovechkin controversial for hart? Probably a bit. Voter fatigue being a thing, it's a bit more expected.
2. 2006 - Thornton over Jagr. To me this was very much a 2 horse race, with Thornton edging out Jagr. Not sure controversial as much as someone had to win. I personally like Thornton being traded mid-year as an extra good narrative, so I had no issues with this
3. 2003 - Forsberg over Naslund. I think here very much the narrative of C over W favors Forsberg (whose also great defensively). Controversial - maybe a bit, but more to do with how ~1 month prior Naslund seemed like the obvious winner, but Forsberg finished super strong
4. 2001 - Sakic over Jagr. I think Sakic also got some bonus points for C over W (but Jagr also had slow start, and the Lemieux factor, to explain his hart loss).
So - I guess there are a few close ones. Insomuch as there are many controversial hart winners period.
I've said for a long time that we can't compare all-star nods at face value across positions - and also eras. This absolutely confirms the first one.
BigBadBruins7708
Registered User
Not necessarily bias, but there is an inherent pecking order in Hart voting. It's easier for an average center to get Hart votes than an average winger. basically the further down the list you go the more exceptional you have to be to get the same Hart votes.
Center
Wing
Defense
Goalie
Sakic over Jagr happens I think even without Mario I.e. if Jagr puts up the numbers he did with no Mario, due to Sakic finishing high in Selke voting (and Jagr being notably very lax in the defensive end).
I agree with your narratives here although Forberg's superior PPG, IMO, gives him a clear edge over Naslund. I don't see any pattern, at least at the high end of Hart voting that puts a C ahead of a W automatically. I think we clearly see notable 2-way Cs getting Hart recognition lower down like Crosby in 2019 and 2021.
JackSlater
Registered User
- Apr 27, 2010
- 19,288
- 14,634
Different topic but I don't see any edge for wingers over centres in Hart voting. Centres are usually better players and have more responsibility than wingers do. In cases like Sakic over Jagr in 2001 or Forsberg over Naslund in 2003 it's obvious that the centre was the better (and more valuable player I suppose) in each year. If wingers were really hurt in Hart voting you'd probably see something like Malkin taking the 2009 Hart over Ovechkin or MacKinnon over Hall in 2018. The advantage clearly seems to be forwards over defencemen in Hart voting, with goaltenders and the Hart almost being a separate thing.
The Macho King
Back* to Back** World Champion
- Jun 22, 2011
- 49,080
- 30,015
This is where my alarm bells start ringing on C bias. Centers have more responsibilities *generally*, but just by virtue of the players we're talking about in a discussion of postseason all-stars, we're talking about outliers.
The most famous case is Gretzky and Kurri. Beyond taking faceoffs, what traditionally "defensive" responsibility did Gretzky take as a C? Kurri often took those responsibilities.
It's also antiquated in the F1F2F3 era of defensive coverage/forechecking. Forwards have become more and more positionless except for faceoffs. And faceoffs as an important event we have recognized more and more as time goes that they simply... aren't.
That's to say - Centers are overrated for their defensive contributions *as a whole*. Individual Cs obviously contribute a ton in some areas (Bergeron) and less in others (Scheifele/Draisatl/McDavid). It's especially annoying to me when we say "X player has more responsibility because they play C" and it's someone like Draisatl who hasn't backchecked a day in his life in comparison to a guy like (just for example sake) Kucherov who... similarly isn't particularly interested in that end of the ice.
It's a way to bolster completely one-way players by attaching the reputation of an entire class of players to which the comparison falls apart with even one critical look.
Michael Farkas
Celebrate 68
This is a really good point about F1/F2/F3. It's about timing usually, and less so about positions unless someone has a specialty.
Gretzky/Kurri: Gretzky was such an amazing boon in transition, in the NZ, in clean zone entries that it made sense to have a much stronger player in Kurri play down in the weeds.
Centers still do traditionally work their way into the lowest part of the defensive zone and take the net-front when a D goes behind the net to address the puck carrier. In transition, yeah, F1 is positionless. But when you work your way into in-zone coverage, some positional "integrity" returns.
There are obviously examples - as you note - about Scheifele not playing defense, while Jere Lehtinen did. But as a whole, centers do have to manage more ice and they're more vulnerable to attack because there isn't anywhere to "hide" - wingers can "hide" on the boards, get their wits about them, and try to advance play. Centers are predator and prey all at the same time in more instances than their counterparts...
As always, everything requires proper talent evaluation and context...but this is a macro-level thread,, it doesn't have to be though...
Gretzky/Kurri: Gretzky was such an amazing boon in transition, in the NZ, in clean zone entries that it made sense to have a much stronger player in Kurri play down in the weeds.
Centers still do traditionally work their way into the lowest part of the defensive zone and take the net-front when a D goes behind the net to address the puck carrier. In transition, yeah, F1 is positionless. But when you work your way into in-zone coverage, some positional "integrity" returns.
There are obviously examples - as you note - about Scheifele not playing defense, while Jere Lehtinen did. But as a whole, centers do have to manage more ice and they're more vulnerable to attack because there isn't anywhere to "hide" - wingers can "hide" on the boards, get their wits about them, and try to advance play. Centers are predator and prey all at the same time in more instances than their counterparts...
As always, everything requires proper talent evaluation and context...but this is a macro-level thread,, it doesn't have to be though...
The Macho King
Back* to Back** World Champion
- Jun 22, 2011
- 49,080
- 30,015
I guess my point is I don't think this is a macro-level thread. We're talking about outliers here already. With a few notable exceptions, none of these guys who are in discussion at the top end of this are more than "meh" defensively, and quite a few were actively terrible. That's where the "Center has more responsibility" just falls apart for me. Lines 2 through 4 (and frankly lines 1 on at least half the league)? Sure, we can make that assumption and feel pretty safe. I don't think that holds true "as a unit" for a lot of star players though.
JackSlater
Registered User
- Apr 27, 2010
- 19,288
- 14,634
You can apply the "usually" to the part about defensive responsibility as well, which I should have indicated. There are examples, as you suggest, where the centre is not the forward tasked with the most defensive responsibility. In general it is still very much the case though, at least in recent decades. Pre-expansion there shouldn't be much distinction when it was a lot more winger vs winger.
Dingo
Registered User
- Jul 13, 2018
- 1,938
- 1,948
Hart voting is just as flawed as Allstar voting.
i mean that, while you get Yzerman not making the top 6 in 89, while his linemate Gallant did, out of Allstar voting, its not any better with voting based on, “the best player to make the playoffs, except also not be on a really solid team where there is another arguably better player.... unless we have all really got excited about a narrative and cant wait to bestow something on this guy” award
i mean that, while you get Yzerman not making the top 6 in 89, while his linemate Gallant did, out of Allstar voting, its not any better with voting based on, “the best player to make the playoffs, except also not be on a really solid team where there is another arguably better player.... unless we have all really got excited about a narrative and cant wait to bestow something on this guy” award
The Panther
Registered User
A bit hard to take your opinion seriously if you actually think this is true.
Michael Farkas
Celebrate 68
I'm not sure I'm following this point. Yzerman finished 3rd in AS voting at C and finished 3rd in Hart voting in '89. It could be just late-day brain fog though...
Dingo
Registered User
- Jul 13, 2018
- 1,938
- 1,948
as flawed as the allstar team selections by position are in determining who the three actual best forwards on a given year were, the Hart is also flawed.
I was only using Yzerman as an example of the AS teams weakness, and i didnt give an example for the Hart... i probably shouldnt have given an example at all and just stuck with the my first sentence in this post.
The Hart is a flawed way of determining the three best forwards in a year, and so is the year end AS team, imo.
Michael Farkas
Celebrate 68
Yeah, I think that's already acknowledged by the OP and by everyone reading this thread already, right?
