clock menu more-arrow no yes

Filed under:

The Oilers in Different Score-States

NHL: Edmonton Oilers at Calgary Flames Sergei Belski-USA TODAY Sports

The Edmonton Oilers are sitting in a great spot right now as the club has an even-strength (5v5) goal-share of 52.28%, having outscored opponents 102-94, which ranks them 11th in the league, and 5th in the Western conference. The Oilers PDO isn’t anything out of the ordinary as the team is relying on league average goaltending and has had a normal team save percentage over the course of the year. This is good news and suggests that the team's goal production should be sustainable, as long as their top players, including McDavid, Talbot and Sekera, stay healthy.

What's interesting is at what score-state (i.e., tie game, Oilers trailing, Oilers leading) the Oilers are scoring their goals. According to Travis Yost’s analysis over at TSN, the Oilers are a much more dominant team when they're trailing in a game, as they run at a goal-scoring rate of 3.11 goals per hour, and have outscored opponents 42-21, which translates into the second best goal-share in the league with 66.67%. It's definitely encouraging to see that the team has the ability to turn up the offence, however it's a bit alarming that  the Oilers play so poorly when the score has been tied this season.

Score State (5v5) Goals For% Goals For/60 Corsi For% Corsi For/60
All Scores 52.28% 2.35 50.52% 55.51
Tied 38.75% 1.74 48.58% 54.78
Trailing 66.67% 3.11 59.18% 62.93
Leading 55.56% 2.42 44.22% 48.47

With the score tied, the Oilers have a 48.58% share of the total shot attempts (Corsi For%), which, along with their 54.78 shot attempts per hour, ranks them 22nd in the league. And in this score state, the Oilers have been outscored 49-31, which translates into the second worst goal share of 38.75%, only ahead of the Colorado Avalanche. Their 1.74 goals per 60 ranks them 26th in the league.

To get a sense of what might be going on here, it's important to look at how the Oilers do with and without McDavid, who has been the team's offensive driver all season. My initial thought was when the score has been tied, the team still does great with McDavid and does really poorly without him. But if we split out the numbers, we see that's not exactly the case.

First, here's how the team does with and without their captain across all score-states at even-strength. The team has had an obvious depth issue as the Oilers can run at an elite rate when McDavid is on the ice, but struggle to  generate shots and goals without him.

All Scores Goals For% Goals For/60 Corsi For% Corsi For/60
With McDavid 59.21 3.11 53.99 61.24
Without McDavid 47.50 1.96 48.78 52.80

Now digging into the various score states, here's how the Oilers have done when the score is tied, with and without McDavid. The team as a whole takes a hit, as reflected in not only McDavid's off-ice numbers, but his on-ice numbers as well. This indicates that the coaching staff is doing something, either through different tactics or deployment, that's impacting the production of all of it's players.

Score Tied (5v5) Goals For% Goals For/60 Corsi For% Corsi For/60
With McDavid 50.00 2.52 51.83 61.98
Without McDavid 32.00 1.35 46.81 51.18

And we can also confirm the coaching staff's influence by looking at how the team does when the team trails. Again, we see that McDavid's on ice numbers go up, but so to do the off-ice numbers, indicating his teammates are able to produce at a very respectable level. A shot share of over 55% with and without McDavid is fantastic.

Oilers Trailing (5v5) Goals For% Goals For/60 Corsi For% Corsi For/60
With McDavid 67.86 4.01 63.11 69.69
Without McDavid 65.71 2.63 56.89 59.35

Finally, here's how the team does with and without McDavid when the team is leading at 5v5. Their 46.06% is right around the league average for this score particular score state, as teams defending a league tend to play more conservatively and don't often generate the same level of offence as when the score is tied or when they're trailing.

Oilers Leading (5v5) Goals For% Goals For/60 Corsi For% Corsi For/60
With McDavid 61.11 2.91 46.06 49.49
Without McDavid 51.43 2.12 43.62 48.31

My key takeaway from all of this is that the coaching staff is doing something when the score is tied that's bringing everyone down, including McDavid. And when the team is trailing, they deploy McDavid and the top line more often, as any team would, but the coaching staff is doing something that is giving the rest of the roster a much needed boost as well.

And part of that something is managing and deploying their defencemen appropriately.

This season, the Oilers have used Klefbom, Larsson, Sekera and Russell as their top four defencemen and have distributed the ice time pretty evenly across the top two pairings. But what we can unover from the available data is that the Oilers make slight adjustments to the distribution of ice time depending on what the score is in the game.

What I've done below is break out each of the Oilers defencemen's proportion of the available ice time by score-state. So starting with the data for score-tied (middle of the graph),  we see that each of the Oilers top four defencemen gets between 30% and 35% of the available ice time.

But when the Oilers start trailing and are looking to generate more offence, the team continues to roll out Larsson, Klebom and Sekera at about the same rate as when the score is tied, but they do scale back Russell's proportion of ice time from 31.27% to 27.96%. That's a difference of about 28 minutes of ice time. In his 25 games before he got hurt, Nurse would see a jump in his ice time proportion when the team trailed compared to when they were tied. And over the course of the season, Benning has seen his proportion of ice time go from 22.60% when the score is tied up to 25.42% when the Oilers are trailing, which translates into about 26 more minutes of ice time. It appears that making these adjustments has been part of the reason why they begin to dominate play when trailing.

The overall issue here is not Russell as a player, but more of how the Oilers coaching staff is approaching the game when the score is tied. Rather than push for offence, they’re playing a far more reserved game and deploying their defencemen accordingly. They've obviously identified Russell as being more of a defensive specialist for them, deploying him less frequently when they need offence and more often when the team is protecting a lead. And they're leveraging some of the the skill they have on defence by deploying guys like Benning, and earlier in the season Nurse, more frequently when they need a goal.

A 48.58% share of the shot attempts and a 29th ranked goal-share of 38.75% has to be addressed for this team to remain competitive down the stretch. The good news is that the team does have the resources to push for more offence instead of playing a safer game and improve their numbers, as Benning has shown that he can produce in a top four role, especially when he’s been paired with Sekera this season. Comparing Benning's performance with Russell's when the score is tied, the team has generated more offence with the rookie on the ice and has outplayed the opponent. Unfortunately, the team's shooting percentage drops considerably when Benning is on  the ice, which we can expect to eventually regress towards the mean (additional detail in Appendix A). The ideal situation would be for Benning to get more ice time and be paired with Sekera, which would positively impact the team's rate of shot attempts and scoring chances. This does mean that Russell would likely play lower down the depth chart and have his ice-time reduced, but that would actually place him at his established NHL level where he could potentially have more success.

Data: Corsica Hockey

Appendix A: Russell vs Benning, 2016/17, Score-tied

Metric Russell Benning
TOI 334.58 241.79
Corsi For% 42.34 51.12
Corsi For/60 46.09 56.58
Corsi Against/60 62.77 54.10
Goals For% 50.00 40.00
Goals For/60 2.33 1.49
Goals Against/60 2.33 2.23
On-ice Sh% 9.56 4.20
On-ice Sv% 92.61 92.37
PDO 102.17 96.60