After 20 games, the key drivers and areas worth monitoring


The Edmonton Oilers are first place in the Pacific division with a 0.650 points percentage (+7 goal differential, 3rd in the division), leading many including myself to question if this team is real or not. We’re 20 games into the season, so it’s worth looking into what the drivers have been for the Oilers success and try to uncover any issues worth monitoring.

Starting with goaltending, which has been good so far. The Smith/Koskinen tandem has posted a team save percentage of 91.27 (all situations), good for tenth in the league. At even-strength, the Oilers’ team save percentage ranks 15th in the league with 92.21% – right around where we should expect them to be. Among 50 goalies who have played at least 250 minutes at even-strength (5v5), Koskinen ranks 8th in terms of save percentage with 93.6%, while Smith ranks 39th with 90.7% (average is 91.7%). Hopefully, Oilers management is seeing the benefits of splitting the workload between two capable netminders and refrain from overspending next summer on the goaltending position.

The powerplay has been dynamite, scoring 10.66 goals per hour (second best in the league). The Oilers are relying on their first unit, similar to what they did last season, loading up with their high end talent and getting major contributions from secondary options like James Neal. While the team can expect to see their shooting percentage regress towards league average rates (13.2% over the last three seasons), it’s at least encouraging to see how great they’ve been consistently generating shot attempts (CF/60) and unblocked shot attempts (FF/60, a proxy for scoring chances). And the team’s rate of expected goals (xGF/60), which uses historical data to give each shot a probability of becoming a goal, reflects their ability to generate higher quality chances.

CF/60 FF/60 xGF/60 GF/60 Sh%
114.56 (3rd) 81.92 (6th) 7.78 (3rd) 10.66 (2nd) 18.6% (2nd)

With the team’s powerplay being this strong, it sure would be nice if the Oilers had the personnel to draw penalties more frequently. The Oilers have the third lowest rate of powerplay minutes per game – something that could be addressed if either the team played with more pace or if they had someone who had that speed and skill-set.

The Oilers penalty kill has also been a key driver for their success as they currently rank 3rd in the league allowing only 4.54 goals against per hour. It appears they’re relying quite heavily on their goaltending, which ranks third in the league, as they haven’t been very good at limiting shots and chances against.

CA/60 FA/60 xGA/60 GA/60 Sv%
97.34 (17th) 75.15 (19th) 7.55 (27th) 4.54 (3rd) 91.74 (3rd)

What’s interesting is that the Oilers have in fact improved over their last ten games, giving some hope that the penalty kill results may be sustainable. Below is a split of how the team has done this season.

Oilers CA/60 FA/60 xGA/60 GA/60 Sv%
Games 1-10 116.3 93.25 10.4 4.19 93.94
Games 11-20 79.74 58.35 4.9 4.86 88.37

Starting with the goals against per hour (GA/60), we see that the Oilers have been consistent allowing about four goals per hour, even though their goaltenders performances have slipped. Thankfully, the Oilers have cut their expected goals in half, indicating that they’ve done a better job at limiting the high quality chances. Worth noting that these numbers from the most recent ten games has them among the leagues best. I’d be curious to know what the coaching staff has done differently tactic-wise that would’ve impacted the results and underlying numbers.

Thinking through this more, I probably shouldn’t be too surprised that the team’s penalty kill figured things out defensively. They have been good all season limiting shots and chances at even-strength (5v5), ranking top ten in the league in various metrics.

A big reason why things have gone well is the emergence of Ethan Bear and how well he’s done playing top minutes against the best competition. Looking at how each defencemen’s on-ice Corsi For% relative to the team, I was surprised to see him leading the way. With Larsson still recovering and Caleb Jones being called up, the Oilers could have some big decisions coming up including how to handle Russell and Nurse. If Jones emerges as a legitimate NHL player, the Oilers left-side is much more solid and gives Oilers management some leverage in their negotiations with Nurse.

Should note that the reason I use Corsi Rel is because of the predictive value of Corsi For% and it’s relationship to goal-share, which I touch on later in the post. It’s not meant to be a definitive metric, but it gives us a sense of what’s happening and can drive further questions and analysis.

20191113 - Defencemen CorsiRel

As great as those things have been, there are definitely issues worth monitoring.

The first issue is the team’s underlying shot-share numbers at even-strength, which are a good predictor of future outcomes. The Oilers currently rank 23rd in the league with a 48.62% Corsi-for percentage (a proxy for possession) and 20th with a 48.93% Fenwick-for percentage (a proxy for scoring chances). These are numbers that a coaching staff can influence through tactics and how they deploy the roster. But so far, it looks like it’s only getting worse.

20191113 - CF Rolling 10.png

The poor Corsi For percentage at even-strength is largely driven by the Oilers inability to generate offensive opportunities, as they rank 26th in terms of shot attempts per hour with 51.23 and 28th when it comes to unblocked shot attempts (i.e., scoring chances) with 37.79. If the Oilers hope to improve their rate of 2.40 goals per hour (17th overall), they’ll need to figure out how to carry the play and spend more time in the offensive zone. It shouldn’t be too much to ask for the team to post shot-share metrics above 50%. But considering the roster construction, we probably shouldn’t be surprised.

That leads me to my next point.

What’s especially troubling is that it’s not just the depth players dragging down the team’s overall performance – it’s the star players as well who are getting out-shot and out-chanced when they’re on the ice.

Below are the Oilers forwards who have played at least 65 minutes at even-strength this season and their Corsi For% relative to the team.

20191113 - Forwards CorsiRel

The key focus here is on McDavid and Draisaitl, who have no doubt been outstanding so far this season. But it’s definitely concerning to see how poorly they rank within a team that’s posting poor shot-share numbers. McDavid specifically, as crazy as it might sound, hasn’t looked 100%, perhaps because of the injury from last season or the fact that he and Leon have been leading the league in ice-time per game.

I usually save these graphs for when I look at depth players and their history relative to their teammates. But I had to see just how far off McDavid has been from his historical numbers. I suspect last season had to do with Hitchcock’s coaching tactics and being overplayed. I’d be curious to know if similar impacts combined with his injury recovery are driving his numbers down like this.

20191113 - McDavid Rel

The fact that a player like McDavid is posting these numbers should shine a light on management and the coaching staff. Yes the player is posting excellent results and the team is doing well. But is there something the coaching staff could do differently to get their star players more scoring opportunities. Are they aware that their even-strength results (i.e.,  goal-share) may not be sustainable? And is there something management could do to mitigate the risk of injuring a star player who is critical for the long-term success of the franchise?

It’s been a great start to the season, and the team has greatly improved their chances of making the playoffs. But there remains major issues both with the roster construction up front and the coaching staff’s offensive tactics that has me questioning the teams ability to keep pace in a competitive division.

Data: Natural Stat Trick

One thought on “After 20 games, the key drivers and areas worth monitoring

  1. Pingback: Running operations | The SuperFan

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s