When looking back on a season, it is always relevant to wonder how much of a team’s successes or failures had to do with luck as opposed to skill or other factors. Luck is a very relevant statistic in sports, it is just one that is generally not quantifiable and mostly seen through the eye test. This makes it tough for analytics and numbers people to utilize. Utilize it they must, however, if they want to create sustained success year after year. If you want to break a team down after a bad year, you better be sure that the team was bad because of something other than luck. And if they were good, you want to make perfectly sure that they were good for reasons other than luck before doubling down on that team and locking it in long-term.
Every year, for the most part, I try to go back and look at the New Jersey Devils season and see how lucky they were. Years ago, I read about a method for attempting to quantify luck from Rob Vollman in his Hockey Abstract, and I thought it a good thing to remember and check in on each offseason. So, that being said, welcome to the annual luck post here at All About the Jersey!
Vollman’s method for quantifying luck boiled down to 5 different statistics, which when added together, can give a broad view of how lucky or unlucky a team was. They are PDO, Special Teams Impact, injuries as measured by CHIP, record in post-regulation games, and record in one-goal games. Let’s take a look at each one.
PDO is perhaps the easiest one of the five to visually see, as it is a statistic regularly tracked by any hockey statistics and analytics website and is a stat that symbolizes luck. It might not be the entire picture, but it is a luck stat. Simply put, it is the combination of a team’s shooting and save percentages at 5 on 5. When added together, luck-neutral teams should have a PDO around 1.00. Anything significantly under that and a team was unlucky, either snakebitten or without even replacement-level goaltending play. On the other hand, anything significantly over 1.00 and a team might expect to come back down to earth next year with more realistic shooting or save percentage numbers. Of course, some of that could be skill. The Isles had a very high PDO last year thanks to a crazy high save percentage, but Ilya Sorokin is also just very, very good, so probably not all of that is luck. Nonetheless, it is a decent piece of the overall picture.
Last season, according to Natural Stat Trick, there were 17 teams with a PDO over 1.00 and 15 teams under 1.00, so it worked out fairly evenly over 82 games as you would expect. And the Devils ended up with a PDO very close to luck-neutral, coming in at 1.003, ever so slightly above. This ranked the team 16th in the NHL in PDO. For all intents and purposes, they were a luck-neutral team in this stat. With a team shooting percentage of 8.68% and team save percentage of 91.63%, they were slightly luckier with save percentage than shooting percentage, but only slightly. They ranked 13th in the NHL in save%, but 16th in shooting%.
For a team that really broke free and rose way above expectations this past year, seeing PDO not be crazy high is a great sign to start us off. Some teams go from worst to first thanks to inflated shooting percentages or insane goaltending. The Devils did not have either of these, but neither were they snakebitten or unable to stop pucks. They were right in the middle, and that bodes well for them being able to replicate that next year.
Special Teams Impact
Next, we look at the team’s Special Teams Impact, or STI. This is a similar stat to PDO in that it combines two distinct stats into one combo stat to paint a better picture. Here, we are combining power play and penalty kill percentages. The closer a team is to 100% here, the more luck-neutral they were on special teams. Again, maybe not all of this is luck, but this is just one piece of the puzzle, so it needs to be looked at in combo with the other stats here.
According to NHL.com, last season the Devils ranked 13th in the NHL in power play percentage, coming in at 21.9%, while they ranked 4th in penalty kill percentage at 82.6%. This makes sense, they were certainly a better PK team than a PP team last year, and that is usually true of most Devils teams of yesteryear to be honest. I am actually a little surprised they ended in the top half of the league in PP%, I expected it to be lower, but maybe in my memory I only remember the bad power plays where they can never seem to set up in the offensive zone. It could be a personal issue on my part. Overall, combining those two numbers gives the Devils a STI of 104.5%, a lucky number overall. That is a number that is pretty far away from the luck neutral 100% all things considered, so we can determine that the Devils were fairly lucky when forced to play uneven hockey, whether up a man or down it, but especially down it. I agree with this, considering I really think they should have had a worse power play number.
Next, and probably the most interesting of the five where I am concerned, is calculating the luck impact of injuries on a team. The ideal way to do this is to look at the cap hit of injured players to a team overall. You cannot simply just look at man games lost to injury, as that does not take into account the caliber of player that is injured. Losing Jack Hughes for 20 games is way more significant than losing Nathan Bastian for 20 games, those 20 man games lost to injury are nowhere near equal. So, to showcase the difference, we look at the cap hit instead of simply the total number of games lost. If you lose Jack’s $8 million cap hit for 20 games, that is way more significant than losing Bastian’s $1.35 million cap hit as it wil be this coming year, and that picture more accurately reflects the effect of the injury on the team.
The place to get this information is from the NHL Injury Viz, which does a good job of tracking this information and putting it into readable charts. I can’t screenshot the entire CHIP summary chart on my screen, the chart is too big, so please go to the link and check it out. Overall, it shows the Devils as having been very luck-neutral in CHIP this past season. The team ranked 15th in the league with a total of $10.19 million lost to injury. In that number, $3.67 million was lost due to forward injuries, only $1.08 was lost due to defensemen injuries, while $5.44 million was lost due to goaltender injuries. For comparative purposes, the worst team hit, Montreal, lost $21.21 million just from forward injuries alone, while the luckiest team to avoid injuries last year, the New York Rangers, lost only $1.48 million total in cap hit from injured players, losing absolutely no cap hit from goaltending.
However, when you dig into it a little more, a lot of that number comes from Jonathan Bernier, who of course was on LTIR for a second year in a row. It feels almost wrong to include him, as he never had a chance of playing. If you remove his entire $4.125 million cap hit from the Devils CHIP last year, they would’ve ended up with a total CHIP of $6.065 million, which would’ve made then the 8th luckiest in terms of injuries. So, it’s up to you how you want to view that. However, I think for the purposes of this article, I am going to stick with where the Injury Viz has the Devils, ranked 15th, and consider this a luck-neutral stat from last year.
Next, we want to look at the Devils’ record in games that went beyond regulation. Whether the game ended in overtime or went to a shootout, the idea here is that a luck-neutral team would be around .500 in these games. The luck especially comes in with the shootout, where it is a crap shoot as to which team is going to win. 3 on 3 overtime constitutes more skill, but in the end it can all come down to possession, and that oftentimes can be the result of puck luck. So, a team that has a really good or really bad post-regulation record can certainly be said to have been affected by luck to some degree.
Last year, the Devils were definitively lucky in games that went beyond regulation. They ended the regular season with a 13-8 record in these games. That is a direct +5 points to the record thanks to those extra wins over losses. Now, a lot of those games ended in overtime and never got to a shootout, and I do think some of that is skill-based. They threw out Michael McLeod to win a faceoff thanks to his strong skill there, and it gives the Devils initial possession, and their skill keeps that possession and opens up strong scoring chances. However, there is also some luck there too, as games did go to a shootout that they won, and sometimes you just get that bounce of a puck. When you shoot it in 3 on 3, the rebound can all be luck, and who is in the right spot to get it and regain possession. So, here I have to say the Devils were somewhat lucky last year.
Finally, this is very similar to the last one, but strictly looks at games that ended in regulation. This is included for a similar reason as post-regulation games. In a full 60-minute game, one goal can often be the result of luck. The way a puck bounces, how it just misses a stick and moves unexpectedly, how it happens to bounce off of a goaltender, or whatnot. You name it, there are a million scenarios where something just works out and leads to a goal that could win a tight game. So, realistically, a luck-neutral team would have a record around .500 in one-goal games, as sometimes the puck bounced their way for that game winning goal, and sometimes it went against them. Again, you can debate that this isn’t all luck-based, and that is ok. The point is, when all five of these stats come together, it paints a picture of luck that is the best we can get, even if not all of the individual stats are completely based on luck.
Last year, the Devils were much more luck-neutral in one-goal games than they were in post-regulation games. In these games during the regular season, they went 7-6, just slightly above .500. And since you can’t get to .500 with an odd number of games played, this really is as luck neutral as it could have been for the Devils.
So, let’s take those 5 categories now and sum them up. With PDO, the Devils were the epitome of luck-neutral. With STI, they were undoubtedly lucky. With CHIP, they were luck neutral with the inclusion of Jonathan Bernier, and were somewhat lucky without including his cap hit. In post-regulation games, they were definitively lucky, but were absolutely luck-neutral in one-goal games. Put this all together, and you get 2 stats where they were luck-neutral for sure, 2 stats where they were very much lucky, and 1 which you could debate on.
You decide how you want to categorize that all when you sum it up, but to me, that comes together to mean that the Devils were somewhat lucky last year. They were not crazy lucky, they did not jump all of those points in the standings thanks to luck and luck alone. But there was a small part that luck played in it, for sure. This especially hits true with the special teams impact and post-regulation games. Next year, watch out for those two stats and see how the Devils perform. Do they lose some overtime games that they might have won last year? Do they give up more goals in penalty kill situations? We will see, but it is certainly possible. However, because they were not overly lucky, they were luck neutral in other areas, I think it still is a positive outlook overall. A lot of what they accomplished last year was done in a luck-neutral environment for the team. They were never really unlucky anywhere, but in many areas were not lucky either. They earned it, and that is a fantastic sign.
Anyway, that sums it up for the luck post for this year. What do you think about these numbers and what it says about the Devils last year? How would you categorize luck for the 2022-23 Devils? Please leave your comments below, and thanks for reading!