Icings by the New Jersey Devils in the First Half of 2016-17

Tonight, the New Jersey Devils begin the second half of their 2016-17 regular season. In their final game of the first half of the season, the Devils ended regulation by icing the puck twice within the final minute of regulation. They tried to waste time by clearing the puck from their zone, only to have the puck go too far and then suffer a defensive zone faceoff twice in a 1-1 game against a team who was picking on them on offense. The game did not end well in overtime, but those events have stuck out to me all season. So much so that I wanted to take a closer look.

For the unaware, icing refers to the act of firing the puck from their own half of the rink (behind the red line) to the end of the other half beyond the goal line without touching anyone. Icing is called when the opposing team is the first to the dots in their own end or if a skater - not a goaltender - is first to retrieve the puck after crossing the goal line. It’s a foul against the team who does it at even strength or if they are on a power play (a penalty killing team can freely ice the puck), but it is not a penalty. Instead, there is a stoppage in play. There is a faceoff in the end of the team who iced the puck, and the team who iced the puck cannot change players while the other team can do so. That is the punishment of an icing: not only does the team who iced the puck have to take a faceoff in their defensive zone but they cannot replace any tired or ineffective players. That is why they are usually not good plays to make. That is why I am annoyed when the Devils do so.

What causes the icing can vary. Three possibilities come to mind. First, icing can be the result of a missed pass going to the other team’s end. Second, icing can be the result of an intentional clearances where the player flings the puck as far away as he could. Third, icing can be called on a dump-in for a line change that is made before reaching the red line. All result in the same: a defensive zone faceoff with the same players on the ice, likely going up against fresh opposing skaters. And whoever wins that faceoff can have an advantage. If the Devils ice the puck and the opposition wins the faceoff, then there is often a shooting attempt or, worse, a goal against. The opposite can happen: an opponent ices the puck and a Devils faceoff win gives them an offensive opportunity.

While icing events are recorded in the play by play log of each NHL game, as far as I know, there are no recorded counts of how many times a team ices the puck in a game or in a season. There is no record in the log for who committed the icing. The only way I know of discerning who did it is by checking where the faceoff is after the event and who was allowed to change their players. My thinking was that a poor possession team like the Devils - after 41 games, Corsica lists their 5-on-5 CF% at 46.8%, the fouth lowest CF% in the NHL - would ice the puck more than their opponents. In person and on TV, I have seen the Devils commit those common acts that lead to icing calls. Whether it’s true or not requires looking more closely at the data.

Therefore, I went through the play by play log of each Devils game from NHL.com and recorded how many icings there were in a game. Yes, I’m assuming the logs are correct. I also recorded when the icing call was made and who iced the puck. I also noted any special information such as who won the faceoff after the icing, whether a goal was scored on the same shift after the icing, and whether a timeout was called to give tired players a breather. I do not have the results of other teams, although I would like to compare the Devils against a better possession team and a worse possession team than them to see if there’s something to it.

How Many Icings Have the Devils & their Opponents Committed?

Data from NHL.com Play by Play logs

Data from NHL.com Play by Play logs

In total, the Devils have taken more icing calls than their opponents, 198 to 179. When the Devils are at home, they have iced the puck thirteen more times than their opponents. When the Devils are on the road, it is more even: the Devils have iced the puck only twice more than their opponents. Still, in general, the Devils have conceded more favorable draws than their opponents.

A period-by-period breakdown showed something interesting. Both the Devils and their opponents iced the puck much more often in the third period than in the other two periods. I wonder if it is due to fatigue; tired players making mistakes at passing the puck or clearing it. Curiously, both the Devils and their opponents have been more guilty of third period icings when they are at home than when they are on the road. I’m not sure why this would be the case at home. Yes, home teams have the last change. But in an icing situation, the team who iced it can’t change players anyway so that would be moot. Are teams just more averse to giving up faceoffs on the road? Maybe? Then again, road games have seen more icings by the Devils in the second period than the third - which may point to a result of their second period issues on the road. It’s something to think about.

Over all 41 games, I can tell you that there has been an average of roughly 9.2 icings per Devils game. The Devils have an average of approximately 4.83 icings per game while the Devils’ opponents average around 4.37 icings per game. That may not seem like much but given how much the Devils have struggled to generate offense and/or play defense over the last month or so, giving away opportunities from icings is worth noting. In terms of how many icings in a game is too many , these figures at least provide some rules of thumb for Devils games.

In recording the icings, I wondered about a time breakdown. Are icings more common in the first five minutes of a period or in the last five minutes? The former represents the beginnings of periods (or games) where teams may be slow to start. The latter represents the ends of periods (or games) where teams are more tired and may focus on just getting to the end of the period than maintaining possession. Here’s what I found among the 41 games (with the exception of the lone Devils icing in overtime

Data from NHL.com Play by Play Logs

While the first five and last five minute groupings both represent a majority of icings in the first and third periods, two figures stand out. The first five minute grouping represented a third of all Devils icings in the first period. The last five minute grouping represented over a third of all Devils icings in the third period. Both point to some issues in various games at either struggling at the start and trying to “ice” a game away. How did the Devils opponents fare?

Data from NHL.com Play by Play Logs

Not as high as the Devils. The Devils’ opponents have committed more of their icings in the middle of periods instead of towards the beginning or ends of them. The Devils’ opponents have committed fewer icings at the starts and ends of games. Whether or not they close games better than New Jersey is a different question, but they’re not throwing pucks away and taking draws in their own end with tired players.

What Has Happened from These Icings?

I recorded who won the faceoff after the icing, whether a goal was scored on the same shift, and whether a timeout was taken. Here’s what I found out:

The Devils should ideally try to win the faceoff of icings. While it does not guarantee that the other team won’t take the puck after a faceoff win and turn it into something, it does put the Devils in a better position. If New Jersey iced the puck, then a faceoff win can alleviate the potential pain from that icing. If the opponent iced the puck, then a faceoff win can give them possession to inflict some potential pain for that icing.

Unfortunately, the Devils have not been meeting this ideal enough. When the Devils have iced the puck, they won 91 out of 198 draws. That’s a faceoff winning percentage of just below 46%. When the Devils ice the puck, the Devils cannot change players, so they’re at the mercy of a drawing someone who’s better and/or fresher at the faceoff dot. When the opponent ices the puck and the Devils can choose who to take the draw, the Devils have won 83 out of 179 draws. That’s a faceoff winning percentage of 46.3%. That’s a little better than when the Devils ice it, but it still means that opponents have been more successful at getting out of that defensive zone start with the puck. In other words, the Devils are more often than not at the mercy of their own icings and not taking enough advantage when the opponent does it when it comes to faceoff wins after the icing call.

Fortunately, icings have not yielded many goals for or against the Devils. When I looked at the log, I paid attention to any goals scored on the same shift after the icing. That meant the scoring and scored-upon team had the same players on the ice from the faceoff after the icing to the time of the goal without any stoppage in between. From the data, the Devils scored four goals on the same shift after an icing and the Devils were beaten for five goals. I was surprised it was only five goals against since two of them were late equalizers that would force overtime (10/20 at Boston, 12/18 at Manhattan). Most of the nine goals came from the other side’s icing. Three of the four goals came off opponents icing the puck and four of those five goals came from the Devils icing the puck. In the excepting cases of the team who iced it and then scored after, they did a great job winning the puck, breaking out, and scoring shortly thereafter. Impressive as that can be (or hard to watch), it only happened once for the Devils and once for their opponents this season. What this means is that icings are not really a significant source of goals within the larger perspective of the season. That said, the potential for damage is there.

Lastly, I counted timeouts. Teams only get one timeout in the NHL and a not unfamiliar usage is after an icing from a long shift. While the other team gets a break, they can put out fresh players anyway since they didn’t ice the puck. So giving thirty seconds for the players who have to stay out there so they can be calmed down, take a moment, and receive some instructions from the coaches can help. Alternatively, a timeout can be called by the non-icing team to set a play up, usually because they’re losing and they want to make the most of an offensive zone start. Two of the six icing-timeouts by New Jersey were taken when the opposition iced it; two of the four icing-timeouts by New Jersey’s opponents when the Devils iced it. None of them resulted in a goal in the first 41 games of this season.

Conclusions and What’s Next?

Looking at this data answered my initial question of how much do the Devils ice it. Knowing that they do so more than their opponents suggests that it is an issue. Perhaps as a result of other, more significant and general issues with the team such as how they play in their own end or how they pass the puck. I don’t think icings necessarily drive wins or good performances (e.g. there have been games where the Devils were blown out but didn’t take most of the icings), but they don’t help in the run of play and that is something the Devils have struggled with. Especially in the third periods of games at home and second periods of games on the road.

It is good to know that few goals have been scored on the same shifts from icings. It leads to a follow-up point of wanting to look at shooting attempts on the same shifts from icings. I think that would point out the real damage of taking an icing call. It is more labor intensive to do so manually (unless there’s an easier way I do not know about), but it can be done in time.

Speaking of labor-intensive, I think the next step would be to hit the videotape of these games and take a closer look at the icings. That would reveal what kinds of icings they were and, just as importantly, who committed them. I would love to know if there’s a Devil or Devils that are just icing the puck more than others. It can be evidence of other struggles by such a player, whether it is in passing the puck or making a zone entry.

As far as value, at least knowing when the icings take place and who wins the draw may be instructive when compared against a shift chart. The team who took the icing call cannot change their players. At the beginning of their shift, this may not be as bad of a situation than if the players were near the end of their shift. Identifying when an icing takes place and when they do or do not win the draw can point to how much that shift is extended. That would provide more of what the cost of an icing call can be from a usage standpoint.

Lastly, I would like to look at other teams at least at a surface level to see how many icing calls in a game would be considered average. We know Devils games have roughly nine on average in the first 41 games in this season. Is that true for other teams? What’s the range for other teams? I’d like to know that just for comparison’s sake.

That said, I’m pleased of what I found and I’d like to explore icings more in the future as far as what they mean. They’re important enough to be stand-alone events. We know they are stoppages where the team who took the icing is punished. There could be more to understand, whether it is for the Devils or in general. Please leave what you want to know about icing tracking and what you have gained from this post so far. Thank you for reading.

A Game by Game Count of Icings in the First 41 Devils Games of 2016-17

As a last bit, here’s the game-by-game count of icings by the Devils, their opposition, and the total by games. It’s not the full log of what I put together, but it’s a good summary to see how often they have happened. By the way, the Devils have out-iced their opponents in 21 of their first 41 games this season.