One of the biggest challenges facing officiating in the NHL is the effort to mitigate bias or persuasion as much as humanly possible.
I say humanly possible because it’s critical to note that the referees calling penalties and overseeing the run of play aren’t automatons. Despite rigorous training and a wealth of experience, it’s inevitable that at some point in time, an official will slide under the microscope of a fan base that believes their team has been unfairly victimized by the whistle.
I will defend referees to some degree here: I think examples like Stephane Auger and Alex Burrows are few and far between. We don’t have more than a handful of examples linking an official to a player or a team in a negative fashion, and I think that speaks volumes about the kind of training referees enjoy before getting to the NHL level. More than anything else, the NHL – all sports leagues, for that matter – work to protect their product against these scathing allegations. The credibility of the game must be preserved, and any thought or effort to submarine it generally ends in an official being pushed out of the league.
What I do find problematic is that referees, collectively, seem to prioritize balancing the power of a game above all else. This obviously conflicts with their actual priority, which is to observe penalties and call them accordingly.
Increasingly this year, I have noticed hockey broadcasters pointing out during games that since their team had received a number of favourable penalty calls in a row, they should not expect the next call to go their way. Or vice versa. The theory is rooted in this is what we touched on earlier: referees are cognizant of how frequently a team is parading to a penalty box, and rather than draw the ire of a coaching staff and his team, they’d rather balance the game by offering an equal number of power-play opportunities.
Believe it or not, there is actually a lot of truth to these observations. Referees are reluctant to allow teams to go on ‘streaks’, racking up man advantage after man advantage.
We can observe this by just looking at actual penalties called versus expected penalties called. For this analysis, I pulled all penalties, excluding fighting majors and delay of games for puck over glass, for the 2014-2015 season. Then, I looked at every game (min. four penalties called) to see how many penalties a team would historically take before their opposition took one of their own.
We can get pretty close to what we would expect to see if we assume that every penalty call is a coin flip; one that can be rewarded to either team on a 50/50 basis. The idea is that we should be able to see a reasonable estimate of how frequently a team should incur three, or four, or five power plays in a row without heading to a penalty kill of their own.
So, what does the data say?
The data confirms a lot of suspicions. NHL referees get a bit gun-shy as a specific team starts to rack up power plays. Referees appear to look for opportunities to even up the game. It’s why there’s such a big variance in our first column between actuals and estimated – even at a one penalty threshold, human referees are generally looking to balance the power.
On the other hand, referees in our estimated model just keep adhering to the likelihood that any team is approximately 50 per cent likely to take a penalty if a penalty is committed on the ice. You’ll see that the referees in our estimated model are far more likely to award consecutive power-play opportunities – sometimes as many as four or five in a row – over the course of a single game. Our human referees, however, are much more resistant to this development.
The next time you hear someone suggest that a team isn’t drawing the next penalty because of a favourable streak of power plays, know they’re not just regurgitating some old-time hockey quote. For better or worse, NHL officials – consciously or subconsciously – work to ensure that a balanced call of penalties is prioritized.