The League of Extraordinary Statisticians: The Goon

Even if you've already read, please take a quick look at Geoff Detweiler's response, which I've just added.  I dropped the ball and lost his response in my email.  I blame booze.

The League of Extraordinary Statisticians (LOES) is a weekly forum bringing together the top analytical minds in the hockey world to answer a variety of questions that straddle the line between stats analysis and something you might hear floating around section 304.  They have agreed to answer these questions in a few paragraphs or less, and with minimal formulae.  Because this is a forum, we'd encourage you to use the comments section to answer the questions yourselves, or to discuss or debate the answers given.

The LOES is not meant to represent the entire of the hockey stats community.  There are a number of people that either were too busy or too difficult to contact for the purposes of the forum.

The "goon" is an interesting role on teams that have them; essentially, they are given about 5 minutes or less per game to either make a few big hits, intimidate the opposing team just by being on the ice, or to engage the opposing team's goon.  Presumably, there are extraneous benefits to teams that have a goon, such as entertainment for fans, though the jury is still out on whether they truly draw more fans.

The debate over a goon's value is well-worn among hockey fans, are pretty evenly divided among those who dislike them entirely, those who love them and believe them to have value, and those who might not feel they have value yet can't help but enjoy watching the fights.

This week, the LOES is going to put the goon to the test (not a real test, only George Parros would pass)... 

The question this go-around is Goons: necessary/positive component, zero-sum gain, or detriment to the team? Please explain.

Personally I think goons are a detriment.  They can't play a regular shift, making your other players more fatigued in the long run.  They contribute nothing, and if they were so good at protecting the stars, how come the stars are constantly getting run anyway?

- Rob Vollman, Hockey Prospectus

Another rule change related to this is that the goon has to be able to play hockey.  Is that too much to ask?  I'm good with the policing aspect, as long as he can actually also play hockey.  Does it make sense that you can have someone who plays 5 minutes a game for 60 games (300 minutes on the ice) as well as spending 300 minutes in the box?  No, that's ridiculous.  You don't see goons in the playoffs, and that's because there are better incentives in the playoffs (you can't afford to lose a game).

I'd prefer the tough-guy, who can play 10-12 minutes a game (say about 700 minutes of ice time), and get 100-150 minutes in penalties.  You still get the physical play, a police-type presence, but with guys who can actually play hockey.  So, you can easily build in a rule based on PIM or number of fights based on a player's ice time.  Say for example, that you need at least 60 minutes of ice time for each fight.  A player playing 10 minutes a game for 60 games would be able to get into up to 10 fights.  If he plays 6 minutes a game for 40 games, then he'd only be allowed to get into 4 fights.

If you don't like the 60 minutes per fight, change it to 40 or 30, or whatever you think is appropriate.  But there has to be some incentive for the player to actually be an NHL-calibre hockey player.

- Tom Tango, tangotiger.net and author of The Book: Playing the Percentages in Baseball

I know it's not the main question, but I believe that it's baked in there...If NHL teams (coldly, coolly, logically) determined that fighting wasn't a necessary or beneficial part of the game, would it be okay to phase out (or ban) fighting altogether? What would the ramifications be? 1. Banning fighting in the NHL would inevitably lead to banning fighting in all North American leagues. 2. The National Hockey League would maintain it's popularity in Canada and in America without fighting due to the high quality of the league. 3. Juniors would maintain their popularity in Canada without fighting because Canadians support the sport at the junior level. 4. But the league that would be significantly hurt is the American Hockey League, the main minor league to the NHL. The vast majority of AHL teams are in America, and fighting is a very significant part of the draw--especially considering the AHL's half-decent level of hockey--for the more marginal fans here in America (I've always believed that the AHL's motto should be: "At least one fight per game, or your money back!"). 5. So the impact of phasing out or banning fighting wouldn't directly impact to the NHL, but could indirectly be crippling to its main minor league, and could consequently really harm the NHL as well. So if you conclude that fighting is bad for NHL teams...please don't tell!

- Timo Seppa, Hockey Prospectus

A no-talent goon who plays maybe 5 minutes a game against weak talent and rarely if ever scores is a detriment to a team.  A more talented player who can play a regular shift without hurting his team or even better helping his team who is willing to fight is usually quite valuable.

 
Fighting in itself rarely turns the tide of a game in practise, but it is a commonly held idea that people are looking for.  If a fight occurs and then a team wins a game, it is often written in the media as though the fight changed the momentum in a significant way, but rarely is that explanation valid.
 
A tough player who hits, wins battles in the corners and is willing to fight and stick up for his teammates is valuable, but not for the fighting so much as for overall toughness.  You don't want your team to be pushed around.  It is hard to win in those circumstances.  Fighting is mostly a spectacle that that does not influence the result of the game.

- Greg Ballentine, The Puck Stops Here at Kukla's Korner

They are a detriment to the team, full stop.  I've looked at this before and found the average value of a goon is one loss per year.  I need to follow that up with a WOWY* between goons and fourth-liners to see if the goons have a measurable negative impact on their linemates and look at the opportunity cost of soaking up cap space with a goon, but I don't imagine either of those studies will look well upon goons.  My guess is the combination of those three things will show the average goon is worth about two losses per year.

The narrative in support of goons always involves something about protecting star players from injury and acting as a deterrent if other goons and Averys were to think about running star players.  It sounds like a logical argument until you realize it doesn't hold up to scrutiny.  Goons are never on the ice at the same time as star players and Averys can never answer for dirty plays.

I have no problem with fighters who have a bit of game -- Theo Peckham and Milan Lucic come to mind.  If protection or deterrence is part of a number of skills, the team is better off for it.  Even guys that can serve as thirteenth forwards or seventh defensemen but have some other serviceable skills like Deryk Engelland or Dan Carcillo are okay by me - the roster spots they take and the cap dollars they use aren't completely wasted.

- Derek Zona, The Copper & Blue

Ah, goons. What exactly is a "goon" though? As a fan of a team who employs nothing but goons, it gets a little confusing. Many people would probably call Scott Hartnell or Chris Pronger or even Matt Cooke goons, but I don't think they are. To me, a goon is Jody Shelley, Colton Orr, Derek Boogard, or Cam Janssen. Because that definition of a "goon" is largely a player who is simply there to fight and won't see more than 6 minutes of ice time a game; they're at best a zero-sum gain. They aren't on the ice to score or even play defense (like Matt Cooke does very well), so they don't contribute much - if at all - on the ice.

But there's something to be said about how these guys keep finding jobs. They're almost all respected around the league, talked of as excellent locker-room types, and sometimes even make space for their star players. None of this can be quantified, obviously, but so long as you have a goon who makes little money (unlike Jody Shelley or Derek Boogard), I can understand teams wanting them around. The way Washington is using D.J. King is probably a best-case scenario for goons: only pay him $637,500, only dress him in 8 games, and give him less than 6 minutes of ice time. At best, the Caps aren't hurting themselves. But Boogard and Shelley? Don't know how any team justifies their multi-year, multi-million dollar, contracts as anything but hurting the team.

- Geoff Detweiler, Broad Street Hockey

As you can see, the goon is not held in high regard by the LOES, and for good reasons: they are limited in their abilities, and their deployment can hurt the other lines.  Tango gives us a good idea to amend the existing stance on fighting without eliminating it entirely, and I think it's a good point to make that should fighting rules be changed, it's not likely that it would result in a complete ban.

That being said, there's a few discussion points that we need to think about: if the NHL's deployment of goons isn't reflective of the LOES's assertion that they are not particularly valuable (and possibly harmful), what are the existing rationale for their use?  How might this rationale be effectively and critically analyzed?  Is the risk to minor leagues such as the AHL great enough to be cause for concern (per Timo Seppa's response)?

* Editor's note: WOWY stands for "With or Without You", a way of measuring player's performance comparatively from their performance when varying teammates are with them on the ice.  Here's a recent example.