Teams do not perform in a vacuum. They have their gameplans, their tactics, and their preferences. However, when a goal or two is allowed or the team puts up a lead, things tend to change. Incentives within the game appear. The team that's losing by a few goals has every reason to push forward and attack more to make up the difference. The team that's leading doesn't have as much reason to keep attacking - and perhaps can't given how the opponent may push forward more in attempt to score. For some time, this concept has been referred to as score effects. This is what I will be looking at today with respect to the 2015-16 New Jersey Devils.
This was really inspired by the poor performance on Friday night against the Islanders. The Isles got a fortunate bounce off a rebound to go up 0-1 just under two minutes into the game. The Devils proceeded to play like it was 0-0 for the most part. As far as I could tell, they didn't change their plan in any meaningful way and they didn't change their mindset or approach to the game. They didn't seem to attack more, especially in the third period where with under seven minutes to play, the Devils were still out-shot 3-5 and the team down a goal took a whopping seven shooting attempts. Credit to the Islanders for giving the Devils so many issues and Jaroslav Halak for playing well on the few scoring chances the Devils did create. But it was a night like I've seen plenty of times this season. A one-goal deficit just seemed too much for this season's squad despite having loads of time and two intermissions to reset and make some adjustments. Since I stated that the 2015-16 Devils don't seem to respond to the score, I now want to see if that's truly been the case.
As far as score effects themselves, I wish I could remember who exactly figured it out, but the ideas have been kicking around since about 2010 or so. I've seen Gabe "Hawerchuk" Desjardens have a few posts back in the day about it (here, here, and here), which acts as good background. The general concept is that a team down one or two goals is going to take more shooting attempts as they are losing. Their CF% will be higher than usual. Likewise, a team leading by a goal or two will not be taking more shooting attempts. Their CF% tends to be lower than usual. When leads or deficits go three goals and beyond - which doesn't happen often - effects aren't as strong. Sometimes the losing team will make a mad dash to get back into it and other times, they just pack it in and just play out the game.
That all said, let's get to it. I went to War on Ice on the morning of February 20 to collect the data from their Team Comparison section. Therefore, this does not include Saturday's game against Washington. All data in this post comes from War on Ice, you can check it out before it gets shut down in the future (I do not know when). At War on Ice, team data can be broken down by scoring situation. For each category, I recorded the following. First, I collected Time on Ice (TOI) to highlight how much time the Devils have spent in that scoring situation. Second, I took down the Corsi For percentage (CF%), the Devils' CF% rank, Corsi For per sixty minutes (CF60), Corsi Against per sixty minutes (CA60), and Corsi For and Against per sixty minutes (CP60) and it's rank. This will give me enough information how possession generally went and how many attempts were involved in each situation. Third, I took down shooting information such as Shooting For percentage (SF%), the Devils' SF% rank, the Shots For per sixty minutes (SF60), and the Shots Against per sixty minutes (SA60). This will tell me how actual shots on goal broke down by scoring situation. Lastly, for each situation, I recorded the goals for (GF), goals against (GA), on-ice shooting percentage (Sh%), and on-ice save percentage (Sv%). This will give me an idea of how the results went in each situation.
A Baseline: All Scoring Situations & Tied Score Situations
In order to provide perspective on how the Devils performed this season with and without the lead, I started with two scoring situations for comparison purposes. The first one is all scoring situations. That's right. All of them combined into one. In other words, it's how the team has done all season in the various stats. The second one is tied score situations. This is the most common scoring situation in hockey. All games start tied. Games that go beyond regulation are tied. Tied scores happen a lot. So the numbers from that situation will be useful to see how they change when the Devils are up or down a goal or two.
Here's the 5-on-5 even strength stats for those situations:
In all scoring situations, the Devils are, well, just not an good offensive team. They've been out attempted regularly, they've been out-shot regularly, and they have a negative goal differential to show for it. Their CP60 being thirtiethmeans the Devils' games have the fewest attempts overall. Things look a little better for tied score situations. The Devils aren't next to dead last in CF%, although they're very much in the red. The SF60 is a little better too, although they're still being out-shot regularly. What's worse is that the shooting percentage is low. Opposing teams have tended to break that tie more than the Devils this season. Combined with a low shooting and a low attempt rate, and goals have been hard to come by.
But you know, why only settle for 5-on-5? Let's look at all strengths. Let's throw power plays and penalty kills in there. Let's throw empty net situations and 4-on-4 play, too. Let's see it all!
It's not that much different than 5-on-5! While the numbers are slightly higher across the board - and the shooting percentage is a little more potent at 8% - the same conclusions apply. This is not a team that generates a lot of offense.
In any case, keep some of the numbers from these sections in mind as we move on. First, how the Devils do while they're trailing this season. This was my point of contention on Friday night and surely in other games. I doubt the Devils really
The 2015-16 Devils While Trailing in Games
Here's the 5-on-5 data:
|Trailing by 1||665.1||49.5||28||44.0||44.8||50.6||20||25.5||24.9||88.9||30||13||17||4.6||93.8|
|Trailing by 2||219.8||53.7||24||46.1||39.8||52.5||25||25.4||22.9||86.0||30||9||6||9.7||92.9|
|Trailing by 3+||82.6||48.7||30||40.7||42.8||40.8||30||21.1||30.5||83.5||30||1||2||3.4||95.2|
Color me surprised. By the numbers, the Devils have actually attacked more while down a goal - and even more while down two goals - than they do when the game is tied. CF%, SF%, CF60, and SF60 all improve over the tied score situation when the Devils go down by one or two goals. This means that, statistically, I'm wrong. This is proof that the Devils do play to the score.
The problem is that, well, they don't really play well to the score. You'll notice that while the CF% is higher when the Devils go down by one or two goals, they still rank rather low in the league. When they're down one, their CF% is still below 50%, one of only three teams to have such a value. While their CF% is solidly above 50% when down two goals, it's nowhere near the league median of something like 58%. Likewise, their SF60 goes up but a shooting rate of 25 and a half shots per sixty minutes is still a really low rate. So even though their shooting percentage has been quite good while down two goals, the goals don't come. Worse, the shooting percentage when they're down one goal is a very low 4.6%. Again, a low shooting percentage combined with a low shooting rate - albeit improved over the tied score situation - will yield few goals. We don't see those goals often in general, so when the Devils are down by one or two, one may be justified in feeling that the game's over for New Jersey when it may not be so for others. Ultimately, I should change my feeling about how the Devils do when they're down a score or two. It's not so much that they don't play to the score, they're just not at all that good at it.
As a quick note, the Devils really tend to pack it in when they're down by three or more. It hasn't happened a lot this season, so it is what it is.
Briefly, let's look at the numbers for all situations to include all of the non-5-on-5 situation play:
|Trailing by 1||830.1||50.1||28||45.2||45.0||50.4||20||25.7||25.3||90.2||30||26||27||7.3||92.3|
|Trailing by 2||292.7||54.1||27||45.5||38.5||54.3||24||26.0||21.9||84.0||30||13||10||10.2||90.7|
|Trailing by 3+||97.7||55.1||25||49.7||40.5||47.7||29||25.8||28.2||90.2||30||4||4||9.5||91.3|
As with the previous section, the numbers look better but the ranks and general conclusions remain mostly the same. The CF%, SF%, CF60, and SF60 numbers are higher but they're not that much higher to make the Devils look better with respect to their peers in the league. Or in general. It's still very much a non-offensive team relative to the NHL. Shooting percentages look better across the board. I suspect that's because of how the team performs on power plays, which has been rather effective over this season so far. They can juice a CF% and a SF% (and possibly a Sh% if they're scoring on them) since those situations are all about attacking for the team who has one. But I do not know if that's the full reason.
As a last note, the Devils remain low-event in general throughout all trailing situations. Their CP60 is dead last regardless of 5-on-5 only stats or all situation stats. I wonder whether this is a root of some of the Devils' issues. While they're not allowing a lot, they're just not taking a lot and I question whether that's the right path. Especially since the Devils' offense isn't generating a lot of attempts or shots per sixty minutes or in general.
The 2015-16 Devils While Leading in Games
So I learned something I didn't expect about this season's Devils team when they were losing in a game. How about we look at how the team performs while leading in a game? The sentiment I get is that the team leans a lot on Cory Schneider (and occasionally, Keith Kinkaid) when they have a lead. Let's go to the numbers and see if that makes some sense. First, 5-on-5 numbers:
|Leading by 1||422.5||41.4||26||37.1||52.4||40.9||28||20.9||30.1||89.5||30||14||12||9.5||94.3|
|Leading by 2||191.8||35.7||29||35.0||63.2||37.0||29||18.8||31.9||98.2||28||6||5||10.0||95.1|
|Leading by 3+||111.1||39.0||17||33.5||52.4||41.4||16||19.4||27.6||85.9||29||5||4||13.9||92.2|
The Devils go from very little offense to even less when they have the lead. Opposing teams certainly have turned it up when they go down to the Devils on the scoreboard. They attempt many more shots and many more of those attempts get on net. That the Devils' save percentage is much higher when the Devils are up by one or two goals than what it is in tied situation speaks to how the team really relies on Schneider (and occasionally, Kinkaid) to hold that lead. How can they not? The Devils certainly aren't generating many shots. Their already low shooting rate gets even worse as the lead goes from one to two. Attempts become fewer, although the Devils are surprisingly near the league median when the lead is by three or more goals. Then again, look at the TOI, the Devils do not often lead by three or more goals. Or even two goals. Still, opposing teams perform as expected to the Devils when they're losing by one or two goals and the Devils' response is mostly to take it as it comes, go for clears even more than trying to get some offense going to keep the opposition honest, and hope the goalie stops them all.
That's the case at 5-on-5. What about all situations?
|Leading by 1||536.5||40.5||26||39.3||57.7||40.4||30||22.5||33.2||97.0||29||23||20||11.4||93.3|
|Leading by 2||278.7||33.3||30||34.9||69.7||31.5||30||16.8||36.6||104.6||25||10||12||12.8||92.9|
|Leading by 3+||140.6||37.8||20||33.7||55.5||39.8||19||20.1||30.3||89.2||30||5||8||10.6||88.7|
It's about the same. The numbers are, well, they're not better. The opposition's shooting and attempts rates are higher. The Devils' CF% and SF% are even worse. I will say CP60 is not always dead last in these situations, it actually got out of the bottom five when the Devils had a two goal lead. Of course, that's largely because they're allowing nearly 70 attempts per sixty minutes in that situation. The save percentages are a bit lower; I suspect that's a result of special teams and empty net situations since they were much higher in 5-on-5 play.
I think it's clear across the board that the Devils are just not a good offensive team. They're on the wrong side of possession in tied situations and in all situations in 5-on-5 play. They get above it when they're down by two goals in 5-on-5 play, but even then, they lag behind most the league. The shooting differential percentage may look a little nicer, but the team is still all about a low shooting rate while trying to keep their opposition below them. Something that definitely doesn't happen when the Devils do have a lead; the opposition just fires away a lot more often, forcing New Jersey to be even less capable to attack and rely a lot on goaltending to preserve that lead.
The big surprise for me was that the possession and shooting rates do improve when the Devils are trailing by one or two goals. It means that, this season, there is a response in favor of the score. The issue is that the response is rather low. Combined with a low shooting percentage, that one or two goal deficit often remains at one or two goals. To put it another way, it wouldn't be right to say that the Devils never play to the score when they're losing. It's more accurate to state that while does happen from time to time, the Devils just attack a little bit more when they're losing by one or two. And that little bit more is rarely enough. So feel free to still gnash your teeth when they come up lame despite having oodles of time to make up a deficit that really doesn't faze most other teams.
How can this be fixed? Well, I think the main answer is the same one that many fans - myself included - have demanded for a while. The team clearly needs more offensive talent. Much more. And it's not just "Go out and get players with points." They need more players that can make and receive passes in stride. They need more players that can make decisions with the puck instead of just robotically dumping pucks away or settling for long shot hopes through traffic. They even need to look at the players that they do have and determine whether they have that talent (keep them) or not (move/dump them). And we know that can't be done immediately. Getting talent, the right talent, takes time. And when that talent does come, we should see a Devils team that attempts more shots, gets more shots, and plays much better in light of the score instead of just relying on low-event hockey. That'll be the day. It won't be today or this season.
What do you make of all of this? What have you learned about the 2015-16 Devils and score effects? Are you surprised at the numbers, or was this what you expected? Is getting more talent really the key answer, or is there more to it than that? Please leave your answers and other thoughts about this season's team and how they play to the various scoring situations in the comments. Thanks to War on Ice for the numbers - which, again, are before the Washington game - and you for reading.