With the recent ungentlemanly behaviour on display in the NHL preseason, many people are putting their two cents into the debate on “goon” culture. People have wondered when it began, how it evolved, and how prevalent it has become. In this post, I’ll try to shed some light on these topics by analyzing the historical data available.
What is a “Goon”?
This is a good question, and really the primary thing we need to establish before we can collect data. Who the hell is a goon anyways? To me, there’s a key dichotomy to establish between who is a Goon and who is merely a tough NHL hockey player who happens to fight a lot. While some people would support the abolition of fighting entirely, I think we can all agree that the most destructive force in the game is the so-called “dancing bear”, or a player whose sole membership in the NHLPA is dependent on throwing punches instead of hockey skill. To me, the archetypal player for this phenomenon is Stu Grimson. Over a 14 year NHL career, Stu managed to score 39 points in 729 games. Over the same period, he took 2,113 minutes in penalties, a rate of 54 PIMs for every point he gained in the NHL. He only had 2 seasons where he scored more than 5 points (offensive explosions of 6 and 7 points). This was a player that had but one purpose: fight other heavyweights.
To define goons, I’ll concentrate solely on forwards. There have certainly been goonish players on defence, but generally they are tougher to identify and are more likely to have hockey skills, as it’s harder to hide one of 6 defencemen than one of 12 forwards on any given night. I’ll define any “goon” season as being one where a forward had 75 or more PIMs, while having 5 or less points. This seemed to be a good methodology to tease out the goons. Perusing the list of such players just sort of “feels” like you’re looking at a list of useless sacks of meat. To get to 75 PIMs, even as a goon, you needed to have played a decent amount to rack up enough fights, so were likely a semi-regular roster player. 5 points seems like a good cutoff to really distill this list down to the most useless of dancing bears. I’m sure some goons had seasons of 6, 7, 8, etc points, but this is just the methodology I used.
Where did Goons come from?
Perhaps the most interesting thing you notice when looking at the list of Goon seasons is that for such a long period in the game’s history, they did not exist. Prior to the 1980-81 season, there had only been 2 seasons that matched my criteria: one in ’58-59, and one in ’77-’78. The early 80s saw 7 such seasons, but in 1985-86 alone the game suddenly had 6 goon seasons. That’s really the genesis of real goon culture: 1985-86. The genie had come out of the bottle, and is still with us today. It would seem that prior to this point, lineup spots were deemed too important to give to players who could not actually play hockey. Fights surely occurred, but were more within the context of gameplay. Guys like Dave Schultz were routinely racking up 300+ PIM seasons in the 70s, but could actually play (Schultz had a 20-goal season the year before he had a 472 PIM season). You simply cannot find guys who were eating up playing time (and probably more importantly, budget) that could not also chip in offensively on the brawling teams of this period.
How many Goons have there been each season?
Here’s a chart of how many players matched my Goon criteria by season:
We see an interesting pattern — the number of Goons trended upwards steadily through the late 80s, 90s, and early 2000s, before trending sharply lower after the 04-05 lockout. The ‘8’ players you see in the 82-83 season actually adds up every player who had a goon season in or before that year. The criteria in the 2013 and 1995 seasons were adjusted to account for the lockout-shortened seasons.
The fever pitch was reached in 2003-2004, when there were 26 goons in the league, almost one for every team in the culmination of the Dead Puck Era. A lot of what made that period of hockey so distasteful was meant to be driven out after that lockout, and you can actually see some immediate results as the number of goons went from 26 to 8 in one year, a drop of almost 70%.
A fairer representation of the growth of Goons would account for the increasing number of teams via expansion that occurred during this period:
To create this chart, I’ve taken the raw number of Goons and divided them by the total number of forward roster spots in the NHL (number of NHL teams * 14 roster spots for forwards per team). This estimates what percentage of forwards in any given year were useless Goons. You can see that the deviations are a lot smoother between the late 80s and early 2000s, with pretty flat growth of the goon workforce (though 2004 was still the peak). Immediately after the lockout, you see the Goon percentage drop from over 6% to under 2% of players. This past season, Goons made up just under 3.5% of all NHL forwards.
Who employs the most Goons?
An interesting question then becomes, which teams are the guiltiest in terms of employing Goons? To accomplish this, I added up the number of goon seasons by team. In seasons where a goon was shared between teams, if the goon played more than 5 games for both teams, I awarded each team 0.5 of a goon season, whereas if the goon played almost all his games with one team and almost none with the other (less than 5 games), I gave a full goon season to the former team and none to the latter. I added franchises together who relocated from one town to another. Here’s the full list of goon seasons by team:
I came up with an adjusted goon seasons score for those teams who came late to the party through expansion, and normalized their goon seasons as if they had been in the NHL since 1985-86, the start of the Goon Age. Some of these teams may surprise you, and some may not at all. In the no surprise at all category, we have Philadelphia leading the entire NHL in goon seasons, with 22.5 in their history.
Interestingly, we have their cross-state rivals in Pittsburgh second overall, with 22.0. This is one example of a phenomenon I noticed throughout this chart: natural rival franchises tend to have similar numbers of goon seasons: PHI & PIT, NYR & NYI, EDM & CGY, TOR & MTL, OTT & TOR, for instance. This would suggest that goon employment may be (in some instances) localized arms races between rivals, whose cyclical number of goons tends to reflect the other’s in some perverted game of Mutually Assured Terrible Hockey (MATH).
We also have a team like Detroit near the bottom of the list, with only 8.5 goon seasons in their history. 4 of these goon seasons occurred in or before 1985-86, or before the beginning of the Goon Age. This means that since 1985-86, the Wings have only had 4.5 goon seasons. They’ve only had 2 goon seasons since 1988-89. Coincidentally, they’ve been pretty damned swell at winning hockey games since that time.
The Oilers are 24th in the league, with only 10 goon seasons since their inception. That may surprise some people, considering some of the famous fighters who’ve played for the franchise (Semenko, McSorley, Dave Brown, Laraque, etc). However, many of those fighters were also actual hockey players. Semenko only ever had one Goon season, his last playing for the Leafs in 87-88. He actually has a 27 point season on his resume. The season McSorley had 399 PIMs, he also had 41 points. Laraque never had a Goon season — his career PIM high was only 148, and the same season he had 29 points & 13 goals.
And that’s kind of the point of all this. It’s really quite ok to have tough guys on your team that can also play hockey. It’s not OK to employ people like Steve McIntyre, who can barely skate, turn, or stickhandle a puck. It’s really fun to go through the lists of teams like Philadelphia or Pittsburgh just to witness all the money they’ve wasted on fighters over the years. Something I noticed quite frequently while putting this together was how short Goon’s tenures on teams really are. It’s like these NHL teams can’t help themselves in repeating the same mistake over and over and over. Have a “need” for goon, employ goon, realize goon’s worth does not outstrip his cost, get rid of goon, then have another “need” for a goon, and repeat. Very few teams have resisted the temptation, for the most part.
How has the role of Goon changed?
Finally, I wanted to have a look at how the role of goons has changed over time. Here’s a chart that shows PIMs per Goon and Games Played per Goon by each season:
Firstly, we see a tentative start to the Goon Age in the late 80s, both in terms of how much teams were willing to play their goons, and how many PIMs those goons racked up. Then, there was an abrupt ramp up starting in about 1989 in the number of PIMs per goon, meaning they were fighting much more (almost double) than they were just a few years prior. In 1992, the NHL brought in the Instigator Rule, which resulted in a 2 minute penalty for “starting” a fight. There was an immediate recalibration in the goon equation: the next season, goons fought much less than they did the previous season, but oddly enough, they started playing in much more games. Prior to 1992, goons averaged about 40 games played or less, used in a specialty role. After the Instigator Rule, goons steadily started playing more games, but fighting less. The gap between decreasing PIMs taken and increasing games played has continued to widen. It’s almost like not being able to arbitrarily jump guys has forced goons to dress for more games, as some kind of faux deterrent. What this means for NHL fans is that we’re being increasingly subjected to a players filling a role that increasingly doesn’t exist — they’re looking for more fights that they’re not getting into. Beyond the ridiculousness of that equation, this also means we’re paying more and more as fans to watch fighters who don’t serve any functional purpose.