Monday, November 13, 2006

On Fighting (Part 1)

I’ve been trying to educate myself on the economy of fighting in hockey. When I first started watching, I found the fighting a bit counterintuitive. I mean, it’s a game when you’re going to get hurt a lot in the course of normal play- even the least aggressive check is still a pretty high-impact hit, and generally speaking, everybody prefers the more aggressive checking anyway. When you throw in all the completely accidental collisions and falling and whatnot, your chances of getting damaged are pretty high even if you’re not a particularly violent player. Now, by getting all brawly on your own time, you’re significantly increasing your probability of taking additional, totally unnecessary injuries, which might cause you to play poorly in future games, or miss them entirely, thereby making your team lose. This perhaps should be referred to henceforth as the ‘Ethan Moreau Principle’ (EMP).

So, if we view hockey as an economy in which wins (for the team) and points (for the player) are the primary currency, and furthermore we (as economists) assume that hockey players are rational beings who choose to fight or not based on an ad hoc cost-benefit analysis with the ultimate goal of winning or gaining points, than fighting is on the surface of it an irrational choice. According to the EMP, fighting not only doesn’t help your chances of winning or scoring, it in fact carries a significant risk of hurting them.

Yet hockey players fight a lot. This means one of three things: A) hockey players are irrational beings; B) hockey players are rational beings and we have mistaken their goal- i.e. they do not actually want wins or points, and are actually playing for some other reason; or C) hockey players are rational beings, and there are some hidden benefits to fighting, which come into effect under particular circumstances, and in those circumstances fighting actually becomes conducive to winning or scoring.

However, since we’re playing at being economists here, we cannot accept choice A, because there is no way to analyze economic behavior on the part of individuals without the general presumption of rationality. We accept that some individuals may not be rational in some instances, but irrationality is only acceptable as an explanation for the occasional outlying datum, not as an explanation for habitual behavior replicated by many people.

Choice B is more feasible, but still unlikely. It is possible that fighting itself is the goal of hockey, and the game is merely a structure set up to make fighting possible. But this would leave us attempting to explain the significant number of hockey players who do not fight, and it’s even more difficult to explain negative behavior (the not doing of things) than positive behavior (the doing of things). Moreover, the idea that winning and scoring are more desirable than losing and not-scoring is the foundation of the sport. If fighting itself were the purpose, why not just go into boxing?

This leaves only choice C: the hidden benefits of fighting. I’ve come across several possible candidates for this category:

  1. Justice- The natural violence of the game is subject to particular regulations. When these regulations are broken, a penalty is called and the offending player is required to go sit in a little glass box with a very stern-looking gentleman in a suit, thereby reducing the strength of his team. This balances the game: in theory the player that broke the rules was attempting to gain some sort of unfair advantage for his team, and therefore the officials try to restore the balance by giving a reciprocal advantage to the victim’s team. However, these laws are not always properly enforced- many penalty-worthy acts are not called as such. In such cases, the players of the aggrieved team rationally decide to try to restore the balance themselves by injuring the guilty party, and fighting is one way to do that. Since a ‘fair game’ is a prerequisite for one’s ability to score/win, fighting may in some cases be a means by which a team or player attempts to compensate for deficiencies in the practice of in-game justice.
  2. Vengeance- The logic is similar to the above, however, this principle applies in cases wherein a penalty is called on the offending player, but the offense was so disadvantageous to the opposing team that a power play is considered insufficient compensation, and fighting becomes a means to effect the additional punishment the team considers necessary.
  3. Psychological Advantage- Generally speaking, I doubt many hockey players are afraid of pain. Some are probably quite masochistic. However, most of them are human, and as such most of them have some kind of flight response to the threat of serious bodily harm or death. Starting fights, even for no reason, and especially if you have a few guys who are particularly talented fighters, could potentially encourage one’s opponents to play less aggressively, give you more space, etc. Even if this avoidance is only on a subconscious level, it’s still a potentially significant edge.

The problem with these 3 arguments is that they do not entirely negate the EMP- they add potential benefits, but the risks are not at all mitigated. The outcome of any given fight is never certain, and the chance that one will end up sustaining a significant injury for an insignificant benefit is high. Are there any possible explanations which reduce or obviate the risk end of this equation? How about this one:

  1. Fans- Fans love watching fights. If you don’t believe this, go to YouTube right now and type in the words ‘hockey fight’. Then do another search for ‘hockey goal’. Guess which one people would rather watch. As a player, you’re in a game which is constantly being reminded of its own comparative unpopularity vis a vis other sports. Moreover, you are also constantly being told that you would make more money if the sport was more popular. Hence, there’s a very strong incentive for some players to act in the way most pleasing to fans. A large, enthusiastic, and engaged audience is in itself a psychological advantage. If fighting engages the crowd more than actual play, and crowd engagement is an additional currency in addition to winning/scoring, than the EMP ceases to apply- the risk of playing less or playing worse becomes less costly, while the potential benefits of the fight increase dramatically. Your play in fact loses value, and being injured even acquires a certain value, insofar as it makes you more entertaining.

So this concludes my preliminary discussion of the economy of hockey fighting. Coming some day in the distant future: the psychology of hockey fighting, wherein I rip off the economist mask, toss that whole rationality assumption out the nearest window, and argue that it’s all about displaced rage and sexual frustration.

[P.S. In between when I began writing this and now, a couple of interesting quantitative studies of the value of violence (although not precisely fighting per se) in hockey have come out, which seem to point to opposite conclusions. On the Forecheck did a study of the correlation between hits given/taken and winning. While not about fighting, it's still phenomenally interesting. Turns out that violence in general may not be such a necessary part of playing well (wherein, again, playing well is defined by winning) as is generally assumed. However, only today Sisu Hockey had a (way over my head) discussion of another study which found that fighting, or more precisely the taking of major penalties, has a positive correlation with winning. I can't make much sense of it, except that I find the explanation offered by the study's authors deeply unsatisfactory- they say something to the effect that fighting 'rallies the team', but that seems to me to be a random supposition on their part, and I hate it when people try to transform correlation into causation via random suppositions. Anyway, if dude over at Sisu is right, their various data parsing techniques might be off, and therefore the correlation itself suspect. Happy reading!]

1 comment:

CheGordito said...

Insightful post. I would, however, challenge your basic principle about player objectives.

Instead of considering points the player objective, consider the fact that a minimum-earning NHL player makes far more than he would in a lower league. Points are a proxy for bargaining power for the salary. This redefines player behaviour - make myself as useful to the team as possible, through playing and fighting, but don't get hurt enough to ruin my NHL career and salary. If I don't earn enough points through, maybe I can fight and show team spirit.

I also wonder how many serious injuries are sustained in fights versus dangerous high-speed actions during the game. Perhaps the fighters aren't bothered by the 'small' losses (a tooth) and are not risking their 'real' assets (eyes, body parts put under most strain during play). Another fact is that not all players fight - my perception is that the more skilled players try to stay out of fights (for more time on ice as well as to not risk their 'assets').

Finally, I think the psychological advantage factor may increase over repeated interactions. I think that fights are probably more common among teams in the same division, where the repeated interactions make each individual advantage worth more and also allow players to get under each others' skins.

If you think of fighting as a game-within-a-game (different skills sets at play than during the game) then the fights give out a clearer signal - you may be beating us on points this time, but we'll 'hurt' you anyways, and we will meet again.

Just a few ideas to throw out. I've recently discovered your blog and am enjoying reading it.