CBC Edmonton News (TV): Recent success, depth players and the GM search

cbc edmonton logoI joined host Lindsay Highmoor on the CBC Edmonton News for my weekly television segment to discuss all things Oilers. Clip is here and starts at the 16:50 mark: CBC Edmonton News (2019, March 28)

Topics we covered:

  • Re-cap of Tuesday night’s 8-4 win against the Kings and the star players leading the way.
  • The Oilers depth issues, and the importance of finding reliable talent on value deals for next season.
  • The Oilers overall numbers at even-strength (5v5) this season
    • Goals/60 – 2.21 (26th)
    • Shots/60 – 28.44 (26th)
    • Goals against/60 – 2.59 (21st)
    • Shots against/60 – 31.34 (20th)
  • Special team results:
    • Powerplay – Goals/60 – 8.06 (7th)
    • Penalty kill  – Goals against/60 – 9.44 (30th)
  • The Oilers search for a general manager, and the importance of collecting as much information and insight as possible before hiring someone.
  • Preview of tonight’s game against the Dallas Stars.

Data: Natural Stat Trick

Advertisements

CBC Radio Active: The countdown is on to end of the Oilers season

cbc edmonton logoI joined host Rod Kurtz on CBC Radio Active to talk all things Oilers. Full segment is here: CBC Radio Active (2019, March 21)

Topics we covered:

  • The season as a whole and the key factors for yet another season without playoffs.
  • Bob Nicholson’s search for a GM, and his missteps and role in the Chiarelli era.
  • The importance of leveraging analytics as part of the larger strategy around decision-making.

Trending downwards

20765233

With the Oilers currently on a decent run, going 6-3-1 in their last 10 games, it’ll be interesting to see how their recent success will impact the management group’s decision-making process going forward.

The reality of the situation is that the Oilers are a bad team, and have been for most of the season. They currently rank 23rd in the league with 69 points, posting an overall goal-differential of -31. At even-strength (5v5), the Oilers have posted a 47.78% Corsi For percentage, a proxy for possession, and a 47.74% Fenwick For percentage, a proxy for scoring chances – both ranking the Oilers 23rd in the league and one of the worst in the western conference. And while the powerplay has been productive, scoring 7.82 goals per hour (10th in the league), the penalty kill continues to be pathetic, allowing the second highest rate of goals against with 9.14 per hour.

So while the Oilers haven’t been mathematically been eliminated, it’s been fairly obvious for most of the season that this team isn’t good enough to contend for a playoff spot, let alone championship.

Hopefully the Oilers management is aware of the fact that even though their results over their last ten games have been positive – at one point being five points out of a wild card spot – the  team has posted only a +1 goal differential (all situations), and have been out-scored 19-17 at even-strength – a goal-share of 47.22%. What’s especially concerning is the team’s underlying shot-share metrics over their last ten games:

Metric Games 61-70
Corsi For% 45.15%
Fenwick For% 44.98%
Shots For% 46.44%
Scoring Chances For% 45.29%
High Danger Scoring Chances For% 46.74%

Those are shot-shares that would put the Oilers in the bottom five of the league, with teams like Detroit, New Jersey, New York Rangers, Anaheim and Ottawa. It’s a good indication that their recent results are not sustainable, and that to continue having success they would need their goaltending to remain top five in the league. And that’s not exactly a reasonable expectation considering team save percentages often fluctuate and goaltender performances tend to regress towards their career averages.

It’s worth noting that in the ten games prior to this recent stretch, which started right after Peter Chiarelli was dismissed, the Oilers posted their best 10-game stretch when it came to shot-share metrics, but couldn’t get results going 2-5-3 and getting outscored 37-25.

Metric Games 51-60
Corsi For% 50.77%
Fenwick For% 51.33%
Shots For% 50.36%
Scoring Chances For% 53.38%
High Danger Scoring Chances For% 47.87%

Not the greatest numbers, as they barely broke even when it came to most metrics. But these are the best the Oilers did since Hitchcock’s arrival. Over this stretch, the Oilers were doing well suppressing offence against at even-strength (5v5), allowing 38.1 unblocked shot attempts (Fenwick) against per hour and 29.24 shots on goal against per hour. Unfortunately, their goaltending was slightly below average, posting a team save percentage of 91.06%. Offensively, they were ice cold not only at scoring goals, but also generating shots and scoring chances – an issue that has plagued them all season.

But as we see in the graph below, things got progressively worse over the most recent ten games, especially when it came to the defensive side of things. The team went from allowing 38.1 unblocked shot attempts (Fenwick) per hour to 47.80 per hour – a significant jump going from one of the best in the league to one of the worst. Note, the black line represents the league average rate of unblocked shot attempts for and against this season (42.2).

20190314 - Rolling Fenwick Rates - 10

A couple things.

One, what exactly did the Oilers do that caused this spike in the rate of unblocked shot attempts against? Was there something about their results in games 51-60 that made the coaches adjust their tactical plans? There weren’t any significant changes in how often players were being deployed as the majority of them were playing around the same number and type of minutes across both 10-game segments. Another tell that the team is having deeper systemic issues is the fact that even McDavid posted poor on-ice numbers over this recent ten game stretch (48.61 CF%, 46.74 FF%, 43.75 SCF%).

Second, and more importantly, which segment are the Oilers going to take their lessons from: games 51 to 60 when they did not get results, but were playing well defensively, or games 61 to 70 where the results were there, but likely not sustainable considering how bad they’ve been possession-wise and how badly they’ve been bleeding shots and chances against.

The Oilers management group has to show courage and apply a critical lens on things regardless of the results, and seek a deeper understanding of not only why things happened but also if the results have been real and sustainable. It’s something we know the Oilers have failed to do in the past, but hopefully their approach changes this off-season considering the dire situation the franchise finds itself in.

Data: Natural Stat Trick

Appendix 1: Results over the last 20 games

Game number Match Result
51 2019-02-02 – Oilers 4, Flyers 5 L
52 2019-02-03 – Oilers 3, Canadiens 4 L
53 2019-02-05 – Blackhawks 6, Oilers 2 L
54 2019-02-07 – Oilers 4, Wild 1 W
55 2019-02-09 – Sharks 5, Oilers 2 L
56 2019-02-13 – Oilers 1, Penguins 3 L
57 2019-02-15 – Oilers 1, Hurricanes 3 L
58 2019-02-16 – Oilers 2, Islanders 5 L
59 2019-02-19 – Coyotes 3, Oilers 2 L
60 2019-02-21 – Islanders 3, Oilers 4 W
61 2019-02-23 – Ducks 1, Oilers 2 W
62 2019-02-25 – Oilers 2, Predators 3 L
63 2019-02-27 – Oilers 2, Maple Leafs 6 L
64 2019-02-28 – Oilers 4, Senators 2 W
65 2019-03-02 – Oilers 4, Blue Jackets 0 W
66 2019-03-04 – Oilers 4, Sabres 3 W
67 2019-03-07 – Canucks 2, Oilers 3 W
68 2019-03-09 – Maple Leafs 3, Oilers 2 L
69 2019-03-11 – Rangers 2, Oilers 3 W
70 2019-03-13 – Devils 6, Oilers 3 L

Appendix 2: Shot-share metrics definitions

  • 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, slightly better than Corsi.
  • Shots For% (SF%) – The proportion of all the shots on goal that the team generated and allowed that the Oilers generated (i.e., Shots For/(Shots For + Shots Against).
  • Scoring Chances For% (SCF%) – The proportion of all the scoring chances (as defined by Natural Stat Trick) that the team generated and allowed that the Oilers generated (i.e., Scoring Chances For/(Scoring Chances For + Scoring Chances Against),
  • High Danger Scoring Chances For% (HDCF%) – The proportion of all the high danger scoring chances (as defined by Natural Stat Trick) that the team generated and allowed that the Oilers generated (i.e., High Danger Scoring Chances For/(High Danger Scoring Chances For + High Danger Scoring Chances Against).

Appendix 3 – Edmonton Oilers, Fenwick/60 (5v5) with key events

Oilers Rolling FF - With Milestones - 20190315

CBC Edmonton News (TV): Loss to the Devils, recent ten games and the general manager search

cbc edmonton logoI joined host Min Dhariwal on the CBC Edmonton News for my weekly television segment to discuss all things Oilers. Clip is here and starts at the 16:00 mark: CBC Edmonton News (2019, March 14)

Topics we covered:

  • Last night’s loss to the Devils and the playoff aspirations. For me, it’s been obvious for a while that the Oilers aren’t playoff contenders.
  • The key drivers for their recent success, which has largely been goaltending. Over their last 10 games, the Oilers have six wins, but it’s due in large part to their team save percentage at even-strength (93.42%, 6th in the league). The team has posted a 45.15% Corsi For percentage (score-adjusted) over that stretch, a good indicator that this team has some major issues to address.
  • The penalty kill, which allowed another two goals last night, and continues to be one of the worst in the league. I mentioned a couple things the Oilers could look into doing, which I had also written about recently: Drivers and Anchors on the Penalty Kill – The SuperFan (2019, March 4)
  • The general manager search, and the importance of expanding their scope and looking for ideas from outside the industry. Related article here, which includes my previous work.

Data: Natural Stat Trick

CBC Edmonton News (TV): Oilers winning streak, Draisaitl’s performance and upcoming games

cbc edmonton logoI joined host Alicia Asquith on the CBC Edmonton News for my weekly television segment to discuss all things Oilers. Clip is here and starts at the 17:45 mark: CBC Edmonton News (2019, March 7)

Topics we covered:

  • The Oilers recent road trip, with three straight wins against Ottawa, Columbus and Buffalo.
  • The Oilers last 10 games, and some of the keys. Worth noting that the team was trending upwards in mid February, largely due to improved defensive play, but are now finally getting results thanks to improved netminding.
  • Leon Draisatl’s performance. Noted that his individual shooting percentage is at 20.0% at even-strength (Data: Natural Stat Trick). Well above his career average of 10.96% heading into the 2018/19 season – so definitely worth tempering expectations for next season.
  • Upcoming home games against Vancouver and Toronto.

Drivers and anchors on the penalty kill

15247148_web1_190123-RDA-Red-Wings-Oilers_1.jpg

In case you missed it, I recently dug into the Oilers recent powerplay success as well as their penalty killing failures.

Aside from seeing the powerplay shot generation and penalty kill shot suppression rates increase, oddly resembling one another, the other takeaway were the poor numbers being posted by defencemen Kris Russell when the team has been shorthanded this season. The team has allowed a team-worst 11.66 goals against per hour (GA/60) and 59.38 shots against per hour (SA/60) on the penalty kill with Russell on the ice.

Listed below are all of the Oilers defencemen who have played at least 15 minutes on the penalty kill this season, sorted by the rate of goals against per hour. Included is the time on ice (TOI), as well as the TOI%, which calculates the players proportion of the team’s total time on the penalty kill (328 minutes).

Player GP TOI TOI/GP TOI% SA/60 GA/60
Gravel 35 55.77 1.59 17.0% 53.80 5.38
Klefbom 44 83.10 1.89 25.3% 51.99 6.50
Larsson 65 137.75 2.12 42.0% 49.22 7.84
Jones 17 19.07 1.12 5.8% 47.20 9.44
Benning 53 67.45 1.27 20.6% 42.70 9.79
Nurse 65 133.20 2.05 40.6% 56.31 10.36
Garrison 17 16.87 0.99 5.1% 28.46 10.67
Russell 55 113.17 2.06 34.5% 59.38 11.66

Note that after 65 games, the Oilers are allowing 8.96 goals against per hour (29th in the league) on the penalty kill, and 51.95 shots against per hour (17th in the league). The league average rate of goals against per hour and shots against per hour this season is 7.16 and 52.0, respectively. As I wrote in my previous article, the Oilers rate of shots against has steadily increased over the course of the season, with the Oilers posting a shots against per hour rate of more than 60.0 over their last 20 games (red line in the graph below). That’s well over the league average rate and deserving of some attention.

Oilers - Special Teams - Shots per 60

Aside from Russell’s struggles, it’s worth looking into which other defencemen are impacting the team’s penalty kill numbers – both positively and negatively. To do this, I broke the season up into three segments: the first 31 games when Klefbom was in the line-up, the next 21 games with Klefbom out (and Russell only available for 10 games), and the next 13 games with Klefbom and Russell back. Losing both defencemen forced the team to change-up the powerplay and penalty kill combinations, giving us a little more insight into how the other defencemen did with more or less ice time.

2018/19 Season Games Team TOI SA/60 GA/60
Klefbom and Russell healthy 31 159.20 48.62 8.29
Klefbom injured, Russell healthy for 10 21 111.40 51.17 9.69
Klefbom  and Russell healthy again 13 57.38 62.74 9.41

Segment 1: Klefbom and Russell healthy (31 games)

In the first segment of 31 games, the Oilers did a pretty good job on the penalty kill limiting shots against (48.62 per hour, 13th in the league), but unfortunately their goaltending was one of the league’s worst, posting a team save percentage of 82.95 (24th in the league). It’s a trend that’s continued from last season when the team would go on decent stretches limiting shots and chances, but couldn’t get consistent goaltending.

Below are the Oilers defencemen who played on the penalty kill  in this first segment (sorted by TOI%), including their on-ice rate of shots against and goals against.

Player GP TOI TOI/GP TOI% SA/60 GA/60
Larsson 31 66.32 2.14 41.7% 48.86 6.33
Nurse 31 65.28 2.11 41.0% 56.06 9.19
Russell 31 65.27 2.11 41.0% 52.40 11.95
Klefbom 31 59.72 1.93 37.5% 44.21 5.02
Benning 26 30.50 1.17 19.2% 41.31 9.84
Gravel 16 20.28 1.27 12.7% 50.29 5.92
Garrison 12 10.22 0.85 6.4% 23.49 11.75

The team had the most success on the penalty kill when Klefbom was on the ice. Not only did the team have its lowest rate of goals against when he was deployed (5.02 GA/60), but the shots against were also low (44.21 SA/60). The other notable player here was Benning, who was fifth in TOI% and posted an on-ice rate of shots against of 41.31, the second lowest among defencemen. Russell and Nurse were on the other end of the spectrum with the team allowing their highest rate of shots against with them on the ice. Thankfully the play of Larsson, as well as the depth players like Gravel and Garrison, offset some of the team’s deficiencies on the penalty kill.

Segment 2: Klefbom injured, Russell healthy for 11 (21 games)

When Klefbom and Russell were injured against the Avalanche on December 11th, the team had to make adjustments to their penalty kill. Over the next 21 games, the coaching staff rightfully increased Benning’s proportion of the Oilers penalty kill ice time from 19.2% in the first segment up to 27.3%. And Gravel saw his proportion of ice time increase from 12.7% to 25.2%. Also seeing ice time in this second segment was young Caleb Jones, who played just under 20 minutes and posted respectable on-ice shot-rate numbers.

Player GP TOI TOI/GP TOI% SA/60 GA/60
Larsson 21 49.08 2.34 44.1% 41.56 7.33
Nurse 21 48.30 2.30 43.4% 52.17 13.66
Benning 18 30.38 1.69 27.3% 45.42 9.87
Gravel 14 28.07 2.00 25.2% 49.17 4.28
Russell 11 24.48 2.23 22.0% 78.42 14.7
Jones 17 19.07 1.12 17.1% 47.20 9.44
Manning 10 11.48 1.15 10.3% 62.70 10.45
Garrison 5 6.65 1.33 6.0% 36.09 9.02
Petrovic 6 5.28 0.88 4.7% 56.78 0.00

Over those 21 games, the team’s penalty kill did see their rate of shots against increase slightly from 48.62 per hour to 51.17 – but this increased rate was still right around the league average of 52.0. The actual rate of goals against remained the same due to poor goaltending, but the skaters were doing their job of limiting the shots getting on net. Key drivers for their relative success limiting shots in this second segment was the play of Benning and Gravel, both of which posted good numbers in their limited minutes over the first 31 game segment. Worth noting that Benning has a history of success on the penalty kill, dating back to the last two seasons.

What might get overlooked is the fact that even when his proportion of ice time was reduced, Russell continued to struggle on the penalty kill, with the team allowing 78.42 shots against per hour with him on the ice. Nurse’s on-ice number’s actually improved, likely due to playing half of his minutes with Benning and spending less time with Russell who was his primary partner in the first 31 games of the season. Manning and Petrovic also struggled and have since fallen down the depth chart.

Segment 3: Klefbom  and Russell healthy again (13 games)

In the recent 13 games with Klefbom back from injury, and Russell playing in every game, the Oilers penalty kill has actually been getting worse. As mentioned above, the team is allowing over 62 shots against per hour, one of the league’s worst, and it’s a little unclear as to what the exact problem might be. Below are the defencemen’s on-ice numbers over the last 13 games, sorted by TOI%.

Player GP TOI TOI/GP TOI% SA/60 GA/60
Russell 13 23.42 1.80 40.8% 58.93 7.69
Klefbom 13 23.38 1.80 40.7% 71.85 10.26
Larsson 13 22.35 1.72 38.9% 67.11 13.42
Nurse 13 19.62 1.51 34.2% 67.29 6.12
Sekera 7 8.05 1.15 14.0% 29.81 14.91
Gravel 5 7.42 1.48 12.9% 80.9 8.09
Benning 9 6.57 0.73 11.4% 36.55 9.14

Despite all of the evidence from the previous 52 games, the coaching staff increased Russell’s ice time on the penalty kill, where he leads the team in total proportion of ice time over the last 13 games. His results have improved from the previous segment, down from 78.42 shots against per hour to 58.93, but we are dealing with a small sample size and can probably expect his numbers to regress towards his career numbers. What’s also interesting is that Russell’s primary partner is Nurse again, as he’s played 19 of his 23 minutes with him. That again goes against the evidence we uncovered in over the previous 52 games when Nurse was doing much better playing alongside Benning (who along with Gravel saw his proportion of ice time drastically reduced).

The other issue has been Klefbom’s play on the penalty kill since his return and his on-ice rate of shots against. Considering how well he was playing prior to his injury (in the first 31 games), and even dating back to last season, it’s surprising to see his on-ice rate of shots against climb up to 71.85. My guess is that we’ll see his numbers regress towards his career averages over the remaining 17 games, but it’ll be worth keeping an eye on. He’s looked fine in every other situation – especially on the powerplay where the team has excelled since his return – so hopefully it’s not a lingering injury issue.

My sense is that the Oilers are fine to continue playing Klefbom on the penalty kill with Larsson, as his numbers should improve as he has had success this season. But the team should also be looking into getting Benning and Gravel more ice time, either together or have Benning with Nurse or Sekera. Benning has posted good on-ice numbers on the penalty kill all season, and it’d be in the Oilers best interest to do whatever they can to reduce the shots against considering how poor the goaltending has been.

Data: Natural Stat Trick

Special Teams Link

nhl-oilers-stars-112718-620

One area of the club that has be driving the Oilers coaching staff crazy is the special teams.

While the powerplay has been solid, generating shots and ranking 9th in the league scoring 8.10 goals per hour, the penalty kill has been the complete opposite. The Oilers rank 29th in the league allowing 8.96 goals against per hour when shorthanded, and have gradually been allowing more and more shots and chances against over the course of the season – more on that in a second.

Simply put, any production from the powerplay has pretty much been off-set by the dreadful penalty kill. The Oilers have scored 40 goals on the powerplay, but allowed 49 on the penalty kill. The biggest problem has been the Oilers goaltending in penalty kill situations, which ranks dead last in the league with an 82.75% save percentage. Had the Oilers received league average goaltending in penalty kill situations (currently 86.07%), the team would have allowed nine fewer goals and would be slightly higher in the wild card standings.

Oilers - Special Teams - Cumulative Goals

Digging into things a bit, I found the shot-metrics for the powerplay and penalty kill interesting, as they’re both trending in very similar directions. Below are the rolling 20 game segments for the special team units – shots-for per hour for the powerplay and shots-against per hour for the penalty kill. Note that the league average rate for shots-for and against this season is 52.0.

Oilers - Special Teams - Shots per 60

Looking first at the rate of shots-for on the powerplay (blue line), you can see how the team started off well but took a hit when Klefbom was hurt in game 31 against the Avalanche in mid December. The Oiler’s low-point came in those first 20 games without Klefbom (generating just under 39 shots per hour), but unsurprisingly improved once Klefbom returned in game 53 against the Blackhawks in early February, with the team generating 64.4 shots per hour since his return. Any team would have an issue replacing Klefbom’s skill-set on a powerplay with four forwards and one defenceman, but the Oilers especially struggled trying to replace his minutes with guys like Nurse and Benning.

What’s especially interesting in the graph above is the rate of shots against on the penalty kill and how it closely mirrors the rate of shots-for on the powerplay. While we can attribute the improved powerplay shots-for rate to Klefbom’s return, I’m not entirely sure what could have caused the spike of shots against shorthanded and why it’s so aligned with the powerplay shots-for numbers. Definitely something worth asking the coaching staff about.

One thing I’d look into is potentially scaling back Kris Russell’s minutes on the penalty kill. Over the last 20 games where the team has been allowing a rate of 60.7 shots against per hour and a goals against rate of 10.41 – both of which being one of the league’s highest – Russell has seen his total ice time (44:37) and ice time per game (2:22) gradually increase on the penalty kill, currently leading all defencemen. In the 20 games prior when the team was doing well on the penalty kill, allowing 50.7 shots against per hour and 7.1 goals against per hour, Russell had missed 10 games due to injury and only played about 14 minutes on the penalty kill (averaging under one and a half minutes per game). In those limited minutes, Russell was extremely poor, posting an on-ice shots against rate of 77.23 (!) and a goals against rate of 12.87 – both rates being the team’s worst among defencemen on the penalty kill.

Oilers Dec. 1, 2018 to Jan. 12, 2019 Jan. 13, 2019 to Mar. 2, 2019
Player GP TOI TOI/GP SA/60 GA/60 GP TOI TOI/GP SA/60 GA/60
Larsson 20 44.25 2.21 50.17 6.78 20 40.77 2.04 50.04 10.30
Manning 3 1.25 0.42 0.00 0.00 9 12.58 1.40 76.29 9.54
Jones 13 10.47 0.81 63.06 11.46 4 8.60 2.15 27.91 6.98
Nurse 20 39.32 1.97 48.83 10.68 20 39.10 1.96 69.05 10.74
Garrison 5 6.65 1.33 36.09 9.02 0 0.00 0.00 0.00 0.00
Gravel 18 31.60 1.76 53.16 1.90 7 12.52 1.79 62.32 9.59
Russell 10 13.98 1.40 77.23 12.87 20 44.37 2.22 68.97 10.82
Benning 17 25.67 1.51 39.74 7.01 16 16.08 1.01 52.23 11.19
Klefbom 6 8.80 1.47 47.73 0.00 13 23.38 1.80 71.85 10.26
Sekera 0 0.00 0.00 0.00 0.00 7 8.05 1.15 29.81 14.91

I wouldn’t say Russell’s the sole reason for the penalty kill gradually getting worse, but he’s one of the factors that the team should probably address. Considering how many goals (and wins) their penalty kill has cost them, it’d be imperative for the Oilers coaching staff to critically examine their tactics and player deployment, and make the necessary changes.

Data: Natural Stat Trick