Friday, August 08, 2008

Ditch the enforcer


I've decided to take another shot at the proverbial goon since there seems to be a lot of them hanging around (and Darryl Sutter can't bear to go without one). Let me first state that I have been a hockey fan most of my life and enjoy the visceral thrill of one of *my* players beating up one of the *bad guys* as much as anyone else. This is not a call to ban fighting from a squeamish bleeding heart. However, as a fan who enjoys winning even more than fisticuffs, and as an amateur analyst, I feel compelled to question the validity of the pugilist.

Certainly the number of lumbering heavyweights has been reduced since the lock-out, in response to such goonish impediments as draconian instigator rules, the salary cap and the obstruction crack-down. In 03/04, 8 of the top 10 PIM getters were guys I'd define as "enforcers" - players that are in the league primarily or solely because they can chuck knuckles (low point totals, small ice-time, triple digit PIMs) - with Sean Avery and Chris Neil (who is basically a relatively high functioning goon) the exceptions. In 07/08, perhaps 3 of the top 10 could be considered "enforcers" (Stortini, Parros and Burish), although only Parros averaged the typical goonish ice-time, with Storts and Burish inching closer to the functional 10 minutes/night crowd (meaning they can actually play some hockey too). It's four if you count 200 pound Jared Boll, whom I tend to classify as a "pest" more than "goon".

Yes, the enforcer is a dying breed. And in my humble opinion, it's a breed that isn't dying fast enough. GM's still cling to the conventional wisdom that "every club needs its heavyweight" like jammy-fingered children to a security basket. Just glancing over the various rosters now, I count about 26 pure pugilists. Im not talking about guys that fight, or even fight a lot. Again, I mean the guys who do almost nothing but fight. The policemen, the nuclear the deterrents, the not-so-gentle giants. The guys whos penalty minutes match or exceed their total ice-time by the end of the year. There's still a bunch of them in the league. And I can't conceive of one good reason why that is.

First off, another hat-tip to James Mirtle who thoroughly investigated this issue previously. His findings and conclusion bear repeating:
The average enforcer in this study has played 16.5 games at 5.45 minutes per game for a total of about 90 minutes of ice time this season. In that time, against the worst opposition in the league, their teams have scored an average of 1.56 goals, allowed 2.66, and been out shot 38-30.

At best, they're not a liability. At worst, they cripple their team, allowing somewhere in the neighbourhood of two goals per 60 minutes more than the rest of their team while generating almost zero offence or shots on goal.

If I'm a coach or GM, isn't there a better option at the bottom of the roster?

One would think so. However, let's proceed on the assumption that the contributions of the archetypal goon can't be captured with the advanced metrics Mirtle employed in his piece. Sabres heavyweight Andrew Peters offered these justifications when questioned about his own usefulness:
"You take fighting out of hockey, you might as well take the triple salchow out of figure skating. I saw someone hurt their tailbone pretty badly and miss a skating competition not too long ago. That's something the International Olympic Committee should be looking at.

"But in all honesty, it's offensive to me. It almost takes away from what you mean to your team. I don't think we are (useless) and I think if you asked 25 guys in this locker room, they'd tell you the same thing. Take fighting out and you're looking at a whole different style of hitting. Now guys will take free rein because they don't have to pay the consequences. I think it would be stupid.

"There's a reason for it. It can change the momentum of a game, send a message that if you come after our players, we'll come after yours. It can be used in so many different strategies, more than just a goon way. And that's such a terrible word to use."

1.) Im personally not arguing against fighting in hockey per se, so his lamentations to that effect are irrelevant. Players would still fight even if there wasn't a designated fighter on every team no doubt. Hell, let 'em.

2.) The "consequences" for malfeasance is a common refrain when I debate this issue with a goon supporter. The claim is that the opposition would suddenly start to hurt or injure the good guys (particularly the stars) without the presence of a heavyweight. This also falls in the "send a message" realm. I'll address this in more detail later.

3.) The "momentum" trope is also a classic one. Despite the fact I think "momentum" is a dubious contributor to success, I find it hard to swallow that a perfunctory clash between two lugheads (since they only ever seem to fight each other) can be of much motivational or psychological value. Iginla scoring a couple goals and then beating on some twerp that's mouthing him off? Maybe. A 30 second bear-hug between Ivanans and Stortini? I doubt it.

4.) As for enforcers "being used in so many different strategies"...well, Im afraid Im a skeptic. Peters names maybe two distinct "strategies" above (message sending and momentum changing) and neither of them are terribly compelling.

As to "message sending", the modern goon is so utterly impotent in the face of "new" NHL obstacles that his ability to right perceived wrongs and deter pests is next to nil. Here's a brief guide on how to neutralize an enforcer:

a.) Just say "no". With the strict instigator rules that include penalties, suspensions and fines, not to mention the existing codes of etiquette and conduct regarding beating on an unwilling partner, the ability of a goon "to do his job" is directly moderated by the oppositions willingness to engage him in fisticuffs. An obvious example that springs to mind is Eric Godard: from Dec.13 to Feb.2*, a 19 game sample, Godard didn't get a single penalty minute, let alone a fighting major. Godard explained in an interviewed during the dry spell that he was "frustrated" because he simply couldn't find anyone to fight him. He'd skate around for five minutes a night, looking for trouble, finding none. He was rendered completely toothless.

(*Notably, Godard's dry-spell ended in Edmonton on Feb.04. His 27 PIM came thanks to a crease scrum in which Godard clumsily attacked anyone and everyone around him (the Flames were being embarrassed at the time). Godard ended up getting thrown out of the contest and the Oilers ended up scoring on the ensuing powerplay(s). Take THAT Oiler scum!)

b.) Have an effective 4th line. Goons only play about 5 minutes a night because they tend to be gross liabilities. They get scored on easily and can't score themselves. They also take a lot of penalties. As such, goons only tend to see the ice when the outcome of the game is no longer in question (blow-outs, in either direction) or against similarly bad players. Therefore, a relatively decent bottom end of the roster would probably convince the opposing coach to bench his big man pretty quickly**.

**(That is, at least, until his team was being blown-out. Then the goon could be sent out to fight and be subsequently penalized, suspended and potentially fined...thereby further feeding the blow-out).

c.) Dont have a goon on the roster. An extension of the point above, but also plays into the "goons only fight each other" thing. How often do you hear "we have to play X heavyweight because they have Y heavyweight?" Well, maybe *they* wouldn't play Y if you didn't have X, capice?

Just say "no", have a half-way decent bottom 6 and don't bother dressing a goon. Voila: you have now effectively castrated the other guy's enforcer. Feel free to "run their stars", if it fits your fancy (because that's what everyone does when there's no goon, right!?), because there ain't a damn thing he can do about it.

---------------------------------------------------------------------

In his interview, Peters hints at a potential psychological calming effect a goon might have on the rest of the roster..."just ask 25 guys in the locker room, they'll tell you the same." Perhaps the enforcer is the gun under-the-pillow for the rest of the team? The insurance of civility, the last resort, the bruiser the guys can turn to if things get out of hand?

Perhaps. And if so, does that have some measurable effect on performance, ie; on winning? If it does, I would wager that the effect is more than neutralized by the fact that goons, as shown over and over again, are bad hockey players and have very real and very observable consequences when it comes to things that most obviously do have effect on winning (goals for, goals against, penalties for).

Just in case, I did a quick correlation-analysis between fighting and other team metrics from last year (wins, losses, points, GF, GA and GD):





Note - tallying fighting majors may be a pale proxy for the "goon effect", since it includes tussles by every team member. Still, teams with lots of fights can be reasonably assumed to have a "fight ready roster", including a goon, and not be susceptible to intimidation (ie: the supposed benefit of employing an enforcer).

As you can see, fighting doesn't really have a strong correlation to any of the other performance measures. wins, points and GD share a positive relationship with total scraps, but nothing even approaching significance. The strongest inverse relationship seems to be between goals against and fighting, oddly, but even that could be a product of chance.

---------------------------------------------------------------

Perhaps there are some good reasons for keeping a goon around. Afterall, it should be safe to assume that each NHL GM isn't just following the ass of the guy in front of him when it comes to these signings, right? RIGHT?

If there's a good argument in favor of enforcers, though, I haven't heard it. I'd welcome a worthwhile explanation because, as far as I can tell, the heavyweight is an anachronism in this league and carrying one doesn't do anything more than further an outmoded convention.