July 7, 2010
Up and Coming
Inductinator: All-Star Teams

by Iain Fyffe

In theory, the players inducted into the Hockey Hall of Fame represent the best at their position for their time. It's difficult to argue that a left wing who was never best than the fourth-best left wing in the league should be enshrined, unless perhaps if the first three have already been enshrined. Even then, the question of true greatness would arise. If he was never the best at his position, can he really be considered one of the greatest of all time?

But how would the Hall of Fame voters (a committee made up of 18 members, including Hall of Fame inductees, hockey writers and others associated with the game) determine who was the best at their position? One clue could be to look at the postseason All-Star Team selections. Clearly a player who is selected for an All-Star Team was perceived to be the best at his position; the question of whether he actually was the best is irrelevant for our purposes. We are attempting to predict who will be inducted into the Hall of Fame, not necessarily who should be. The All-Star Team selections are generally good at identifying the best players in the NHL, so there should not be a significant difference between will and should in this case.

Once again, we'll be looking only at modern players, as defined last month(players whose careers are substantially in 1967-68 and later). Counting up the All-Star Team selections (both First- and Second-Team) for those modern players who have been eligible for induction at least one year is simple and produces the following results.

19 All-Star Teams: Ray Bourque
15 All-Star Teams: Wayne Gretzky
 9 All-Star Teams: Phil Esposito, Mario Lemieux
 8 All-Star Teams: Bobby Orr, Mike Bossy, Paul Coffey, Luc Robitaille
 7 All-Star Teams: Brad Park, Denis Potvin, Al MacInnis
 6 All-Star Teams: Ken Dryden, Guy Lafleur, Borje Salming, Larry Robinson, Patrick Roy
 5 All-Star Teams: Tony Esposito, Mark Messier, Michel Goulet,
 Jari Kurri, Scott Stevens, Brian Leetch
 4 All-Star Teams: Ed Giacomin, Yvan Cournoyer, Guy Lapointe,
 Bobby Clarke, Rick Martin, Marcel Dionne, Bryan Trottier, Cam Neely
 3 All -Star Teams: Bill White, Bill Barber, Steve Shutt, Doug Wilson,
 Rod Langway, Tom Barrasso, Kevin Stevens, Pavel Bure
 2 All -Star Teams: Ken Hodge, Pat Stapleton, Mickey Redmond, Bernie
 Parent, Rogie Vachon, Gil Perreault, Glenn Resch, Lanny McDonald, Clark 
Gillies, Don Edwards, Charlie Simmer, Mike Liut, John Tonelli, Grant Fuhr,
 John Vanbiesbrouck, Alexander Mogilny, Eric Desjardins
 1 All -Star Team: John McKenzie, Vic Hadfield, Dennis Hull, Wayne 
Cashman, Barry Ashbee, Steve Vickers, Rene Robert, Reggie Leach, Darryl 
Sittler, Serge Savard, Danny Gare, Jim Schoenfeld, Randy Carlyle, Dave Taylor, 
Mario Lessard, Billy Smith, Rick Middleton, Brian Engblom, Pete Peeters, Denis 
Savard, Rollie Melanson, Pat Riggin, John Ogrodnick, Pelle Lindbergh, Dale 
Hawerchuk, Mats Naslund, Bob Froese, Ron Hextall, Tim Kerr, Hakan Loob, Brad 
McCrimmon, Gary Suter, Joe Mullen, Gerard Gallant, Mike Vernon, Brian Bellows, 
Darren Puppa, Adam Oates, Phil Housley, Kirk McLean, Pat LaFontaine, Al 
Iafrate, Adam Graves, Alexei Zhamnov, Theo Fleury, Jim Carey, Vladimir 
Konstantinov, Byron Dafoe, Steve Yzerman

It's immediately obvious, reading the list from the top down, that a high number of All-Star Team selections is a big point in favor of being inducted into the Hall of Fame. The more eagle-eyed among you may have noticed that every player with five or more All-Star Team selections is in the Hall of Fame, and that if we extend it down to four or more appearances, there is but one player who is not in the Hall (Rick Martin). It makes you wonder what Martin did (or didn't do) that kept him out of the Hall. It doesn't seem like career-ending injuries are an issue, as long as you did enough before the injury (Cam Neely, Bobby Orr, Mike Bossy) to merit inclusion. Maybe it's because Gil Perreault was considered the Sabres' key player, and the voters didn't want to enshrine too many players from a team that never won the Cup. There's no way of knowing right now and perhaps as the Inductinator continues to develop we might get some insight. Or perhaps it's merely an oversight on the selection committee's part.

Predictably, the lower the number of All-Star selections, the less likely the player is to make the Hall of Fame:

All-Star Teams  Number of Players  Hall of Famers  HOF %
5 or more	22	           22	           100
4	         9	            8	            89
3	         8	            3	            38
2	        17	            5	            29
1	        50	            9	            18

There is a massive drop-off between four selections and three. It's quite rare, when looking at these types of things, to find such a clear “magic number”. If you have four All-Star team selections, you are almost assuredly in, since past modern players with such qualifications have made the Hall 97% of the time. (I hope Rick Martin's not reading this. Actually, it would be pretty cool if he were. So I hope he is reading this.)

We had an interesting case this year which put this to the test. In his career, John LeClair was a First Team All-Star twice, and a Second-Team All-Star three times, giving him five total selections. Based on the above, that would put him in the category of having a 100% chance of making the Hall. However, in his first year of eligibility (2010) he was not selected. So it may take several more years to know whether this relationship between All-Star teams and Hall of Fame induction holds up. I don't generally think of LeClair as a Hall of Fame player, but perhaps I should. He put up a heck of a lot of goals for the era he played in. Eventually the Inductinator will have an opinion on whether he meets the implicit Hall standards, but given his five All-Star Team selections, it does seem likely at this point that he'll get into the Hall of Fame in the years to come.

Other players who have already earned at least four All-Star team selections (through 2009) include Nicklas Lidstrom (ten), Martin Brodeur (seven), Rob Blake, Scott Niedermayer, Chris Pronger, Jarome Iginla, Zdeno Chara, and Alex Ovechkin (four each). There are several sure-fire Hall of Famers in this list, but a couple of question marks as well.

Of course, All-Star Team selections are not the be-all and end-all of Hall of Fame qualifications. There are five modern Hall of Famers with two All-Star selections, nine with one All-Star selection, and ten who never made an All-Star Team at all (Jacques Lemaire, Gerry Cheevers, Bob Gainey, Peter Stastny, Mike Gartner, Viacheslav Fetisov, Bernie Federko, Ron Francis, Igor Larionov and Glenn Anderson). If we break this group down by position, we come upon one of the most obvious weaknesses of using All-Star team selections in this way.

Of these Hall-of-Famers, fully half played center, while only 8% played defense, even though “defense” is really two positions. We would expect 16% of players at each position (33% for defencemen) if they were distributed randomly. This is a huge discrepancy, and it has nothing to do with sample size. There is a reason why the center number is so high, and the defense number is so low.

The number of centers is so high because the best forwards play center. The best center is almost always better than the best left wing or right wing, and this relationship holds as you move down through the second-best at each position, the third-best, the eight-best, the seventeenth-best, et cetera. Center is a more important position than wing, so that's where you put your best players. Since 1967-68, 27 of the 41 Hart trophies for league MVP have gone to centers, and even if you take out Gretzky's nine awards, the position still has a huge advantage.

The number of defensemen in the Hall is so low because defense consists of two positions on the ice, which means that defensemen have twice as many All-Star Team positions available. To make an All-Star Team, you only need to be among the top four defensemen in the league, while any other position requires you to be in the top two. This is one reason Ray Bourque was able to make 19 All-Star teams; even if he slips to being the fourth-best defenseman in the league, he can still get the honor.

In summary, there are two important observations that inform the development of the Inductinator. First, multiple All-Star team appearances are an important predictive factor for Hall of Fame enshrinement. The fourth such selection seems to be the most important. Second, an All-Star selection varies in value by position. It is less valuable for a defenseman, and much more valuable for a center. There's a third observation as well: you need more than All-Star team appearances to predict who will become a Hall of Famer. But we knew it wasn't going to be that easy.