Debunking the “Proven Winner” myth in the National Hockey League

by: Issues: May / June 2012. Tags: Leadership. Categories: Leadership.

Hiring a proven winner in major league sports is usually seen as the solution for jump-starting a moribund team. But not so fast, at least in the National Hockey league, where hiring a winning coach or general manager from another team can make a team that is already skating uphill struggle even more. This Ivey professor has lots of compelling data to support this argument.

The debate

In the past few years there has been an ongoing debate about just where National Hockey League teams should look to find their next General Manager and/or Coach. Should they hire experienced “proven winners” from other NHL teams? Or should they look for GMs/coaches who have never won a Stanley Cup with other teams? In December, 2011, the focus of the debate was the hiring of Randy Cunneyworth to replace the fired Jacques Martin as the interim head coach of the Montreal Canadiens.According to Associated Press writer Benjamin Shingler:

“Some fans suggested they wouldn’t be so upset if the Habs had brought in a highly regarded Stanley Cup winner, such as the Detroit Red Wings’ Mike Babcock, a McGill alumnus, instead of an assistant without NHL head coaching experience.”

On the other hand:

“Others said the most important thing is winning. They say that if Cunneyworth, a young, tech-savvy coach and former NHL captain, manages to revive the Habs, then he should stick around.[1]

This debate is reminiscent of a similar one surrounding the hiring of Brian Burke by the Toronto Maple Leafs as that team’s General Manager (GM) in 2008. Kevin McGran[2] wrote a provocative article titled, “Cup winners need not apply?” He described how the Leafs were insisting that their next GM would be a “proven winner.” This “proven winner” would be someone like Ken Holland, (who had won two Stanley Cups in 1998 and 2002 with the Detroit Red Wings and would win a third in June 2008, Brian Burke, who had won a Stanley Cup with Anaheim in 2007, or Jim Rutherford, who had won a Stanley Cup with Carolina in 2006.

Leafs management insisted that the new GM be a “proven winner” despite the fact that in the history of the National Hockey league, only one GM has won the Stanley Cup with more than one team. This exceptional GM was Tommy P. Gorman, who won seven Stanley Cups — three with the early edition of the Ottawa Senators (1920, 1921 and 1923), one with the Chicago Blackhawks (1934), one with the defunct Montreal Maroons (1935) and two with the Montreal Canadiens (1944 and 1946).

McGran suggested that should the Leafs hire a “proven winner,” their Stanley Cup drought would only be extended. I was quoted in the McGran article as saying that the Leafs were: “… being seduced by the glitter of someone having won the Stanley Cup,” if they kept on insisting on hiring a GM who had previously won the Cup with another team. Richard Peddie, then president and CEO of Maple Leaf Sports and Entertainment Ltd., the owner of the Leafs, suggested that he was not about to change his mind about going after a “proven winner.” He was quoted as saying:

“We’re looking for a proven winner and the best example is a Stanley Cup,” says Peddie. “If we end up hiring an individual who’s already won a Stanley Cup, I guess we’ll break the string.”[3]

In a 2010 article, Lance Hornby suggested that Peddie was only following popular public sentiment when he hired proven executives such as Brian Burke to lead the Leafs and Bryan Colangelo to lead the NBA Raptors.[4]                                                                  

Ken Campbell’s input

Two days later, Ken Campbell, a senior writer for The Hockey News, jumped into the debate and suggested that not going after a “proven winner” was a very bad idea.[5]

He went on to point out that as of February, 2008, only seven men had gone on to be the GM of an NHL team after leading another team to the Stanley Cup: these were the aforementioned Tommy Gorman, Bob Gainey (Dallas – 1999), Glen Sather (Edmonton – 1984, 1985, 1987, 1988 and 1990), Cliff Fletcher (Calgary – 1989), Bill Torrey (New York Islanders – 1980, 1981, 1982, and 1983), Milt Schmidt (Boston – 1970 and 1972) and Punch Imlach (Toronto – 1962, 1963, 1964 and 1967). He did acknowledge that only Gorman had won with more than one team. Campbell concluded that there was a one-in-seven chance of a “proven winner” leading a different team to a Stanley Cup, and that he would take these 14.3 percent odds anytime. Of course, this list now includes Brian Burke (Anaheim – 2007) and Jay Feaster (Tampa Bay – 2004), which means that the odds are now down to 11.1 percent.

In his article, Campbell then asked:

“You don’t think Sam Pollock or Frank Selke or Jack Adams would have won Cups if they had changed teams? Does that mean Lou Lamoriello or Holland would suddenly lose their ability to build a winner if they changed organizations?”

The surprising answer is “Maybe.” Maybe Pollock, Selke, and Adams would not have won had they changed teams; maybe Lamoriello and Holland would have lost their ability to build a winner had they changed organizations; and maybe that is why GMs like Pollock, Selke and Adams (and Allen, Dandurand, LaCroix, Craig Patrick, Lester Patrick, Art Ross, Serge Savard, and Smythe, to name multiple cup winners who won with the same team) stayed where they were. All of them realized how hard it is to win again with the same team and sensed that it is nearly impossible to win with another team.

Maybe Lamoriello and Holland (the only multiple Stanley Cup winners with the same team, New Jersey and Detroit, and who are still active as GMs in the NHL) have a better sense of history. Maybe they realize that their chances of winning another Stanley Cup are better with the team that they know and have already won with than trying to win with a different team.

Campbell answered his questions above by stating:

“Of course it doesn’t, which is why if the Maple Leafs search committee doesn’t do everything it possibly can to get permission to speak with Ken Holland, then hire him if they do. Then they are simply abdicating their responsibility and doing the organization a complete disservice.”

This is a pretty strong indictment of the Maple Leafs ownership. I agree with Campbell that Holland is the best GM (or least one of the two best if we include Lamierello) we have seen in the league since 1995. I would go further and say that these two GMs are among the best in the history of the NHL. They are among a group of only twelve GMs who have won three or more Stanley Cups with the same team from 1918 to 2011 and they have done it in a league that included 26 to 30 teams.[6]

What does history say?

Jason Farris, in his seminal new book on NHL general managers,[7] lists 174 individuals who have been an NHL GM from 1926-1927 to 2010-2011. Of these individuals, three (Lou Angotti – St. Louis, May to September, 1974; Howie Meeker – Toronto, May to October, 1957; and, Rudy Pilous – California, May 1966 to June 1967) did not serve their respective teams for even one game during the regular season. A fourth, Kevin Cheveldayoff, was hired on June 8, 2011 by Winnipeg, after the team had moved from Atlanta. This leaves 170 individuals who served as an NHL GM for at least a portion of a regular season.[8]

Thirty-two of these GMs have led their respective teams to the Stanley Cup. If we add the GMs who also led their NHL teams to a win between 1917 (the year the National Hockey League was formed) and 1926 (the last season before the NHL assumed de facto control of the Stanley Cup[9]), we can add two more names. Table One lists these individuals, the number of Stanley Cups each won and the teams with which they accomplished this feat.

Nine of these individuals were asked to pursue the Stanley Cup with another team after having won it with a previous team. These are noted in Table One with an asterisk. As mentioned, this lowers the odds from Campbell’s 14.3 percent to 11.1 percent. These odds, combined with the idea that perhaps many of these GMs decided to stay with the team with which they had already won, led to the following question (see Table Two): What are the odds that a GM who has won the Cup with one team can win again with the same team? Or, how many Cups does a GM have to win with the same team before we reach odds that are similar to 11.1 percent?

In Table Two we see that a GM has to win approximately 5 or more Cups with the same team (with odds of 13.5 percent) to have the same odds of winning another Cup with a different team.[10] This suggests that as hard as it is to win a Stanley Cup it is easier to win with the same team than a different team. (Only 32 out of the 170, or 18.8 percent of GMs who led teams from 1926-1927 to 2010-2011, won one or more Stanley Cups.)

Table Two also shows analyses for coaches and captains who have their names engraved on the Stanley Cup. Twenty-three coaches who won Stanley Cups with one team were invited to try to win with another team. Only three succeeded (Scotty Bowman, Dick Irvin and Tommy Gorman). This is a ratio of 13.0 percent, a slightly higher ratio than winning four or more Stanley Cups with the same team (11.3 percent) as a coach.

Regarding captains, three who won Cups were given the opportunity to do so with other teams (Wayne Gretzky – Edmonton (4), Los Angeles and Saint Louis; Darien Hatcher – Dallas (1) and Philadelphia; and, Mark Messier – Edmonton (1), New York (1), Vancouver and New York again). Only Messier succeeded – a 33.3 percent chance. As we see in Table Two, this is much lower than the odds of winning 2 or more Cups (43.8 percent) with the same team.

This suggests that the difficulty of winning with a “proven winner” – someone who has won with another team — is difficult at all three leadership levels in the NHL


Table One

Stanley Cup Winners as a General Manager





Pollack, Sam



Smythe, Conn



Adams, Jack



Gorman, Tommy*

Ottawa (3), Chicago (1), Maroons (1),

Montreal (2)


Selke, Jr. Frank



Sather, Glenn*



Imlach, Punch*



Torrey, Bill*

New York Islanders


Patrick, Lester

New York Rangers


Ross, Art



Lamoriello, Lou

New Jersey Devils


Holland, Ken



Querrie, Charles



Dandurand, Leo



Schmidt, Milt*



Allen, Keith



Savard, Serge



Patrick, Craig



LaCroix, Pierre



Gerard, Eddie



Gill, Dave

Ottawa (early edition)


Tobin, Bill



Ivan, Tommy



Fletcher, Cliff*



Grundman, Irving



Smith, Neil

New York Rangers


Devellano, Jim



Rutherford, Jim



Gainey, Bob*



Feaster, Jay*

Tampa Bay


Burke, Brian*



Shero, Ray



Bowman, Stan



Chiarelli, Peter



Table Two

Odds of Winning More Than One Stanley Cup With Same Team (1917-1918 to 2010-2011)







% age


% age


% age

Two or more







Three or more







Four or More







Five or More







Six or more








1) For the GM analysis, Tommy Gorman’s four times as a Stanley Cup winning GM were treated as if he were four different individuals. This increased the denominator in the GM ratios from 34 to 37.

2) For the coach analysis, Scotty Bowman’s three times, Dick Irvin’s two times and Tommy Gorman’s two times as Cup winning coaches were treated as if they were seven individuals. This increased the denominator from 49 to 53.

3) For the captain analysis, Mark Messier’s two times as a winning captain was treated as if he were two different individuals. This increased the denominator from 47 to 48.

How does the next team do?

Table Three presents the aggregate records for the 8 GMs (Tommy Gorman is excluded) who won with one team but not with the next team they served as the GM. These 8 GMs won 19 Cups with their initial teams and none with their next teams. In fact, these “next” teams did not even make it to the Stanley Cup finals.[11] Noteworthy is the fact that only one GM had a better regular season record with the second team, Jay Feaster with a 25-11-9 (0.656) record over the latter part of the 2010-2011 season with Calgary.


Table Three

 A Comparison of Performance Between Cup Winning Teams and Next Teams for GMS




Cup Teams

Next Team

Opportunities to be in Playoffs*



Years in Playoffs



% of years in Playoffs



Regular-season record



Winning %



Cup appearances



Cup wins



% of Wins/Appearances



Won/Loss % in playoffs



* Opportunities to be in the playoffs include coaches who joined their teams for the last part of a season and their teams made it into the playoffs.

Table Four presents the records for the 20 coaches (Scotty Bowman, Dick Irvin and Tommy Gorman were excluded) who won with an earlier team but not with their next team.

Table Four

 A Comparison of Performance Between Cup Winning Teams and Next Teams for Coaches




Cup Teams

Next Team

Opportunities to be in Playoffs*



Years in Playoffs



% of years in Playoffs



Regular-season record



Winning %



Cup appearances



Cup wins



% of Wins/Appearances



* Opportunities to be in the Playoffs include coaches who joined their teams for the last part of a season and their teams made it into the playoffs.

These 20 coaches won 29 Cups with their initial team and none with their subsequent teams.

Tables Three and Four suggest that, after being a “proven winner” with one team, winning with another team is extremely unlikely. Except for one of the 170 GMs who served in the NHL from 1926 to 2011 (Tommy Gorman) and three of 333 coaches who served in the NHL from 1917 to 2011 (Tommy Gorman, Dick Irvin and Scotty Bowman) it just did not happen.

Is this any different than in other major sports leagues? Jason Cole’s analysis suggests that it is not. In a recent Yahoo! Sports article, he examined the difficulty of winning a Super Bowl with a second team.[12]

Table 5

A Comparison of Performance Between Super Bowl Winning Teams and Next Teams

for National Football League Coaches


Super Bowl Teams

Next Teams

Seasons coached



Regular-season record



Post-season record



Super Bowl appearances



Super Bowl championships



Overall winning record (in %)



Table 5 presents Cole’s analyses of how the 12 Super Bowl winning coaches fared with their “next” team. (In Bill Parcells’ case, second, third and fourth times.) Twenty Super Bowls out of 23 appearances the first time versus none out of two appearances with their next team. This suggests that it is very difficult for a “proven winner” of a Super Bowl to repeat his success with another team. Cole states that, “… there is only one conclusion to draw about hiring a coach who has won a Super Bowl. Don’t.”[13]

Cole also mentioned (without analysis) that only two field managers have won MLB championships with more than one team (Tony LaRussa and Sparky Anderson) and only three coaches have won NBA championship with more than one team (Phil Jackson, Pat Riley and Alex Hannum).

Is it possible that it is just as difficult to transfer “proven winners” in Major League Baseball and the National Basketball Association as it is in the National Hockey League and the National Football League?

Why is it so difficult?

Building an NHL team by hiring a front office and scouting staff, firing and hiring the coach and his staff, and bringing together an appropriate group of players through drafts, trades and promotions from the minors/juniors is a daunting task. In today’s NHL, GMs have 82 games to build a team that will make it into the playoffs.

Lou Lamoriello fired New Jersey Devils’ coach Robbie Ftorek with 8 games left in the season. Ftorek’s replacement, Larry Robinson, led the Devils to a 4-4-0 record over those last eight games. In the playoffs, he led his team to a 16-7 record and the Stanley Cup.

In Behind The Moves, Lamoriello describes his decision:

“I let a coach go … It was Robbie Ftorek, one of the toughest decisions, but I knew—I just didn’t feel—I knew we had a chance to win, but I didn’t think we were going to recover [with Ftorek]. I can’t sit here and tell you I felt good about [the decision]. I made it. Nobody else made it. Nobody knew. I hired Larry Robinson and we won the cup.[14]

Coaching an NHL team to a Stanley Cup is just as difficult as being the GM – maybe even slightly more difficult (18.8 percent of GMs and 14.7 of coaches have led their teams to Stanley Cup wins).

Why is it so difficult?  Perhaps the main reason is that “proven winners” have developed socially complex relationships within their organizations. The winners cannot leverage these relationships when they switch to a new organization. The fact that they can’t be leveraged has been demonstrated in another field, capital markets. In the results of a study published in 2004, Harvard University professors Boris Groysberg, Ashish Nanda and Nitin Nohria, found that the performance of 1,052 star stock analysts in 78 U.S. investment banks plummeted by as much as 20 percent —permanently — after they were hired by a new firm. (The study tracked analysts’ performance between 1988-1996). .[15] The authors argued that only 30 percent of a star’s performance is attributable to individual capabilities, while70 percent is due to the resources and capabilities specific to the analysts’ former firms, where they developed into stars. Resources and capabilities include leadership, reputation, team chemistry, training, and information technologies, all of which the “star” leaves behind when he or she goes to another investment bank. The successful socially complex relationships mentioned earlier are also left behind and might never be established in another organization.

With the rare exception, this seems to be the case in the NHL. Only four individuals were able to achieve success in their new organizations even though they were not able to leverage the socially complex relationships they had developed in their former organization: Mark Messier as a captain, Scotty Bowman and Dick Irvin as coaches, and Tommy Gorman at the coach and GM level. Of course, Lester Patrick could be added to the list as one who won as a GM and a coach across not only different teams, but different leagues. A pretty amazing group of five people! To this point, all others have failed or declined to take advantage of the opportunity.

What advice do Groysberg, Nanda and Nohria have to offer?

  • Winning a star war might be the worst thing you could do for your organization.
  • Grow your stars [proven winners] internally; do not buy them on the market.

[Note: Ken Holland worked in the Detroit organization for twelve years—first as a scout, second as director of amateur scouting and third as assistant GM—before becoming the GM on July 18, 1997. As mentioned, he has led Detroit to three Cup wins and is one of only two GMs to do so in the last 24 years. Lou Lamoriello is the second. I do not believe that it is a coincidence that Lamoriello was the GM for 8 years (thereby, being allowed to develop internally) while he built a team that won New Jersey’s first cup (the first of three) in 1995.] 

  • Find and recruit bright people through a disciplined hiring strategy.
  • Develop them through training and mentoring, e.g., Sam Pollack had a 17- year mentorship and training period with Frank Selke, Sr.—arguably one of the greatest GMs to serve in the NHL.
  • Work to retain outstanding GMs.

I have often wondered why Sam Pollack left the Canadiens in the prime of his career as a GM. He resigned at the age of 52 after working with the team for 31 years – his last 14 as GM – a period that saw Montreal win 15 Cups and make it to the finals another six times. I have wondered why he never accepted any of the offers I assume he received from other NHL teams. I can only surmise that he concluded – astutely – that doing it again with another organization would be much more difficult.

What are the implications for owners/presidents of NHL teams who have a proven winner in their ranks?

  • Keep him. Do whatever it takes as long as the value created is more than the value required to retain. Support the “proven winner” with the appropriate structure, controls and rewards.

What are the implications for “proven winners”?

  • Stay with the team that developed you into a proven winner as long as they continue to reward, control and structure in a manner conducive to remaining a “proven winner.” Your chances of winning with the same team are better than with a different team.
  • Of course, there is one exception – if there are family reasons for moving to a new team—then move.

What are the implications for a team that needs a “proven winner”?

  • Find the brightest, youngest person in your organization (.e.g., Pollack and Holland).
  • Patiently mentor, develop and train this person.
  • If you need to go outside, find individuals who have been involved in winning organizations and give them time to learn your organization (e.g., Selke and Lamoriello).

We see this model in various forms in the stories of four NHL GMs. We have already mentioned Holland who was trained in the Detroit organization for 12 years.

Pollack has a similar story. He was hired by the parent Canadiens after being asked to scout local talent for the Junior Canadiens by retired Montreal goalie Wilf Cude. He was hired n 1947 and appointed the clubs director of player personnel in 1950. Under Frank Selke’s tutelage, he built the Canadiens’ farm system and led two junior teams (Montreal Junior Canadiens and Hull-Ottawa Junior Canadiens) to Memorial Cups in 1950 and 1958, respectively. In addition, he led two minor pro teams to championships. These were the Hull-Ottawa Canadiens in the Eastern Pro League in 1961 and the Omaha Knights in the Central Pro League in 1962. He became Montreal’s GM in 1964 and served until 1978. [16] He won in his first year as a GM as did Ken Holland.

Frank Selke Sr. received his training from Conn Smythe who led Toronto to 13 Cup appearances and seven Cup wins from 1927-1928 to 1953-1954. Selke won a Memorial Cup and worked with a semi-pro team before being hired as the assistant GM by Smythe in 1927. He worked with Smythe for 19 years during which Smythe lead the Leafs to nine Cup appearances and six Cup wins. Jason Farris suggests that the Leafs success was the result of the Leafs farm system, which Selke was responsible for building. Selke left the Leafs in 1946 after a falling out with Smythe and was immediately hired by Montreal. Under Selke Montreal became the premier team in the NHL because he “stocked the team with a deep farm system.” [17] It took him six years to win his first Cup while leading the Canadiens to another five Cup appearances in his first nine years. Then he let Dick Irvin (four-time Cup winning coach) go and hired Toe Blake as his coach. The Canadiens won five Cups in a row—the only team to do in the history of the league.

Lou Lamoriello took a very different route. He developed his skill set in the U.S. university hockey system. He played college hockey (1960 to 1963), served as an assistant coach immediately upon graduation and became the school’s head coach in 1968. In this role his record was 248-179-13. In 1983, he became his school’s athletic director and concurrently served as the commissioner of the Hockey East Conference from 1984 to 1987. He became the president of New Jersey in April of 1987 and appointed himself GM in September. He took over an organization that Wayne Gretzky had called a “Mickey Mouse operation” that was “ruining hockey.” Over the next several years he patiently built a winning team mostly through the draft. However, when one of his draft choices left to sign with Saint Louis as a free agent, he received Scott Stevens as compensation. Interestingly, Stevens was New Jersey’s captain from 1992-1993 to the early part of the 2003-2004 season. It was during this period that New Jersey won their three Stanley Cups.[18]

The average tenure of a GM with any one team is slightly more than five seasons. Selke took six years to win his first Cup and won a total of six while Lamoriello took eight years to win his first Cup and has won a total of three. On the other hand, Pollack and Holland won in their first seasons as GM. Pollack went on to win another eight cups and Holland has won another two. The implication seems to be that it takes more time for a GM who comes from outside the organization to do what it takes to build a Stanley Cup winner than it does for those developed within the organization.

The data presented suggest that it is very difficult for a “proven winner” to leave one organization and replicate that success in a different organization; i.e., moving from one NHL team as the GM to another NHL team as the GM. A “proven winner” in one organization develops a skill set that is specific to that organization. In fact, as Jay Barney suggests, some of these skills “… are so taken for granted, so much a part of the day-to-day experience of managers in a firm, that these managers are unaware of them.”[19] It is possible that these skills develop because of tacit characteristics that managers take for granted. Some of these characteristics are innate to an organization’s culture (Jason Farris’ book describes the culture in the California NHL team in the late 60s and early 70s – there is no doubt it was very different from Montreal’s culture during the same period). Examples are teamwork among top managers, relationships between suppliers and customers, and relationships among employees. Barney argues that an organization’s resources and capabilities may be such socially complex phenomena “… that although a particular leader may be able to transform the socially complex resources and capabilities in one firm, this same leader will be unable to transform the socially complex resources and capabilities in another firm.”[20] Winning a Stanley Cup is very difficult. Owners who hire GMs and GMs who hire coaches need to remember that spending money to get a “proven winner” from another team is very unlikely to buy their team the big one – the Stanley Cup.


[1] Shingler, Benjamin, 2011 (December 19), Habs strike a nerve in language debates, The Associated Press.

[2] McGran, Kevin, 2008 (February 9), Cup winners need not apply? The

[3] Ibid.

[4] Hornby, Lance, 2010 (November 30), Richard Peddie to retire from MLSE, Toronto Sun.

[5] Campbell, Ken, 2008 (February 11), Campbell’s cuts: Leafs must do everything they can to hire Holland as GM, The Hockey News.

[6] Statistically more difficult than winning multiple cups as the league grew from four teams in the 1917-1918 season to ten in the 1926-1927 season and back down to six at the start of the 1942-1943 season and from six teams in the 1942-1943 season to 26 at the start of the 1993-1994 season.

[7] Farris, Jason, 2011, Behind the moves: NHL general managers tell how winners are built, West Vancouver:

[8] The total tenure of these 170 individuals range from Jeff Gorton’s 11 games with Boston at the end of the 2005-2006 season to Glen Sather’s 2406 games with Edmonton and the New York Rangers from 1980-1981 to 2010-2011.

[9] Prior to the 1926/1927 season, the champions of the NHL and the Western Canada Hockey League had competed for the Cup. See for a more complete history of who controlled the Stanley Cup prior to the 1926/1927 season.

[10] Campbell noted that Lester Patrick had won a Stanley Cup as the GM of the Victoria Cougars (from the Pacific Coast Hockey League) before the NHL assumed control of the Stanley Cup. Adding this win to the denominator and treating it as a separate individual lowers the odds slightly. The odds become 52.6, 31.2 and 18.4 for 2+, 3+ and 4+ respectively. If we include Patrick into those who went to another team we get a 2 (Gorman and Patrick) out of 10 ratio or a 20.0 percent chance of winning with another team after winning with a previous team. This compares to winning approximately 4+ Stanley Cups with the same team.

[11] Only the records of the two teams (the Cup winning team and the subsequent team) were used for the analysis presented in Table Three.

[12] Cole, Jason, 2011 (October 21), Mission impossible: Duplicating Super Bowl success, Yahoo! Sports.

[13] Ibid.

[14] Farris, 2011, p. 39.

[15] Groysberg, Boris, Nanda, Ashish, Nohria, Nitin, 2004 (May), The risky business of hiring stars, Harvard Business Review, 1-10.

[16] Farris, 2011, p.212.

[17] Farris, 2011, p. 215.

[18] Farris, 2011, p. 203.

[19] Barney, Jay, 2011, Gaining and Sustaining Competitive Advantage (4th Ed.), Toronto: Prentice Hall, p. 131.

[20] Ibid., p. 133.