In the middle of this past week, the National Hockey League made three changes to their rules for the 2017-18 season. Scouting the Refs has a breakdown; I’ll summarize the first two. The first is a big one: coaches’ challenges for goals due to offside play will charge a team with a minor penalty if they are wrong. They’ll still have a timeout, they would just be down a man if their challenge failed - meaning they’ll be on a penalty kill after conceding a goal. The second is that a puck played with a high stick will result in a neutral zone faceoff nearest to where the puck was played. So such plays will not result in automatic offensive or defensive zone draws. The third is that a team can no longer call a timeout after they ice the puck. And that’s a rule change I want to delve into because I have a lot to say about icings.
During this Summer, I wanted to finally answer a question I received quite a few times last year from very smart and respectable people: “The Devils ice the puck a lot, who does it the most?” Icings are not actually counted by the scorer beyond that there was an icing event, the time of the event, and who they thought was on the ice for it. Nothing about who committed it and definitely nothing about what happened after or how it happened. Icings are not good events because they give the other team an offensive zone draw and the team who took the icing cannot change players. But there was little data beyond that; I wanted to see what other impact there was in addition to answering that question. Therefore, I counted and recorded information of every icing that happened in every game by both teams in 2016-17 New Jersey Devils’ season. I even included some of the feedback from readers like yourself in the counting. Now, I can provide some actual data about this rule change. What will it actually change? What can be its impact?
Icings & Timeouts
In the 2016-17 Devils season, there were 689 icing events recorded and 681 actual icings called on the ice. Yes, there were eight instances where a puck out of play or an offside play was recorded as an icing. I repeat: the castles are built on sand. Just to get it out of the way, a quick breakdown. Out of those 681, 372 of them were by the Devils - or 54.6% of all icings in Devils games. I do not have every other team’s count so I cannot tell you if 372 by one team is bad or not. It does not appear good that the Devils committed more icings than their opponents, though.
Out of those 681 legitimate icings, 19 of them led to timeouts being called. That’s 2.7% of all icings in 82 Devils games. Of those 19, 13 of them were called by the Devils. The other six were by the opponents. Teams only have one timeout to call in a game. There was only one game when both teams used their timeout after an icing: January 3, Devils at Carolina. That means there out of 82 games, 18 of them involved a timeout being called after an icing - and 13 of those were by the Devils. From this, it can be said that a timeout being used after an icing was occasional but not common occurrence in a Devils game last season. It can also be said that the Devils used it more often than their opponents.
In terms of timing, the majority of these timeouts called after icings happened in the third period. Only two of the nineteen occurred in the second period and none were in the first period. Those two in the second period were split; one by New Jersey and one by an opponent. There were seventeen third period icings that led to timeouts. Eight were at even strength situations, five happened when the opponent had pulled their goalie, and four happened when the Devils had pulled their goalie. Among the eight even strength timeouts, six were by the Devils. All four timeouts called when the Devils had their goalie pulled and there was an icing were by the Devils. Among the five when the opponent had their goalie pulled, the Devils called timeout two of those times. From all of this, we can conclude that these timeouts were commonly called after icings late in the game with a slight edge toward extra skater situations. The Devils utilized it more often in both even strength (seven to three) an extra skater situations (six to three). As a quick aside, the Devils never iced the puck while having their goalie pulled in 2016-17. Neither did their opponents.
In most cases, the Devils called their timeouts after icings after they iced the puck. Nine of their thirteen timeouts after icings were called when the Devils iced the puck. It is enough to not just assume that a Devils timeout would come after a Devils icing, but it is enough to expect it. Out of the six instances of the opponent using a timeout after an icing, four of them were after the opponent iced the puck. In effect, this rule change would have prevented thirteen timeouts if it was in place last season and assuming all other events would have remained the same.
The Real Time of an Icing & Timeout
When an icing is called, the play and the game clock are stopped. Real time never stops. The team that took the icing stays on, the other team can (and does) make player changes, and the referee skates to the appropriate faceoff dot, waits for both teams to settle for the draw, and does it. For the 662 icings in the 2016-17 Devils season that did not lead to a timeout, the average real time from whistle to puck-drop, was 31 seconds. For the 19 icings that did involve a timeout before play resumed, the average real time before play resumed was 1:46. That is a over three and a half times the average length of time for other icings.
How? A timeout is thirty seconds long. What I observed from the 19 icings is that the timeout is usually called before the faceoff happens. Players can start to be set-up and then a referee gets word that a coach is calling their timeout. This is already after twenty to twenty-five seconds has already passed by. Players on both sides skate at their own pace to the bench for the thirty seconds. There is some leeway after the thirty seconds ends for players to leave the bench. Then they all have to skate back to the faceoff and start setting up again. This set-up can be delayed further if the player adjusts equipment for a few seconds, or the referee re-does a faceoff or removes a player from the dot, which usually takes more time than what a player can do. That’s how a thirty second timeout can lengthen a thirty-one second average time between icing call and a puck drop.
The range of the real time in these 19 icing situations was 1:25 to 2:24. It is not exactly a full rest between shifts, but it is enough time for a group to catch their breath, get some instruction from the coach, and prepare to resume play. The removal of timeouts after icings will mean that is gone. The team that takes the icing will have to settle for around a half-minute of rest since their coach cannot use his timeout. Which is not evenly distributed depending on what happened prior to the icing; especially if someone was chasing that puck down and lost the race to it.
This half-minute might be stretched out for equipment reasons, but from the 662 icings I observed, any real elongation of time came from the referees from performing a faceoff twice, removing players from the faceoff, or other decisions like video review. Even those did not happen often enough to skew the average well beyond thirty seconds. Going to get a new stick or fixing a skate was rare and any serious equipment issues like a skate missing its blade led to the player going to the locker room. I doubt such events will increase in 2017-18 because they were not commonly done last season. It is a hard thing to fake and I would think the league would crackdown on that quickly if it does start happening more often.
The Impact of the Icings with a Timeout Called
I defined impact as shooting attempts taken by the non-icing team before there is a zone exit or a stoppage in play. I counted the attempts myself instead of relying on a scorer that may or may not have understood what icing is. Among the 19 icings with a timeout called, twelve were icings by New Jersey. Six of these twelve led to no shooting attempts whatsoever by their opponents. New Jersey won three of the post-icing faceoffs among those six instances. Two of those six instances included an opponent who pulled their goalie.
The other six that led to shooting attempts all included faceoff wins by the Devils’ opponents. Those six totaled twelve attempts. Three of those six included opponents who pulled their goalie. Three opponents put up three shooting attempts; the other three only put up one. None of them scored. Only two did not have a proper zone exit, as regulation time ended to stop the attack.
The other seven icings with timeouts were by the Devils’ opponents. The Devils did not generate much from those situations. Despite winning the faceoff in five of those cases, they only generated two shooting attempts on one of those five. The Devils were able to attempt one shot in the two situations where the opponent iced it, a timeout was called, and the opponent won the faceoff. All three attempts did not lead to a goal.
These special circumstances did not yield much in the way of offense. When a team, the Devils or their opponents, generated a shooting attempt after an icing with a timeout, it often occurred after that team won their faceoff. This will be more apparent when I cover the whole season of icings, but icings alone will not lead to attempts. Winning that faceoff appears to be a factor - and even that does not guarantee an actual offensive play.
What about the icings that did not lead to a timeout? There were not a lot of shooting attempts either. There were a grand total of 352 attempts combined the Devils and their opponents out of the 662 icings without timeouts. It is an average of a little over a half of a shooting attempt per icing. 143 of those 352 were by the Devils after their opponents iced the puck. 113 of those 143 attempts happened after the Devils won the offensive zone faceoff (they won 146). Out of those 140 attempts, six were goals. Just six goals. For the opposition, they took 209 total attempts off Devils icing calls. 168 of those attempts happened after the opponent won the offensive zone faceoff (they won 188 draws). The opponents combined scored a total of just six goals. You may remember two of them as heartbreakers from last season but at the end of the day (and good examples of why icings have a potential for being really bad), it’s still only six goals over the course of a full season. Both the Devils and opponents were able to generate most post-icing offense if they won the post-icing faceoff. Even then, it’s less than one shot attempt - not even a shot on net, just an attempt - per icing and even per icing plus a post-icing faceoff win.
Therefore, I conclude that icings really did not yield a lot of offense with or without the timeout. Not by the Devils or their often-superior opponents last season.
The Scouting the Refs post quotes Rangers GM Jeff Gorton and NHL Commissioner Gary Bettman justifying this rule in the name of offense. On paper, it makes some sense. However, the icing data from what I tracked from the 2016-17 Devils season suggests otherwise. It is possible that the Devils are outliers and other teams have conceded more offense after they ice the puck. However, given that this is involving 681 icing events across a full season, I doubt that the Devils were exceptional all season long. Especially since the Devils were a poor team last season.
Based on last season’s data, I really do not believe this rule change for icings will really impact the Devils or even the league as a whole. Yes, no timeouts may mean the team that committed the icing cannot get one longer-than-usual break. Yet, if the team that commits the icing wins the faceoff or makes the right defensive play if they do not, then they can get a zone exit and get some breathing room anyway. For offense to actually be generated by way of a shooting attempt - not even a shot on goal, just an attempt - more has to happen than just receiving an offensive zone faceoff and the ability to change players.
There will be a little real time saved, but even that may not happen as most games last season did not include a timeout taken after an icing or a timeout taken by the team who just iced the puck. Therefore, even the “quality of life” improvement of reducing the real time length of a game is likely to be less than an already very small reduction.
I will not dispute that icings are actually bad if only because the team who commits one suffers from not being able to change players and starting in their own end. They are not good situations to be in. However, they were not damaging throughout last season for the Devils or their opponents on a given night. If that is the case for a bad Devils team, I suspect it would be the case for better teams. Taking away the timeout from the team that iced the puck seems like it could do some good, but I do not see the evidence backing it up. In short, it is a rule change that really, well, just makes a change.
I’d like to know what you all think about this rule change and the data presented in the comments.
I would also like your opinion on the next step with this data. I can and would like to present more of what I found from my tracking of the Devils’ icings from last season. I know a new season is upon us. But if this coming season is supposed to be about taking a step forward, then this data can provide some areas to make some small improvements - which is important as the coaching staff and most of the blueline remains the same from last season. Further, there really is not much data available regarding icing events. Not that many were clamoring for it, Would you prefer to see that as a separate post or series of posts? Or would you like a video of a presentation showing all of this off instead, similar to what I did last year with time on penalty kills?
Additionally, during the tracking project, I realized there was one additional thing to track for icings. Pressure. Problem is: I realized this after tracking the first 203 icings. For a variety of reasons, I did not go back to re-track those 203 icings (a little less than a third) just to define pressure differently. I do not know how long it’ll take. Do you think it is worth doing? If so, I’ll try to push through it. If not, then I’ll jump into summarizing the data and presenting it in someway. I thank you all for your previous feedback with this project.
Oh, and the answer to the question? A tease: It is who you just loved out there last season. But he wasn’t the team leader in icings until he surpassed #2 guy in Devils icings around early March. And #2 may surprise you. I’ll leave that for the next post or the presentation - whichever you would prefer.