Talking Oilers, defence pairings and trade deadline on The Lowdown with Lowetide (TSN 1260)

Joined Lowetide this morning on TSN 1260 to talk about the past week, including the overtime win against the Flames. We touched on the potential defence pairings, with and without Larsson, and what areas of the roster the team needs to focus on as the trade deadline approaches. Big week ahead with three home games against Arizona, Florida and Nashville, and a Saturday night tilt in Calgary.

Segment starts around the 20 minute mark.

Below are the various defence combinations that the Oilers have rolled with this season at 5v5, sorted by Corsi For% (score and venue adjusted). If the Oilers insist on playing Russell, they have to pair him with Sekera as that’s who he’s best with. Klefbom-Benning looks like a good tandem so far. And I’m also a fan of Gryba-Davidson as the third pair. (Data: Corsica Hockey. Ideal pairings for me: Klefbom-Larsson, Sekera-Benning, Davidson/Gryba/Nurse.


This past weekend, I also looked into how the OIlers are doing when it comes to generating shots and what the key drivers/drags have been: Volume Shooting – The Copper & Blue (2017, January 14)

And just a heads up that I’ll be on the CBC Edmonton News to chat about the Oilers and the playoff race on Tuesday night. Following that segment, we’ll be doing another Facebook Live session on the CBC Edmonton Facebook page to talk Oilers and take questions from the community.

Volume Shooting

Early in his first season as head coach of the Edmonton Oilers, Todd McLellan emphasized the value of volume shooting, and its importance in generating offence.

Volume shooting, I don’t know what that does to Corsi or Fenwick because I don’t even know what those things are, but volume shooting is important. I think it breaks down defensive zone coverages, gets players out of position, taxes the opposition, makes them play more minutes in their zone. (Source)

Taking a look at the rate of shot attempts (i.e. Corsi For/60) the Sharks generated when McLellan was behind the bench, we see that they were always above the league average and typically ranked in the top five.

Season Corsi For/60 League Rank
2008/09 57.87 7th
2009/10 58.80 6th
2010/11 61.91 1st
2011/12 60.22 5th
2012/13 59.74 5th
2013/14 64.78 1st
2014/15 60.60 5th

It appears that the Oilers have gradually made progress when it comes to generating shot attempts under McLellan, as they currently rank 12th in the league, 5th in the Western Conference, with 57.18 shot attempts per hour at even-strength. The top five teams: Boston, Toronto, Montreal, Pittsburgh and Washington.


As you might guess, one of the key drivers for the team’s rate of shot attempts includes Connor McDavid. When he’s been on the ice this season, the team has generated 62.66 shot attempts per hour, which is just below what Boston, who ranks first in the league in this metric, is generating. Without McDavid, the Oilers generate 54.55 shot attempts per hour, which is below league average, and would rank them 19th in the league. Also worth noting that the Oilers top line of McDavid, Maroon and Draisaitl is currently generating 71.12 shot attempts per hour.

If we break out the Oilers rate of shot generation over rolling 10-game segments, we see that they had at one point been generating over 60 shot attempts per hour, but steadily declined starting around the end of November. As I mentioned in my previous article, I suspect this has to do with two things. One, the team lost Darnell Nurse, who was showing progress in his offensive game, to a long term injury at the end of November. And two, the team began giving more and more ice time to Kris Russell, who provides very little to a team’s offence. More on individual players later.


In the graph above I have the team’s rate of shot attempts, but I’ve also added two additional lines: one for the rolling 10-game averages of when McDavid is on the ice (orange line), and one for the rolling 10-game averages of when McDavid is off (blue line). The team is having issues this season where they’re relying heavily on one line, more so than other teams with elite players. Knowing his ability to escalate the play of his team, and the importance of having depth to win a cup, we’ll need to know how the rest of the roster is doing without McDavid on the ice.

What we can start to do is look at each player this season, and how the team does when it comes to generating shots with and without them on the ice. I’ve ranked the table below by Corsi For/60 Rel, which tells us how the team does with the player on the ice, compared to how the team does when they’re on the bench. So when Patrick Maroon is on the ice, the Oilers generate 65.65 shot attempts per hour. Without him, that number drops by 11.77 shot attempts.

Full article is at The Copper & Blue.

Couple Thoughts on Benoit Pouliot

benoit-pouliotWithout question, the 2016-17 season has been a terrible one for winger Benoit Pouliot. In 443 minutes this season, Pouliot has scored 5 goals and added 3 assists (all of which have been primary) at even strength, which ranks him 10th on the team among 15 forwards who have played at least 50 minutes this season. His 1.08 points per hour (P/60) is well below his career average coming into this season, ranks him dead last among the 15 forwards.

Player GP TOI G A P P/60
CONNOR.MCDAVID 43 678.51 10 20 30 2.65
TYLER.PITLICK 31 291.26 8 3 11 2.27
ANTON.LANDER 20 143.49 1 4 5 2.09
PATRICK.MAROON 43 572.52 15 4 19 1.99
LEON.DRAISAITL 43 574.32 5 12 17 1.78
MARK.LETESTU 40 360.13 3 7 10 1.67
JESSE.PULJUJARVI 28 286.12 0 7 7 1.47
ANTON.SLEPYSHEV 15 164.79 2 2 4 1.46
RYAN.NUGENT-HOPKINS 43 547.87 3 10 13 1.42
ZACK.KASSIAN 40 427.14 2 8 10 1.40
JORDAN.EBERLE 43 597.46 3 10 13 1.31
MILAN.LUCIC 43 609.46 5 7 12 1.18
DRAKE.CAGGIULA 25 266.27 2 3 5 1.13
MATT.HENDRICKS 18 161.85 1 2 3 1.11
BENOIT.POULIOT 38 442.72 5 3 8 1.08


If we look at Pouliot’s relative-to-team stats, we see that across the five metrics (see Appendix for definitions), the teams that he’s been on have often done better with him on the ice than without him, with the current season being an exception.  This ability to be a positive influence is likely what drew the Oilers to him when he hit free agency in the summer of 2014. The Oilers at the time desperately needed an experienced winger who could support the developing core at the time, and made the right call signing a versatile forward like Pouliot while he was still in his prime.


Unfortunately for Pouliot, the team has done better without him on the ice this season, which is surprising considering his history of being a positive influence. This season, the team has been outscored 13-15 with Pouliot on the ice, but has maintained a 50% share of all of the shot attempts with him. The issue for Pouliot has been the penalties he has taken, as he’s tied for the worst penalty differential with Kassian, sitting at -6 (i.e., he’s taken 12, but only drawn 6).

We also have to keep in mind here the McDavid effect when assessing the 2016-17 rel stats above. When McDavid is not on the ice, the Oilers see a drop in production as they have a  49.98% share of the shot attempts, and 45.54% of the goals. Pouliot has only been on the ice for 20 minutes with McDavid this season, as Maroon and Lucic have been the regular left wingers, so it’s fair to compare how Pouliot’s on-ice numbers compared to McDavid’s off ice numbers. I’d be a lot more concerned with Pouliot’s production if his on-ice numbers were lower than the McDavid off-ice numbers.

Metric McDavid off ice Pouliot on ice
CF% 49.98 50.00
FF% 49.86 50.74
SCF% 44.71 45.86
xGF% 45.89 45.28
GF% 45.54 46.43

What I do wonder is what impact the coaching staff’s constant line juggling this is having on players including Pouliot. What’s become pretty obvious is that McLellan bases a lot of his lineup decisions on actual results – specifically goals. That’s completely understandable if you’re trying to ride a hot hand when a combination of players are scoring, but you’re also quick to break up what could be a good line if they’re not scoring. To assess what could be a successful line combination in the future, the share of shot attempts (Corsi or Fenwick), and the expected goal share (xGF%), the more predictive metrics, needs to be reviewed. Relying solely on goal-share, while understandable, is risky as goal production is often influenced by more random factors, as players (shooting percentage) and goalies (save percentage) can run hot and cold through a season.

Couple issues I see for Pouliot. One, he hasn’t had consistent linemates this year, as he’s been constantly shuffled around the bottom nine, sometimes with Caggiula, sometimes with RNH, and sometimes in the press box. In my opinion, he should be lined up with RNH at any given moment, as they could be relied on to match up against the other team’s top lines. RNH has struggled mightily this season, but if he can be paired with Pouliot and Eberle for a long stretch of games, I think the team can benefit. Keep in mind that the trio posted a 53% Corsi For percentage together coming into this season and that was not in any sort of sheltered role. They did get outscored 26-28, but again we have to put that into context as the goaltending was pretty bad last season and impacted their goal-share.

The other issue I see is how quickly Pouliot has been punished for taking penalties. Again, this is understandable from a coaches perspective as these penalties have been in the offensive zone and have lead to a powerplay goal against. Unfortunately for the Oilers, playing on the edge is a part of Pouliot’s game and it’s made him an effective NHL player. Yes, penalties do get called against him, but more often than not in previous years his fore-checking and aggressive, often risky, play is what’s lead to turnovers and shots on goal. Punishing a player who has relied on this type of play can easily have a detrimental effect as they may hesitate, even for a split second, to engage along the boards in an attempt to create offence.

One last thing – the Oilers need to roll out Pouliot on the penalty kill as often as they can. The club is still allowing a higher than average number of unblocked shot attempts against (68.60 per hour). But when Pouliot is on the ice, the rate of shots against drops to 61.13 per hour, fourth lowest on the team among all players  who have played at least 30 minutes on the penalty kill.

Data: Corsica Hockey

Appendix – Definition of metrics

  • Corsi For% (CF%) – The proportion of all the shot attempts the team generated and allowed that the Oilers generated (i.e., Corsi For/(Corsi For + Corsi Against). This is used as a proxy for possession and can predict a team’s future share of goals.
  • Fenwick For% (FF%) – The proportion of all the unblocked shot attempts the team generated and allowed that the Oilers generated (i.e., Fenwick For/(Fenwick For + Fenwick Against). This is used as a proxy for shot quality and considers shot blocking a repeatable skill. It can also predict a team’s future share of goals, slightlty better than Corsi.
  • Scoring Chances For% (SCF%) – The proportion of all the scoring chances (as defined by Corsica Hockey) that the team generated and allowed that the Oilers generated (i.e., Scoring Chances For/(Scoring Chances For + Scoring Chances Against),
  • Expected Goals For% (xGF%) – This is a weighting placed on every unblocked shot based on the probability of the shot becoming a goal. This depends on the type of shot, location and uses historical shot and goals data to come up with the probability for each unblocked shot. This has been found to be a better predictor of future goals than Corsi and Fenwick. (Detailed explanation can be found at Corsica Hockey)
  • Goals For% (GF%) – The proportion of all the goals that the team scored and allowed that the Oilers generated (i.e., Goals For/(Goals For + Goals Against).


The Edmonton Oilers With and Without McDavid (Part III)

An area that the Oilers will need to address going forward is their scoring production when McDavid is not on the ice. The young captain is the offensive catalyst that drives play and elevates his team when he steps on the ice. But when he’s on the bench, there’s been a significant drop in the team’s goal-share at even-strength.

Heading into Tuesday’s night game against the Sharks, the Oilers had outscored their opponents 36-22 at even-strength with McDavid on the ice, which translates into a 62.07% goal share. Without him, the Oilers have been outscored 43-51, a goal-share of 45.74%. The Oilers do appear to have an okay proportion of the shot attempts (Corsi For%) without McDavid as they have a 50.51% share without him, and a 54.71% share with him. Corsi provides value here as it serves as a proxy for possession and predicts a team’s future goal share. When we look at the expected goals for percentage, which measures the quality of the shots generated and also predicts future goal share, the Oilers drop from a 56.25% share with McDavid on the ice to 45.71% share without him.

If we look at the rolling 10 game averages of the different metrics with and without McDavid, we start to see a pattern that should be of some concern for the team.

First up is a graph with the Oilers shot-share broken up into three lines: the team’s goal share with McDavid (blue), the team’s goal-share without McDavid (orange) and the team’s overall goal-share (black) (with and without McDavid).


Full article is at The Copper & Blue.

Talking Oilers on The Lowdown with Lowetide (TSN 1260)

Joined Lowetide this morning on TSN 1260 to talk all things Oilers, including the demotion of Puljujaarvi and Gustavvson to the AHL. Full clip is below, starting around the 40 minute mark.

Couple notes:

  • The new top line of Maroon-McDavid-Draisaitl is absolute gold right now. In 66 minutes together so far this season, they’ve outscored the opponent 6-1 at 5v5 and have a 67% CF% (adjusted). This line should stick together, but also allow for Draisaitl to center another line as needed.
  • Another line I’d like to see get an extended look this season is Pouliot-RNH-Eberle. Heading into this season, they were a 53% CF line, but were outscord 26-28. Part of that was the team’s shoddy goaltending (89% save percentage). It obviously wouldn’t be a popular move, considering the funk all three have been in. But I’d like to see the coaching staff have more patience with their forwards, as there’s been very little stability in terms of line combinations all season. I’m convinced the head coach doesn’t like the roster he has.


Checking in on the Oilers Pacific Division Rivals

Forty games into the 2016/17 season, and the Oilers are holding a playoff spot, ranking third in the Pacific with 47 points, good for fifth in the Western conference. The team is coming off of a big win against a good Boston club last night, with Maroon, McDavid and Talbot leading the way. Up next is a back-to-back set against the Devils, who will have played the night before, and the Senators.

With the team in a playoff race now, it’s worth checking in on how the Oilers Pacific division rivals are doing in terms of goals and the shot metrics that predict future goals. The three California teams are going to be the ones to beat, but if the Oilers can get consistent even-strength scoring from the lines that don’t feature McDavid, they could potentially challenge for that third spot.

Here’s what the Pacific Division looks like today. I’ve included each team’s record, along with their points percentage, which is the points divided by the total points available (Source: Hockey Reference).

Team Games Record PTS PTS% GF% CF% XGF%
San Jose Sharks 39 23-14-2 48 0.615 52.98 52.03 53.07
Anaheim Ducks 40 20-12-8 48 0.600 43.16 44.72 41.48
Edmonton Oilers 40 20-13-7 47 0.588 52.38 51.10 49.00
Calgary Flames 40 21-17-2 44 0.550 47.16 49.59 47.14
Los Angeles Kings 39 19-16-4 42 0.538 50.55 54.07 53.03
Vancouver Canucks 40 19-18-3 41 0.513 46.05 47.63 46.57
Arizona Coyotes 38 11-22-5 27 0.355 49.17 49.45 51.51

What we can also do is look at how each team is doing over 10-game, rolling segments this season. What we do here is take games 1-10, find the average of the metric, then take games 2-11, find the average, games 3-12, and so on. Doing so can highlight trends and can give us a better perspective of how a team’s entire season has been going.

I’ve focused on even-strength play (5v5), as it evaluates a team in a natural game setting where coaches are rolling their lines, trying to score and defend, and is where the majority of the game is played. These numbers are also score adjusted, based on Corsica Hockey’s methodology. This way, we eliminate score-effects, as team’s that trail in a game start to take more risks and generate more shots, while team’s that lead in a game tend to play safer, and generate far fewer shots. Without adjusting, team’s that trail a lot, like the OIlers have in the past, start posting better shot-shares, for example, mainly because they’re taking more risks trying to tie a game. So adjusting for score-effects puts more emphasis and value on the events that occur in close game situations and gives us a more accurate assessment of a team. More on adjusted metrics can be found at Corsica Hockey.


Full article is at The Copper & Blue.

Talking Oilers on the CBC Edmonton News (TV) + Facebook Live


Had a fun experience today talking Oilers with Adrienne Pan on Facebook Live over the lunch hour. We talked about the team, their playoff aspirations, and took questions from the CBC Edmonton Facebook community. A big, big thank you to everyone who sent questions in!

Also, a big shout out to the team at CBC for putting it all together. We’re looking into doing more in the future.

Below is the full session via Facebook.


I also made my bi-weekly appearance on the CBC Edmonton news in the evening. Link is here: (clip starts around the 8:55 mark).

Backup link