Oilers Officiating Review – The Big Finale

Over the previous seven articles, I have detailed the officiating for the final seven Oilers games of the 2015-2016 season. Links to the previous articles are at the bottom of the page.

Through this process I have tried to communicate my approach, relay the results that I collected, remove opportunities for bias, and propose some questions worth exploring.

Disclaimer – I acknowledge and confirm that 7 games is an almost negligible window of data from which to draw any conclusions. I also realize that for many, this small a sample size would outright discredit the entire exercise. All that being said, I hope this will prove enough to at least act as a starting point, a sounding line so to speak, by which we might begin to survey the obscured landscape of NHL officiating. As such, my approach in this exercise isn’t going to be specifically a statistical one, but rather something akin to a sociological perspective using empirical data to corroborate or disprove fan observations.

In this article I will be revisiting those questions as well as adding some specific data collected by comparing the sum total of the data from the Oilers games to that from the non-Oilers games.

I should note that due to viewing limitations the data from non-Oilers games is made up primarily from Flames games. However, these games included approximately half of the same teams played by the Oilers as well as some not included in the Oilers’ data set. As well, because of the wider range of available teams and the NHL schedule, the non-Oilers data is comprised of eight games to the Oilers’ seven.

I’ll begin with the complete data from the Oilers’ games.

For those that have been reading regularly, there isn’t anything here that should be particularly surprising. The penalties meted out are more or less even and the difference in uncalled infractions is about 2/1 in favour of the opposition. Scan through the numbers and I’m sure you can find something of particular interest. Certainly, there is enough here to warrant a more thorough investigation.

Below are the numbers after we remove the borderline calls and separate what remains into the technical and physical categories.

Rather than dither on about this or that specific bit of data, I’ll just point out that those large grey swatches in the charts represent the number of times the opposition committed a technical or physical foul, respectively, worthy of a penalty and yet received none.

Even though these numbers were collected by an avowed fan of the team in question, that kind of result bears further scrutiny.

Here are some specific numbers and data points I pulled from the data collected.

• The most frequently fouled Oiler was Hall (tripping), followed by McDavid (hooking, holding, interference), Eberle (hooking), Hendricks (hooking, holding), Draisaitl (tripping).
• Edmonton was most likely to commit hooking infractions followed by interference.
• Edmonton was most likely to be targeted with hooking infractions followed by interference, slashing and tripping.

Below are tables that show the breakdown of how often a penalty of a particular type and severity was called overall, and how often they were called in Edmonton’s favour compared to those called in the opposition’s favour, as well as the same rates of call in the non-Oiler games. (I apologize for the poor resolution, you can click on the image to enlarge)

First, the “chance of call” column represents the average likelihood that a penalty in that category would be called.

But, notice the discrepancy between the chance that same penalty would be called in Edmonton’s favour versus in the opposition’s favour.

Now, and this is very important to what we’re trying to determine here, compare those percentages with the far-right column that shows the likelihood of a call for the same penalty category in the non-Oiler games I recorded.

The numbers there are higher still than any favouring the Oilers yet lower than those favouring the opposition faced by the Oilers in all of the games recorded.

To put that in plain English, it looks as though teams facing the Oilers were more likely to get calls in their favour when playing the Oilers than against some other teams.

Were our sample size larger, this would represent a very clear, and distinct argument towards bias by the officials.

More on the sample size later.

Here you can see the average rate that a given severity of penalty was called in the non-Oiler games, compared to those same rates in the Oiler games. I have noted on the far right side the number of powerplays required in order to bring the Oiler games up to the average seen in the non-Oiler games.

Over the season the Oilers had 2.89 powerplays per game, and 3.10 penalty-kills per game.

If the Oilers were to draw penalties at a league average rate and based on the number of uncalled infractions, they would be gaining approximately 4 extra chances with the man-advantage per game, bringing the per-game average to around 7 powerplays a game. Conversely, if the referees chose to even up the game by calling fewer penalties against the Oilers, they would need to cut back by 1.5 penalties per game, or roughly halving their per-game average this past season.

Here is a breakdown of the chance of the Oilers’ receiving a powerplay based on the kind of infraction committed by the opposition.

You can see that some penalties are far more likely to be called (boarding, holding and high-sticking, in descending order) while an infraction like cross-checking, which can result in serious injury, was highly unlikely to be called at all and interference calls were essentially non-existent.

Now, over the course of the previous articles I had highlighted certain questions that I wanted to address once we had looked over the entire data set.

Game one – Referees will tend to try to find a balance in the number of penalties called in a game, outside of a few circumstances, despite the actions of either team…what does it mean about the integrity of the game when teams receive equal powerplay time but commit infractions, some of them potentially injurious, at wildly different frequencies?

Consider this: if a team knows that they are going to be penalized roughly as often as their competitors, why not simply use every effort to bully and intimidate them, committing virtually every foul at your disposal shy of putting your opponent on a stretcher, knowing that if you do get penalized you are at least reasonably likely to get a “make-up call” at some point in the game?

Game two – We begin to ask ourselves whether the game is being well-served by balancing the penalties against each team?

Game three – ought the Oilers to engage in a more contentious style of play, willfully breaking rules, under the assumption that in either case penalties will be doled out at a relatively even pace regardless of how often infractions are committed.

Game four – did the Oilers leave unpunished physical infractions on the table, so to speak, out of concern they would be penalized or because of a lack of willingness to engage in that style of game?

Game five – Based on this information I believe we are faced with two options:

One – If we wish to ascribe to the “game management” theory of officiating, that is, that referees remain acutely aware of the penalty count to either team and work to match penalties to a more or less even state depending on the circumstances of the game, then we must accept that the application of this method failed completely and utterly in this instance.

Or

Two – If we wish to ascribe to the theory that officials are more prone to offer powerplay opportunities to teams that are playing from behind, then this game would stand in direct contradiction as the Oilers were penalized four times before receiving a powerplay in their favour, by which time they had been playing from a position of deficit for some time and were, for all intents and purposes, out of it.

Game six – I think we’ve arrived at the point where we have to begin to look at the columns in the tables below and ask ourselves “what is fair?” Is fair an even distribution in the “Called” columns, or in the “Missed” columns? Given that a powerplay goal can have a significant impact on the outcome of a game, allowing a team a lead, a tie, or cutting a deficit, is it reasonable to suggest that powerplays hold greater importance within the overall game structure than uncalled penalties, or unrealized powerplays? As such, should on-ice officials operate under a mandate that emphasises in equal measure they are to mete out powerplays as well as punishment? Does this place officials as the gatekeepers of sorts to a game-defining position that is perhaps at odds with the best interests of the game? In other words, is it better in the long run that games be considered fair by the measure of both teams having equal opportunity with a man-advantage than that games be measured as fair by way of both teams being held accountable to the rules and structures of the game as a whole?

Game seven – the (game 7 physical) table and chart represent the only instance in this brief exercise where we have seen the Oilers carry the majority of uncalled infractions, notable because it is the exception that at the very least appears to suggest a rule.

The general theory going into this exercise was that the referees called infractions against either team based on an inconsistent threshold relative to their perception of both the teams and the players involved, with an eye to ensuring that penalties were approximately even at the conclusion of the game.

That theory has proven true to the extent that referees try to keep penalties even by game’s end.

There isn’t enough data yet to confirm that any specific team, or players on that team, receive a greater share of the benefit of the doubt when it comes to calling infractions.

However, there does appear to be some hint that there exists a level of bias being exercised by the officials to the detriment of the Oilers. In order to either confirm or deny this early indication I will be weighing the work involved in gathering a larger, season-long data set of both Oiler and non-Oiler games for this coming season.

My personal opinion at the conclusion of this series is this:

1. The officials have a standard for rule enforcement that does not entirely align with the written rules but they are well within their professional prerogative to apply that standard as they see fit.
2. All standards of rule enforcement are valid only so long as they are applied universally and without bias or preference.
3. The Oilers do not receive a level of rule enforcement that is on par with that of their opponents.
4. The Oilers have some leeway to be more aggressive and to commit a slightly larger share/wider range of rule infractions in order to compete in games.
5. The Oilers would be justified in being reticent to play a more aggressive style given the likelihood of a higher rate of call on infractions that benefit their opponents, thus putting themselves at a disadvantage.
6. There is a strong case to be made that the appearance of the officials’ preference to “even-up” the calls for each team is, in fact, detrimental to the overall entertainment value of the game from the perspective of the fan.

I thank you for reading this series and welcome any questions or constructive comments you may have in this regard either in the comments section below or to me directly via twitter (@codexrex).

For those interested, below is the raw data for every Oilers game covered in this exercise, followed by the raw data for the corresponding non-Oiler games.

Here is the raw data from all the recorded non-Oiler games.

Game one

Game two

Game three

Game four

Game five

Game six

Game seven