Hockey Players' Career Lengths

Watching the 2001-02 Stanley Cup finals, I found myself somewhat surprised at how long some of the participants have played for. Ron Francis has been in the NHL since 1981-82 (in his rookie season, he was a teammate of Dave Keon), Chris Chelios and Steve Yzerman have been playing since 1983-84, and several others have been NHL regulars since the mid-1980's. Also, several other players (such as Igor Larionov) had logged several years in Europe's elite leagues before joining the NHL.

This seems rather unusual to someone like me, who first learned about hockey in the 1980's. While it is nice, in a way, to see these players continue playing, it strikes me that, not all that long ago, people's careers weren't that long. In the 1980's, after Keon retired, I don't recall a lot of ancient players in the NHL. People like Gilbert Perrault, Guy Lafleur and Larry Robinson were very much an exception. Nowadays, Detroit has several ancients, and there are others scattered throughout the league. This raises two questions: First, is my perception correct, or am I just imagining that things were different in the 1980's than they are now, and second, why?

In an attempt to answer these questions, I'm going to look at one aspect of the big picture. Specifically, I'm going to look at every year during which major-league professional hockey was played (In this exercise, I use a definition of "major league" similar to that used by Klein and Reif in The Klein and Reif Hockey Compendium; i.e. it includes the NHL, WHA, NHA (1909-10 to 1916-17), ECHA (1908-09), PCHA (1911-12 to 1923-24), and WCHL/WHL (1921-22 to 1925-26).). Out of all players who made their major league debut in any given year, I'll find the last year during which at least one of those players was still playing major league hockey. In other words, the one aspect I'm looking at is the longest careers of each era. Before I reveal the results, keep the following points in mind:

So, here goes:

Year Last season of any skater who debuted that year Player (seasons)
1908-091923-24Joe Malone (15)
1909-101927-28Lester Patrick (15)
1910-111927-28Odie Cleghorn (17), Sprague Cleghorn (17)
1911-121926-27Ken Randall (14)
1912-131929-30Frank Nighbor (18)
1913-141925-26Leth Graham (8)
1914-151929-30Mickey MacKay (15)
1915-161931-32Buck Boucher (17)
1916-171932-33Reg Noble (17)

Apart from Leth Graham (who played a grand total of 27 NHL games over six seasons, if you can call them that) sneaking in, there are few surprises here. Ignoring Graham's "numbers", the mean and median number of years played of the eight other people is 16. Keep this figure in mind for future reference.

Also, keep in mind that, since I've defined major professional hockey to start with the ECHA of 1908-09, by definition no-one started their major league career before then. This does penalise people like Newsy Lalonde and Lester Patrick, whose careers started before this date.

Also, the last player to hang up their skates isn't always the player who played the most number of seasons. For example, Newsy Lalonde played 18 seasons between 1909-10 and 1926-27 but doesn't appear here because Lester Patrick played in a single playoff game in 1927-28. Similarly, Billy Bell played 10 seasons between 1913-14 and 1923-24.

1917-181926-27Jack Adams (10)
1918-191925-26Wilf Loughlin (8)
1919-201930-31Babe Dye (11)
1920-211930-31Frank Frederickson (11)

Looking at the above numbers, keep in mind that few people started their careers around this time, and the debuts of those who did may have been postponed due to World War I.

1921-221943-44Frank Boucher (18)
1922-231937-38Aurel Joliat (16)
1923-241936-37Frank Finnigan (14), Sylvio Mantha (14), Howie Morenz (14)
1924-251940-41Hooley Smith (17)
1925-261939-40Hec Kilrea (15), Nels Stewart (15)

Ah, now these numbers are rather similar to those from between 1908-09 and 1916-17. It seems that careers in the 14-to-18-year range are the maximum that a hockey player can expect, at least in this era.

1926-271938-39Paul Thompson (13)
1927-281946-47Dit Clapper (20)

The first player to play major league hockey for 20 seasons. Also, the first to play for 19.

1928-291944-45Mush March (17)
1929-301945-46Syd Howe (17)
1930-311940-41Johnny Sorrell (11), Paul Haynes (11)
1931-321945-46Earl Seibert (15), Ott Heller (15)
1932-331944-45Bill Thoms (13)
1933-341945-46Herb Cain (13), Lorne Carr (13)

In five out of the preceding six seasons, the last player to still be playing finishes out their career in 1944-45 or 1945-46. I'll discuss some possible reasons for this clumping below.

1934-351947-48Toe Blake (14), Nick Metz (12)
1935-361953-54Woody Dumart (16)
1936-371954-55Milt Schmidt (16)
1937-381950-51Bob Hamill (12)
1938-391953-54Sid Abel (14)
1939-401953-54Doug Bentley (13)
1940-411953-54Elmer Lach (14), Max Bentley (12)
1941-421958-59Kenny Mosdell (16)
1942-431959-60Rocket Richard (18)
1943-441955-56Don Raleigh (10)
1944-451964-65Ted Lindsay (17)
1945-461957-58Jim Thomson (13), Tony Leswick (12)

Nothing too unexpected here. While the difference between the starting year and ending year is quite similar to that of previous years, the total number of years played is smaller in several cases, as the careers of many of these players were interrupted by World War II.

1946-471979-80Gordie Howe (32)
1947-481968-69Doug Harvey (20)
1948-491968-69Allan Stanley (21)
1949-501973-74Tim Horton (24)
1950-511973-74Alex Delvecchio (24)
1951-521973-74Eric Nesterenko (22)
1952-531975-76Harry Howell (24)
1953-541974-75Doug Mohns (22)
1954-551975-76Larry Hillman (22)
1955-561977-78Johnny Bucyk (23)
1956-571977-78Frank Mahovlich (22)
1957-581979-80Bobby Hull (23), Carl Brewer (13)
1958-591979-80Stan Mikita (22)
1959-601978-79J.C. Tremblay (20)
1960-611981-82Dave Keon (22)

Now these numbers are quite different. In each of the above fifteen years, there is at least one person whose career started that year who played for at least twenty years (twenty-five skaters in total, some of whom aren't listed above). Previously, only Dit Clapper had played for twenty years, no-one else had reached 19, and only a small number of players played for 18 years. So, this is quite a departure from everything previous.

1961-621978-79Ed Westfall (18), Pit Martin (17)
1962-631980-81Terry Harper (19)
1963-641980-81Phil Esposito (18), Ron Ellis (16)
1964-651982-83Wayne Cashman (17)
1965-661981-82Don Marcotte (15)
1966-671982-83Carol Vadnais (17), Serge Savard (17)

Yep, those 15's and 16's and 17's and 18's look more familiar. Interestingly enough, even though someone who started their career in 1946-47 was still playing in 1979-80, namely - all together now, everybody - Gordie Howe, everyone who started their careers in the inappropriately-named "original six" era (1966-67 and before) had retired by the end of 1982-83. Actually, it's sort of unfair to classify Cashman, Marcotte, Savard and Vadnais as representatives of that era, as they played a total of 15 regular-season games before expansion (with Vadnais picking up 11 of that total). The last player remaining who played a full "original six" season was Dave Keon, who retired at the end of 1981-82.

1967-681982-83Walt McKechnie (16), Garry Unger (16)
1968-691984-85Brad Park (17)
1969-701984-85Butch Goring (16)
1970-711986-87Gilbert Perrault (17)
1971-721990-91Guy Lafleur (17)
1972-731991-92Larry Robinson (20)
1973-741994-95Mark Howe (22)

Hmmm, we've got a few people now who are playing for twenty years or more again... The numbers between 1967-68 and 1971-72 aren't too surprising, though.

1974-751990-91Harold Snepsts (17)
1975-761993-94Gordie Roberts (19), Brian Trottier (18)
1976-771992-93Randy Carlyle (17)
1977-781994-95Kent Nilsson (11)

For now, the survey ends here, as there's still an active player who started their career in 1978-79, namely Mark Messier, who as of 2003-04 has played in 26 major league seasons, giving him the second-longest major league career of all time. This era of hockey seems to be seeing a renaissance of the 20-year veteran. In the five years between 1997-98 and 2001-02 (when I originally wrote this article, in case you're wondering about the seemingly arbitrary date), 12 skaters (as well as one goaltender, John Vanbiesbrouck) logged 20 major league seasons. In alphabetical order, they are: Dave Andreychuk, Ray Bourque, Paul Coffey, Ron Francis, Mike Gartner, Wayne Gretzky, Phil Housley, Al MacInnis, Mark Messier, Larry Murphy, Scott Stevens, and Pat Verbeek. Since then, they have been joined by internationally acclaimed superstar Marc Bergevin, as well as Chris Chelios, Ken Daneyko, Doug Gilmour, James Patrick, Steve Thomas, and Steve Yzerman, with, no doubt, more to come to a hockey "centre" near you (not a lot of "arenas" around now, it seems) really soon. It certainly seems that I wasn't hallucinating, and that there really are more ancient players in the league now than there were 20 years ago.

Okay, so seeing all of those familiar names was refreshing, and I've answered the first question I posed above, but what about the other question? In other words, what does all this data mean? While it's probably not a good idea to extrapolate about all players who debuted in a given year based on the length of one player's career, the numbers above give a rough idea, at least, about the lengths of careers throughout the ages. Combine this rough idea with an analysis of the conditions of the day, and we can probably start to be able to draw some conclusions. Let's graph those years in the middle column (which represent, as I hope you remember, the last year in which a player who debuted in that year still played). This is represented by the red lines at the bottom. The green lines represent the number of major league teams, which is also somewhat interesting:


The interesting bits are the graph's clumps and gaps. The first clump occurs around 1926-27, after the Western Hockey League disbanded. With three fewer major league teams, the competition for jobs would be higher, so ancient players who would previously had a spot on a roster might find themselves out of a job. There's another clump around the start of the 1930's, when franchises started dropping out of the NHL.

The next clump is found around 1944-45. Even though the number of teams was static around that time, the talent pool had shrunk drastically, with most of the talented players in an army. So, we could assume that a lot of players' careers were lengthened by the lack of talent in the league at the time (at least those players who were still around), but by the end of the war these people started to retire.

There's a bit of a peak around 1953-54, but that seems just a coincidence. Moving right along, there's an incredibly large clump in the ten years between 1973-74 and 1982-83. Between 1975-76 and 1979-80, there's a steady decline in (mostly WHA) teams, which might explain why the ancient players started retiring, but how did their careers get so long in the first place?

Certainly the huge increase in major league teams (from 6 in 1966-67 to 32 in 1974-75) caused the demand for hockey players to greatly outpace the supply, and so inflated the careers of several of the people on the list above. However, this isn't a complete explanation. By the end of 1966-67, before expansion, some players (Gordie Howe, Bill Gadsby, Red Kelly) who had already logged 20 seasons, and there were others who were close.

Looking at the list of players above, none of them played their final season between 1960-61 and 1963-64, only Ted Lindsay played his final season in 1964-65 (although that was a bit of an exception, as Lindsay hadn't played the previous four years), and no-one else between 1965-66 and 1967-68. Any sort of balance between age and youth that existed prior to the 1960's was lost. Veterans with 10, 15, even 20 years under their belts became the norm, and the number of rookies declined drastically.

Just to show you that this really was the case, and I'm not just making this up to justify the numbers in the previous graph, here's another graph. This time the red lines represent the number of people who made their NHL debut in each year (note that, for example, 47 represents 1946-47 here). The green lines represent how many of those players played a full season:


Of course, the green bar isn't exactly equal to the number of rookies in each season, as it doesn't include people who only played a small number of games in a previous season.

There's a significant difference between the left-hand side of the graph (1946-47 to 1956-57) and the right-hand side of the graph (1957-58 to 1966-67). The average number of players who played a full season in their first NHL season in the first 11 years is about 7.8, whereas for the last 10 years it's only 4.4. Why is this? Not because there were no talented younger players (as can be shown by the number of talented "rookies" in 1967-68 who were in their late 20's or even in their 30's). This graph is rather hard to explain by taking an objective look at the numbers, but a subjective look might help.

Probably the best explanation is that, by the late 1950's, team owners figured out that, in order to attract fans to their arenas, it was a much better idea to have a lineup full of veterans than one full of rookies. Who'd want to pay money to see a bunch of people they didn't know? The reason why people knew who all of the players were wasn't because they didn't wear helmets, it's because they were the same players that they'd seen for the past 5, 10, even 15 or more years.

What about after 1966-67? In one word, expansion. Expansion increased the demand for jobs, and prolonged the careers of many of the people listed above. In 1972-73 the WHA came along, and with 28 major league teams, many veterans who might otherwise have been forced to retire managed to retain a job somewhere. With the demand for anyone who could lace a pair of skates far exceeding the supply, many were able to prolong their careers to hitherto unimaginable lengths.

After the demise of the WHA after the 1978-79 season, major league hockey settled down to a more manageable 21 teams, and after a while the longest career lengths too shrank down to what they had been before.

That was a refreshing history lesson. Let's apply this to what's going on today. In the 1990's, the NHL expanded five times, bringing the total number of teams in the league to 30. Once again, the demand for hockey players now far exceeds the supply, and there is still demand for a 20-year veteran.

To summarize: When you see people who have been in the NHL who have been there for 20 years, it means one of two things. Either the pool of available talent is no longer able to support the number of teams around, or that the owners now value veterans highly in order to draw spectators in. The career of Dit Clapper, who as I mentioned above was the first player to play 20 seasons of major-league hockey, was prolonged by World War II, when young hockey players who would otherwise have taken his place were in the armed forces. The careers of a lot of people who debuted in the late 1940's and 1950's were prolonged by two factors: First the owners started to place a premium on age, and expansion and the WHA combined to exhaust the talent pool and keep grizzled veterans employed. Nowadays, the NHL has more teams than the NHL and WHA combined did during most of the 1970's (1974-75 and 1975-76 being the exceptions), and once again the demand for hockey players surpasses the supply, so we once again are seeing the return of players who have played for 20 or more seasons.

I'll close this off with a diversion: What player has played against the most major league hockey teams? Well, since there were a total of 34 different franchises between the WHA and the NHL in total in the mid-1970's, there are probably a fair number of people who have played against about 34 franchises. Could anyone have played against more? The only way they could do that is if they played in the modern-day expansion era (or if they played in the 1930's, but that's obviously not a reasonable possibility).

Looking at the table above, a reasonable candidate who played in both the WHA and NHL is Mark Howe. Looking at this career, he would have been able to play against all 16 different WHA franchises, the 17 NHL franchises that weren't previously part of the WHA, and 5 of the franchises the joined the NHL ranks in the 1990's (while I'm not 100% sure that Howe played against Florida, he did play 44 games in 1993-94 so it's probably not a bad bet that one of them was against Florida). That's 38 franchises in total. I can't think of anyone else who would have played against more. Neat, eh?

Actually, I'm going to close this article off with yet another diversion. I got a question a little while ago about how long the average career length is for a defenceman. This is a somewhat tricky question, which is why I avoided it completely in the article above. I thought I'd take a look at it here though. I decided to look at the 1999-00 season, which is recent enough that the results I get have a hope of being relevant today, but long enough ago that it's likely (but certainly not certain) that people who haven't played in the NHL since then won't be playing in the NHL in the future either. Anyway, in 1999-00 there were 32 defencemen who played in the NHL that year but haven't played in the NHL since. These 32 defencemen spent the following number of seasons in the NHL:
Number of seasonsNumber of players

These data have a mean of 6.1875 seasons, a median of 5 seasons, and a mode of 1 season (which is sort of revealing). So, it probably makes sense to say that the average NHL career length of a defenceman is about 5 to 6 years. One final thing I should say about this diversion is that I've counted seasons, not full seasons. So, for example, if someone played one game in the NHL, I've counted that as a full season. If I had only counted "full" seasons (whatever that means), quite a lot of people would have ended up playing zero seasons.

Well, enough diversions (for now, anyway). On to the footers...



January 21, 2004
Updated the list of players who had played 20 or more seasons, and added the new diversion at the end of the article.
March 20, 2004
Added Johnny Gottselig to 1928-29 player list.

For questions or comments, please e-mail James Yolkowski (

This article is Copyright © 2002-2004, James Yolkowski. You may reprint or reproduce this article, as long as this paragraph is also reproduced. The original article and others like it can be found at