« NotSneaky on remittances | Main | Time travel back to 1000 A.D.: Survival tips »
The incentives for violence in hockey
Here is my source. Here is the paper. Here is the abstract:
The level of violence in the National Hockey League (NHL) reached its highest point in 1987 and has reduced somewhat since then, although to levels much larger than before the first team expansions in 1967. Using publicly available information from several databases 1996–2007, the incentives for violence in North American ice hockey are analyzed. We examine the role of penalty minutes and more specifically, fighting, during the regular season in determining wages for professional hockey players and team-level success indicators. There are substantial returns paid not only to goal scoring skills but also to fighting ability, helping teams move higher in the playoffs and showing up as positive wage premia for otherwise observed low-skill wing players. These estimated per-fight premia, depending on fight success ($10,000 to $18,000), are even higher than those for an additional point made. By introducing a “fight fine” of twice the maximum potential gain ($36,000) and adding this amount to salaries paid for the team salary cap (fines would be 6.7% of the team salary cap or the average wage of 2 players), then all involved would have either little or no incentives to allow fighting to continue.
In other words, there are substantial incentives for violence in hockey.
Posted by Tyler Cowen on June 6, 2008 at 07:14 AM in Sports | Permalink
Comments
There is very little genuine violence in hockey fights. After all, both participants are on skates, on ice: they grab on to each other to try to maintain balance and awkwardly try to flail away with one free fist. It's impossible to put any significant body weight behind a punch, most of which miss anyway, or merely land a glancing blow. Sooner rather than later, one participant slips and falls and then the referee and linesmen move in to break it up.
Once upon a time, bench-clearing brawls were a genuine problem, but the league dealt with that quite effectively (by imposing harsher game penalties on anyone piling into a fight, rather than financial penalties).
Bottom line: this idea sounds a solution in search of a problem.
Posted by: at Jun 6, 2008 7:40:20 AM
...er, sounds like...
Posted by: at Jun 6, 2008 7:43:16 AM
Why would the NHL want to reduce violence? I suspect it helps ratings.
Posted by: FXKLM at Jun 6, 2008 8:44:35 AM
In response to FXKLM, the NHL has attempted to reduce violence in recent years by implementing a penalty for "instigating" a fight.
However, the instigator penalty has been criticized (by those such as Don Cherry) since it may lead to more "dirty" plays. A player - knowing that a player on the other team is unable to retaliate (otherwise they will be penalized) - may go after top players on the opposing teams. For instance, when the instigator penalty did not exist, it is argued that opposing players did not go after players like Wayne Gretzky since Gretzky's teammates would be able to stick up for him. The same cannot be said today: if a player goes after someone like Sidney Crosby in the playoffs, Crosby's teammates have a strong incentive NOT to retaliate (and this is magnified in the playoffs) otherwise they will be penalized.
I think the "externalities" of the elimination of the instigator penalty would be a very interesting question for future research in hockey.
Posted by: Samir Nurmohamed at Jun 6, 2008 8:56:33 AM
I just skimmed the paper, but I noticed what seems to be one big problem—they seem to have only counted fights that were penalized as fighting, which is a minority of all of the encounters that a non-hockey person would consider a "fight." Five-minute major penalties for fighting are generally a "ticking off the referee" penalty. Most participants in hockey fights are penalized with two-minute minor penalties for roughing, unless they tick off the referee by refusing to stop the fight once the refs try to break them up or something similar. If the NHL really wanted to get rid of fighting, they could tell the refs to begin penalizing all fights as fights and not as roughing.
But they don't want to get rid of fights, and not just because of what the first poster said about nobody actually getting hurt in hockey fights (which is very true), and not because the league thinks that fighting brings in fans (which may or may not be true). It's because a lot of people believe that hockey fights are much safer than the alternative, which is an increased number of other, more dangerous penalties, like slew-footing ("accidentally" knocking another player down by sweeping their feet out from under them from behind, which at best ends in the player landing heavily flat on their back and at worst ends with them landing heavily on the back of their neck or head). I've been told that there's little to no fighting in European hockey because of official crackdowns, but that there are a whole lot more penalties like that. I don't know if this is true (I don't watch much European hockey), but it's another thing that I'd like to see an actual study on before the NHL tries to get rid of fighting.
Posted by: AnonymousLibrarian at Jun 6, 2008 9:04:31 AM
I think the data may be distorted. I remember hockey being very violent in the sixties and seventies. However, the rules that increased penalties for coming off the bench to fight may have created some premium for players who can fight one on one.
The goon part of hockey was often a result of the fact that many of the skilled players tended to me rather small. Hockey players tend to be more average in size then most professional athletes. To keep these skilled players from being physically intimidated teams employed physical players or enforcers.
Montreal had John Ferguson, Boston had Teddy Green, and the list goes on. The officials can not create rules to formally protect smaller players but the teams use this system to enforce some codes of conduct.
I assume that your salary data reflects that one enforcer can protect numerous skill players. Teams may be willing to pay a premium for an enforcer who is effective at keeping a set of skilled players from being intimidated. Owners have a larger selection of skilled players to select from if they don't have to worry about the physical intimidation of smaller players. (BTW the French Canadian players that often dominated with the Montreal teams tended to be of average size.) The fact that some fans enjoy the fights just encourages the system.
Posted by: DanC at Jun 6, 2008 9:09:46 AM
Just echoing what others have said: fighting in hockey is a way of preventing worse forms of violence. Hockey fights are done with purpose and have certain norms that go with them. They are not usually uncontrolled brawls (see: baseball beanings). Hockey has a number of strong and tacit codes of conduct that players are expected to abide by and fighting often happens at very predictable times, frequently as controlled retaliation for dirty play by the other team or as a way to spark one's own team. And fighting is normally the province of a small number of players for whom it is part of their role in their team (e.g. Georges Laraque on Pittsburgh or Aaron Downey on my beloved and Stanley Cup Champion Red Wings).
The league should NOT be trying to reduce fighting and Don Cherry is largely correct about the unintended consequences of the instigator rule. Fighting was never about violence but about preventing people from doing real damage in a high-speed game with sticks and skate blades and lots of protection.
Plus, if you've ever seen a good fight at the appropriate time in an arena where they understand hockey, you'd understand why the league has no interest in reducing it. It's part of the game, not an aberration (again, see beanball brawls, or the Pistons/Pacers brawl).
The Stanley Cup is back where it belongs in Hockeytown!
Posted by: Steve Horwitz at Jun 6, 2008 9:22:52 AM
"Five-minute major penalties for fighting are generally a "ticking off the referee" penalty. Most participants in hockey fights are penalized with two-minute minor penalties for roughing, unless they tick off the referee by refusing to stop the fight once the refs try to break them up or something similar."
I don't agree with this. A fight in hockey is a highly ritualistic event which can't in any way be confused with roughing. Both players drop their gloves, move away from the rest of the game to some extent, grab each other and both try to land punches, while all of the other players stop playing the game, and the linesman don't break up the fight until the players are off their feet. Nobody stops playing on a roughing penalty until the offending team gains possession and the whistle blows. The player who commits the roughing penalty doesn't drop his gloves either.
"If the NHL really wanted to get rid of fighting, they could tell the refs to begin penalizing all fights as fights and not as roughing."
Unless there is an instigating penalty, which is pretty rare, fighting penalties are off-setting so neither team goes down a man. Roughing almost always leads to a power play for one side.
Posted by: josh at Jun 6, 2008 9:51:18 AM
The most violent acts in hockey have nothing to do with 'fights.' They come from hits at the knees, high-sticking, crosschecking, etc. when the victim is largely not expecting it.
Posted by: meter at Jun 6, 2008 10:00:02 AM
"I just skimmed the paper, but I noticed what seems to be one big problem—they seem to have only counted fights that were penalized as fighting, which is a minority of all of the encounters that a non-hockey person would consider a "fight." Five-minute major penalties for fighting are generally a "ticking off the referee" penalty. Most participants in hockey fights are penalized with two-minute minor penalties for roughing, unless they tick off the referee by refusing to stop the fight once the refs try to break them up or something similar. If the NHL really wanted to get rid of fighting, they could tell the refs to begin penalizing all fights as fights and not as roughing."
You are 100% incorrect. If anything what used to be considered a minor penalty for roughing is no considered a Major for fighting.
Posted by: Kostoglotov at Jun 6, 2008 10:10:46 AM
I'm sure the NHL wants give the appearance of trying to reduce violence and they might want to reduce violence to some extent, but I doubt the optimal level of violence from their perspective is zero.
Posted by: FXKLM at Jun 6, 2008 10:28:42 AM
It depends on whether you define fighting as "any time two players trade punches" or if you define it only as the classic two-players-squaring-off hockey fight. I think most laypeople would define anything that involves two people trading punches as a fight, and using that definition most fights(=punching) are penalized as roughing.
Posted by: AnonymousLibrarian at Jun 6, 2008 10:41:43 AM
AnonLibrarian:
I think you're wrong. In particular, I'm not sure that the majority of roughing penalties involve trading punches in any case. At best, it might be a push to the face with a glove that gets you the roughing penalty, as opposed to a genuine punch. Every once in a while, a cheap shot punch will get a player a roughing penalty, but no one would confuse that with a fight.
Posted by: Steve Horwitz at Jun 6, 2008 10:54:12 AM
There's a problem that I think is unique to hockey. It's the fact that there's no way to distinguish skill players and the physical enforcer players. In the other major collision sport, football, they have special rules that apply to QBs. This prevents teams from sending in that 3rd string meathead to go after the star player on the other team. There's no mechanism for that in hockey. You don't want to take out the physical part of the game, but you can't have guys taking cheap shots at the star skill players.
Posted by: Yogi at Jun 6, 2008 11:04:20 AM
To put into econ-speak, the Pareto optimal amount of fighting in the NHL is not zero.
Posted by: Steve R at Jun 6, 2008 11:17:43 AM
“It depends on whether you define fighting as "any time two players trade punches" or if you define it only as the classic two-players-squaring-off hockey fight. I think most laypeople would define anything that involves two people trading punches as a fight, and using that definition most fights(=punching) are penalized as roughing.”
You aren’t making any sense and are flat out wrong.
Are you confusing a face-wash (sticking a gloved hand in an opponents face) with a punch? Otherwise, there are only a handful of incidences, usually in a scrum where a gloved punch is thrown and a Major isn’t assessed.
Posted by: Kostoglotov at Jun 6, 2008 12:23:17 PM
To anonymous:
There is very little genuine violence in hockey fights. After all, both participants are on skates, on ice: they grab on to each other to try to maintain balance and awkwardly try to flail away with one free fist. It's impossible to put any significant body weight behind a punch, most of which miss anyway, or merely land a glancing blow.
Response:
Interesting. I play hockey, and I assure you that people who know how to skate can put significant body weight behind punchs.
After all, if they couldn't do that, how on earth could you put significant body weight behind a slapshot or a bodycheck? This accounts for the occasional KO, bloody nose, etc.
The real impediment to punching is all the equipment, and the fact that most hockey fights are in-close, grab-on, and wrestle with the opponent rather than trading blows toe-to-toe. It's more MMA than boxing.
Posted by: John Koetsier at Jun 6, 2008 3:30:47 PM
Admittedly I haven't read the paper - although it sounds interesting - I smell some bad econometrics!
If there is in fact a fighting premium for willing players - and assuming, of course, that players rationally respond to incentives - then shouldn't we observe an *increase* in the number of fights over time? Given that the number of fights has actually declined, I think something else is going here.
Perhaps the fighting premium is picking up an unobservable characteristic of fighting players. My hunch is that players who fight earn more not *because* they fight - but I concede that this is plausible at the margin. Rather, they earn more because they are of a particular type. In other words, fighting players have characteristics that non-fighters don't. They are self-selected workhorses and are willing to sacrifice themselves for the team.
Bottom line: this research seems questionable...
Posted by: Econometron at Jun 6, 2008 4:20:24 PM
Let me just point out (for those who are interested), there are similar studies regarding violence in NASCAR (i.e. the number of crashes).
Also, it seems like everyday there are "bench clearing brawls" in baseball, which always perplexes me since it seems to have very little relevance to the sport (unlike contact sports such as hockey, etc.). And of course, there seems to be very little hoopla surrounding it...when fighting happens in basketball though, it's an entirely different story.
Posted by: anonymous at Jun 6, 2008 4:45:27 PM
Beach clearing brawl irelevant? does anyone remember Marichal runnnig around the field with a bat after the pitcher, Roseboro?.Base ball player fight even against their own team. Bradley and Black.Or Zambrano and his own catcher .If a player does not enter the baseball field during a brawl will be fined by the team.
Posted by: karl at Jun 6, 2008 6:03:33 PM
John Koetsier,
We're not really disagreeing. I was indeed referring to the in-close clutching-each-other phase of hockey fights, which is what every hockey fight turns into after a few seconds. Under those circumstances, they're punching only with the power of the muscles of one arm and can't put much if any body weight into it.
PS, meter was right that the most violent acts in hockey have nothing whatsoever to do with ritualistic fighting. Some are even perfectly legal. A well-timed hip check, for instance...
Posted by: at Jun 6, 2008 6:05:19 PM
"There's a problem that I think is unique to hockey. It's the fact that there's no way to distinguish skill players and the physical enforcer players."
From these comments it seems like many of you don't watch much, if any, hockey.
It is *well known* who on a team is a physical enforcer. When Tie Domi stepped on the ice, the opposition had a pretty good idea what the Maple Leafs were up to.
Posted by: meter at Jun 7, 2008 9:14:03 AM
Meter,
I would argue that the difference between hockey and other major sports is that, rules wise, it is far easier in other sports to distinguish between types of players. I don't know precisely how a player in football is defined as "the quarterback" (although 'roughing the passer' penalties suggest that a player isn't considered the quarterback until making a pass attempt). Although the intent of putting a player like Tie Domi onto the ice is pretty obvious, there's no real way to distinguish his function in the game itself from that of another player.
Posted by: Will at Jun 7, 2008 10:20:59 AM
Meter,
I would argue that the difference between hockey and other major sports is that, rules wise, it is far easier in other sports to distinguish between types of players. I don't know precisely how a player in football is defined as "the quarterback" (although 'roughing the passer' penalties suggest that a player isn't considered the quarterback until making a pass attempt). Although the intent of putting a player like Tie Domi onto the ice is pretty obvious, there's no real way to distinguish his function in the game itself from that of another player.
Posted by: Will at Jun 7, 2008 10:21:38 AM
"We're not really disagreeing. I was indeed referring to the in-close clutching-each-other phase of hockey fights, which is what every hockey fight turns into after a few seconds. Under those circumstances, they're punching only with the power of the muscles of one arm and can't put much if any body weight into it.
PS, meter was right that the most violent acts in hockey have nothing whatsoever to do with ritualistic fighting. Some are even perfectly legal. A well-timed hip check, for instance..."
Youtube
Kimble Laforge
Twist Ray
Grimson Simpson
Kocur Kyte
Gilles Hodspdar
Wilson Hillworth
etc
And then comment
Posted by: Kostoglotov at Jun 8, 2008 3:16:26 AM
Yeah, whatever. Show me a fight that resulted in a Todd Bertuzzi-like outcome and then we can talk parity between squaring off and hitting someone from behind at more than standstill velocity.
Posted by: meter at Jun 8, 2008 11:38:25 AM
“Yeah, whatever. Show me a fight that resulted in a Todd Bertuzzi-like outcome and then we can talk parity between squaring off and hitting someone from behind at more than standstill velocity.”
That wasn’t your initial claim which was
“We're not really disagreeing. I was indeed referring to the in-close clutching-each-other phase of hockey fights, which is what every hockey fight turns into after a few seconds. Under those circumstances, they're punching only with the power of the muscles of one arm and can't put much if any body weight into it.”
Again look at those fights and then talk about how much power can be generated. As for damage inflicted – In the Bertuzzi/Moore incident was from Moore’s head hitting the ice, not the punch.
As far as damage – Twist broke Ray’s orbital bone. Boogaard broke Fedoruk’s jaw, Playfair broke Ruskowski’s jaw, etc, etc, etc
Posted by: Kostoglotov at Jun 9, 2008 11:06:21 AM
Sorry, that quote is from someone else. I thought you were disputing the bit after that which is attributable to me.
I grew up rooting for and watching the Broad Street Bullies, so I'm aware what a real fight looks and sounds like.
Still, I'd rather be hit at close range with a fist (from anyone) than hit in the head/knees with a hockey stick at 30+ mph.
Posted by: meter at Jun 9, 2008 12:26:25 PM
Phil Birnbaum has an interesting look at this article at
http://sabermetricresearch.blogspot.com/2008/06/hockey-fighting-study.html
Posted by: Cyril Morong at Jun 11, 2008 5:16:22 PM
