Wednesday, May 07, 2008

When the whistles go away
An embarrassing look at NHL officiating

Those that frequent this site know that, in general, I don't focus all that much on officiating. For the most part, that old axiom about hockey being difficult to referee holds true, and I think the league's latest attempt at a crackdown the past few years has been a generally positive thing.


As the following stats will show (and I owe another thank you to Gabe Desjardins for some help on this), the calls in this year's playoffs simply do not cut it. Too often referees are disappearing late in games, so much so that it's affecting the integrity of games.

So far in these playoffs, there have been 68 games played. In those games, there have been by my count a total of 766 penalties called, or 11.3 per game, a figure that includes everything: fights, misconducts, majors, etc.

Here's the minute totals, by period:
Period 1 = 580 PIM (8.53 per game)
Period 2 = 576 PIM (8.47 per game)
Period 3 = 571 PIM (8.40 per game)

Let's throw out fights and misconducts, the bulk of which occur in the third:
Period 1 = 550 (8.09 per game)
Period 2 = 556 (8.18 per game)
Period 3 = 381 (5.6 per game)

With 68 games played, 20 minutes a period, there have been 1,360 minutes each of first, second and third periods. That's .40 PIM per minute in the first period, .41 in the second and .28 in the third. (And I'm including coincidental penalties for now.)

Anyone want to guess what overtime looks like?

Fifteen games have gone to overtime these playoffs, nearly a quarter of them, and more than 164 minutes of extra time played.

There have been exactly four penalties, all minors, called in overtime this year, which is 0.049 PIM per minute. Three minutes in penalties over 60 minutes of action, in other words.

Now, playoff overtime's obviously an extreme, but the fact is, penalties have been disappearing as games go along. I've broken down all 68 games into 12 five-minute intervals, eliminated coincidentals, which gives us a look at the man advantages dealt at different points in games in these playoffs:

The first five minutes of a period are very penalty free, in general, but have a look at the third five-minute interval (10-15 in the first, 30-35 in the second and 50-55 in the third): Players are half as likely to receive a penalty in the final period as in the first two in this portion of games.

(And the small spike at the very end is attributable to some of the end of game shenanigans that always seem to take place.)

Minus misconducts, majors and coincidentals, games average about 8.5 penalties per game, roughly four power plays a side. First and second periods average between 9.2 and 9.5 penalties per 60 minutes, a figure that declines about 30 per cent in the third, and 35 per cent over the final 10 minutes of games.

From the rate in the first two periods, overtime penalty calling drops 85 per cent to 1.46 per 60 minutes.

Types of penalties by period (coincidentals included):

1st 2nd 3rd Total
Roughing 69 55 32 156
Hooking 46 45 21 112
Interference 31 30 19 80
Holding 22 29 16 67
Slashing 21 25 20 66
Tripping 25 21 19 65
High stick 11 16 20 47
Cross check 19 9 8 36
Interf. goalie 8 6 4 18
Unsportsmanlike 2 10 5 17
Puck over glass 3 8 5 16
Boarding 5 3 5 13
Elbow 2 6 3 11
Holding stick 4 3 3 10
Other 1 4 4 9
Diving 3 4 1 8
Charging 1 1 0 2
Total 273 275 185

There's no question that's a problem. The question is, how big of one is it and how do you fix it?

UPDATE Danny Tolensky had more on this topic last week.



At 6:59 a.m., May 07, 2008, Anonymous Ryan said...

I think slashing, tripping, and high-sticking are three of the most clear-cut penalties, and unsurprisingly, they drop the least in the third period. Refs are afraid to make the borderline calls in the third, so they overcompensate by calling too many in the first and second.

I think you do have to ref situationally, by the score and time left, but not to this extent. But who knows? Teams trailing after two periods have won 22% (12 of 55) of games this year, higher than any year since at least 2001. (The next highest was the penalty-filled 2005 playoffs, at 19%, 14 of 60, and the worst was 11% in 2002, 7 of 64.) But common sense tells me that teams will catch on to the pattern, and this trend of relatively penalty-free comebacks is too good to last.

At 7:04 a.m., May 07, 2008, Blogger E said...

three random thoughts:

1) it's always seemed to me that most hockey fans are intuitively aware that officiating isn't consistent over the course of a game, but that a lot of them approve of it anyway on the justification that there are some circumstances (particularly the third period of an important game) where it's best to 'just let 'em play'.

2) if officiating was really done to the letter of the rulebook, there'd be a lot more calls in every game. although there's been a crackdown on hooking and holding, apparently, there are still a lot of things (i'd definitely mention boarding and cross-checking) that are called way, way less often than they technically happen. given that a lot of people aren't thrilled with the frequency of current rates of hooking calls, would they really like to see an even more literalistic application of all the rules?

3) but if you're going to allow for subjectivity in calling (i.e. not cleaving zealously to the most precise possible application of the rulebook), then is it sufficient for the calls to be 'fair' in the sense of 'unbiased' rather than in the sense of 'consistent'? your data seems to point to a kind of consistent-inconsistency- if this kind of variation in calling depending on the point in a game is completely pervasive, does that make it paradoxically equitable? if officiating practices are uniform, do they have to be literal?

At 7:14 a.m., May 07, 2008, Anonymous Anonymous said...

Great comments.
I personally do not like to see a marginal penalty decide a hockey game, especially in the playoffs.
The late holding call on Datsyuk last year against the Ducks changed the whole series.

Maybe players sctually commit less penalties late in games and in overtime, focusing more on the score.

It stands to reason they would want to make a physical statement earlier, and focus on clean play late in the game.

At 7:15 a.m., May 07, 2008, Blogger sacamano said...

1) How have you controlled for the possibility that players are less inclined to take penalties in the 3rd with the game on the line, than they are in the first two periods?

2) It used to bother me that NHL reffing had "make up calls" and that they "let 'em play" at the end of games. But I'm increasingly coming to see it as simply part of the NHL's charm. As 'e' points out, most people who grew up watching the NHL intuitively understand how it works, and consistency (on a game by game, not necessarily period by period) is more important.

At 8:01 a.m., May 07, 2008, Anonymous Anonymous said...

i agree with the above comments about the possibility of players taking fewer penalties in the third and overtime when the game is on the line.

and i seem to see a worse trend where any time a player falls or gets hit cleanly the commentators or players complain for a penalty.

i like "just let them play"

At 8:22 a.m., May 07, 2008, Anonymous Anonymous said...

@ "E":

To your third point, i'd say if you asked the players, they'd tell you "I don't care what you call, as long as you call it consistently, and at both ends of the ice". There's nothing more frustrating as a player than to watch an opponent do something that is typically considered a penalty, not get punished, then do the exact same thing to one of their players and get the hook (see: Scott Hartnell and Brooks Laich, 2008 1st round) In my mind, that's the worst thing a ref can do because at that point, you have no idea what you are or are not allowed to do in the game and cannot make adjustments for it.

At 9:26 a.m., May 07, 2008, Anonymous Anonymous said...

You have to account for lopsided games in the third period. If a team has a 2-3 goal lead they dump and play a trap in the third. They're much less likely to take a penalty. Even the losing team in a game like that often just runs out the clock for 2-4 minutes.

At 9:38 a.m., May 07, 2008, Blogger Mr. Thursday said...

I don't think I really mind the "Let 'Em Play" mode of referring. In a close or tied game, in the 3rd period or overtime, goals come at a premium. With man advantages giving 20%+ chance of scoring, in a low scoring environment, the difference is far more significant that it is in the 1st and 2nd.

That's to say, 2 minutes for hooking in the 3rd/overtime impact the game a lot more than they do earlier.

I agree with Bill Clement's philosophy: I don't really mind if the refs are strict or lenient, as long as they make the same calls for both teams. I don't even really like seeing my team win as a result of a PP goal from a bad call. I appreciate the win, of course, but it feels dishonest.

At 9:44 a.m., May 07, 2008, Blogger Steve said...

How do these numbers compare to the regular season?

At 9:53 a.m., May 07, 2008, Blogger Doc Nagel said...

"Let 'em play" is repugnant to me. It reduces the sport to an uncontrolled melee, a testosterone-fueled festival of thuggery. "Let 'em play" doesn't let anyone play, at least, it doesn't let anyone play the game of hockey. Refereeing is an ongoing and evident scandal in hockey. "Let 'em play" cements hockey's reputation (in the US) of being nothing more than random assault and battery on skates.

At 10:23 a.m., May 07, 2008, Blogger breed16 said...

Thank you for highlighting this, Mr. Mirtle. As a Caps fan I am still bitter with how the refs influenced the outcome of game 7.

As your statistics show, refs have consistency problems. One ref to the next you'll get a different opinion on what constitutes a slash, a hook, interference; that is OK. The problem is when each ref cannot maintain his own standards. The result is a whirlwind of confusion amongst players, coaches, fans, and too much attention for the men in stripes. They should fade into the background instead of being at the forefront of discussion.

At 10:36 a.m., May 07, 2008, Blogger Adam C said...

It's a fair point that players may be more careful not to take penalties late in the game. It's also fair, I think, to argue that perhaps referees are "setting the tone" by calling penalties early, and that the players learn to avoid committing those same infractions later on. Either of these effects would be hard to incorporate statistically.

However, my experience watching the NHL leads me to doubt it; I'd more likely believe that the referee's standards change.

Is this still fair as long as it's done equitably? I don't think so. Suddenly players who have been successful all through the regular season can't be successful anymore because they're getting hooked and slashed with no repercussion. They have to play a different style of game that they may be less effective at, all because the rules have changed.

As Doc Nagel wrote, "let 'em play" means 'let no-one play'. In the bad old days, this meant no comebacks were possible because any scoring chances were shut down by hooking and holding. I don't want to go back there, and I don't want to see this kind of refereeing creep back into the game.

At 10:53 a.m., May 07, 2008, Anonymous Anshu said...

Suddenly players who have been successful all through the regular season can't be successful anymore because they're getting hooked and slashed with no repercussion.

That's the crux of my concern with officiating too. You build your team for speed and skill, cruise to 100+ points in the regular season, and then get gooned out in the 2nd round because the rules change.

The league needs to improve their overall officiating strategy in order to enhance the league's credibility in non-hardcore markets.

I'd argue this is true for other sports as well - soccer being a good example. For the non-fanatic viewer, the unpenalized and/or rewarded diving is a real turnoff.

At 11:05 a.m., May 07, 2008, Blogger sd57 said...

Is it not possible that the officials are being instructed by the league to put their whistles away?

At 11:52 a.m., May 07, 2008, Anonymous Anonymous said...

Dugg this one.

At 12:53 p.m., May 07, 2008, Anonymous Anonymous said...

The problem with situational officiating, such as not calling borderline penalties in close games, is that it undermines the purpose of even having a rule book in the first place. A penalty in the first period of the opening game of the season should be the same as a penalty in the third overtime of a Cup Final Game 7. Maybe a flowchart should be added as an appendix to the rules: "Is the game close? Yes -> call a penalty / No -> what penalty?"

It's no wonder that players get so frustrated with 'phantom' or make-up calls since they can commit the same infraction on several different occasions but only end up in the box when the game situation dictates that it's acceptable to actually call a penalty. If refs want to put the whistle away late in a game, then the NHL needs to institute a second rule book entitled "The NHL Official Rule Book in One-Goal Playoff Games" (the cover can be the same as a traditional rule book with all the bothersome pages between the table of contents and index removed).

At 1:16 p.m., May 07, 2008, Blogger James Mirtle said...

1) How have you controlled for the possibility that players are less inclined to take penalties in the 3rd with the game on the line, than they are in the first two periods?

I don't think it's possible to, statistically, account for that, but I've caught the majority of the games this season my eyeballs tell me the trend has been refs putting away the whistle rather than any "extra caution" used late in games.

That certainly is a factor, but not the factor.

Also, note that in games where the score's out of hand, players will take more penalties. I think it would balance out a bit.

At 1:36 p.m., May 07, 2008, Blogger mike said...

Does the league grade referee performances at the NHL offices? I am curious, I'd imagine they do but that the info is not for public consumption. And if so, are playoff performances awarded on merit or on union-mandated length of service clauses?

If anyone from James on down knows the deal, I'd appreciate the shared knowledge....

At 1:44 p.m., May 07, 2008, Blogger James Mirtle said...

Yes, the referees are all graded based on performance and they're given assignments based on those grades.

We've seen veteran officials on the sidelines at this time of year more often in recent seasons.

At 2:00 p.m., May 07, 2008, Blogger mike said...

I thought as much, James, but I was hoping that this year's playoff refs didn't represent the top referees in the's a tough job, but the refs seem to be ruining many a playoff game this year.

To let 'em play or not? It's a tough question. Protecting players from injury should be a priority but a game with less calls means in most cases there's less chance for a ref to screw up. The NHL guys are still better than NBA refs, and they haven't been caught cheating or gambling like European football refs.

At 2:07 p.m., May 07, 2008, Blogger Jawsh said...

I just wanted to say that the work/research you put into this blog shows and is appreciated.

At 3:17 p.m., May 07, 2008, Anonymous Anonymous said...

"Let 'em play" is a travesty of the pre-lockout era and it's creeping back in.

I'm no Rangers fan, but Jagr took a consistent bushwacking in the series v PIT. But Cosby's phantom phlops got called.

Yet Bartman calls a Royal Commission and changes the goalie interference call in the middle of the playoffs because of one guy's antics.

I don't know how anybody can take NHL hockey seriously, frankly. It's like the league and its officials are just making things up on the fly, WHA-like.

At 5:10 p.m., May 07, 2008, Anonymous baroque said...

I have zero problem with "letting the players play" ...

PROVIDED you follow that with "according to the rules."

Otherwise, I agree with an earlier commentor - you might as well not have a rulebook at all, abandon all pretense at being anything other than entertaining, and instruct the officials to call penalties based on who is ahead or behind and make sure that the game stays close so the fans don't get bored with the contest.

At 9:29 p.m., May 07, 2008, Blogger J. Michael Neal said...

1) How have you controlled for the possibility that players are less inclined to take penalties in the 3rd with the game on the line, than they are in the first two periods?

I don't think it's possible to, statistically, account for that, but I've caught the majority of the games this season my eyeballs tell me the trend has been refs putting away the whistle rather than any "extra caution" used late in games.

You could make an attempt at it by carefully watching video and deciding what is and isn't being called. I'm not sure it would be worth it, though.

What you can do is control for the size of the lead. You could turn James' graph of penalties during each five minute stretch into a three dimensional matrix, with the third dimension being the lead. You could also throw in type of penalty.

It seems as if there is an interesting multiple regression equation to be built here, but I'm probably not going to take the time to do it myself. I'll just leech off of whatever James decides to do for us.


Post a Comment

<< Home

Links to this post:

Create a Link


Free Page Rank Checker
eXTReMe Tracker