Powerplay expectations


Quick follow-up to my recent post about the Coyotes penalty kill with Dave Tippett as head coach. Key takeaway from that: the Coyotes penalty kill had poor results for the most part, and the coaching staff did a poor job implementing tactics to limit the rate of scoring chances and shots against. They saw little progress over those eight seasons, leaving me to wonder what metrics they were using (or ignoring) to evaluate their penalty kill success and why the coaching staff couldn’t figure out a solution. Also learned today that it was Jim Playfair who was running the penalty kill in Coyotes, and he’s rumored to be joining the Oilers.

The Oilers made some changes to their coaching staff today, parting ways with Manny Viveiros, who was the eye-in-the-sky and had a history of powerplay success prior to joining the Oilers, as well as Trent Yawney who was responsible for the defenecmen and the penalty kill.

Couple thoughts on this. My expectations of the penalty kill under Yawney were low considering his teams in Anaheim were awful at limiting shots and relied heavily on John Gibson. And his results in Edmonton were poor, so it’s not surprising to see him leave. As for Viveiros, it’s disappointing to see someone who could have given the team a different perspective on the powerplay and possibly share new ideas depart. We don’t know for sure what influence he had on the Oilers powerplay last season, which was solid.

Speaking of powerplays, I went back and looked at how well the Coyotes did with the man-advantage under Tippett. Below are the rates of goals-for per hour (actual results), along with the underlying shot metrics which gives us a sense of (a) if the team’s success or failure was real or not and (b) what impact the coaching staff had. I’m of the belief that you don’t need star talent to generate shots and chances on the powerplay, but rather solid tactics that can drive offensive opportunities. And you can get a good sense of what influence the coaches have by looking at the rate of Fenwick (i.e., unblocked shot attempts, a proxy for scoring chances) per hour and shots per hour.

Season Goals for/60 Fenwick for/60 Shots for/60 Shooting%
2009/10 5.26 – 28th 64.65 – 28 45.77 – 29th 11.50 – 24th
2010/11 5.82 – 21st 68.06 – 21st 48.07 – 24th 12.11 – 19th
2011/12 4.67 – 30th 60.07 – 30th 40.55 – 29th 11.53 – 25th
2012/13 5.41 – 24th 63.45 – 23rd 44.61 – 23rd 12.14 – 21st
2013/14 7.44 – 4th 79.49 – 5th 54.5 – 10th 13.66 – 9th
2014/15 7.16 – 7th 81.06 – 4th 56.79 – 7th 12.60 – 14th
2015/16 6.38 – 19th 71.58 – 19th 51.37 – 16th 12.41 – 14th
2016/17 5.81 – 25th 64.87 – 25th 44.98 – 26th 12.93 – 15th

Looking at the actual results (i.e., the rate goals-for per hour), there were only two seasons of the eight when the Coyotes powerplay ranked top ten league wide – 2013/14 and 2014/15. And in those two seasons, their success was real as they ranked in the top five when it came to the rate of scoring chances, and top ten when it came to shots on goal.

Every other season, they ranked poorly when it came to shots and scoring chances, which played a big role in their actual results (goals-for) You could point to the lack of finishing talent and injuries along the way, but if your tactics aren’t enabling offensive chances, you’re not going to find success.

Worth noting that the Coyotes found success on the powerplay after they hired Newell Brown as an assistant coach. In his first two season with the Coyotes (2013/14-2014/15), the team did well generating shots and scoring chances, but then in his last two seasons with the team (2015/16-2016/17) things dropped off. I’d be curious to know what happened there, which coaches were involved and what influence Tippett had on the decision-making.

The results on special teams weren’t great in Arizona, and it’s especially concerning that the underlying shot metrics that the coaching staff have a bigger influence on were poor for the most part. It begs the question what metrics the new Oilers coach looks at and how he tests his methods over time – a critical component of analytics.

More importantly, Tippett’s past results emphasize the importance of establishing an analytics department that could support the coaching staff’s decision-making process – not only by collecting the data the coaching staff values but also testing the validity and reliability of it. It’ll be imperative that the Oilers invest in a group of people that can support the full hockey operations – including roster construction, drafting, player development and salary cap management. But especially the coaching staff responsible for optimizing the roster and their decision making around tactics and player deployment.

Data: Natural Stat Trick, Behind the Benches


Penalty kill expectations


Digging through Arizona’s on-ice results with Dave Tippett behind the bench, I was expecting, and really hoping, that his special teams numbers would have been average-to-good – especially the penalty kill. The Oilers have had dreadful results shorthanded, and it’s an area that will have to improve if the team wants to compete for a playoff spot.

Unfortunately, the penalty kill results over the course of eight seasons under Tippett were poor. Only twice did the Coyotes post a goals-against rate that ranked in the top ten league wide – 2009/10 and 2011/12. And those were largely due to outstanding goaltending performances from Ilya Bryzgalov as the club ranked near the bottom of the league when it came to the rate of unblocked shot attempts against (i.e., Fenwick, a proxy for scoring chances) and shots-on-goal against. Six of the eight seasons, the Coyotes would rank near the bottom of the league when it came to the rate of goals against.

Below is a summary of the Coyotes penalty kill, including goals, Fenwick and shots. Included is the team’s ranking in the league.

Season Goals against/60 Fenwick Against/60 Shots against/60
2009/10 5.66 – 6th 77.51 – 24th 54.51 – 19th
2010/11 8.12 – 26th 89.35 – 30th 63.84 – 30th
2011/12 5.2 – 8th 79.6 – 29th 55.62 – 27th
2012/13 7.53 – 22nd 70.24 – 20th 49.63 – 19th
2013/14 7.73 – 27th 77.97 – 22nd 57.36 – 21st
2014/15 8.42 – 29th 86.39 – 30th 62.69 – 30th
2015/16 8.09 – 28th 71.9 – 10th 50.79 – 12th
2016/17 8.29 – 26th 80.98 – 27th 53.84 – 16th

While the coach can’t control the goalies performance on the penalty kill, they do have a big impact on the structure to prevent the rate of shots and scoring chances against. Coaches are significant drivers on special teams, and their influence can be detected using the shot-based metrics.

And that’s what I found surprising looking at the Coyotes’ numbers. I had thought the penalty kill results (i.e., goals against) were driven by poor goaltending. But looking at the shot metrics, it appears the team did a poor job preventing scoring chances and shots against. The Coyotes were regularly allowing a higher than normal rate of shots, making me wonder what exactly the Coyotes coaching staff was thinking when assessing their results and making adjustments. They had eight seasons to correct things, so it’s a little concerning that they weren’t only bad league-wide, but consistently bad.

I’m doubtful that it was a lack of talent that drove these results – the Coyotes had the “grit” and role players that often find success shorthanded. So I’d really be curious to know what information Tippett and his staff were looking at over those years and how they came to decisions regarding tactics and deployment. It’ll be interesting to see how Tippett fills out his coaching staff this off-season, and what he has in store for the penalty kill.

Data: Natural Stat Trick

Thoughts on the hiring of Ken Holland + CBC Edmonton News segment



Professional hockey  remains in the dark ages when it comes to managing teams and looking for competitive advantages in the front office. Whenever there is a job opening for a general manager or when club re-structures itself to have an executive overseeing hockey operations, you realize that the list of potential candidates is small and the overall talent pool is extremely shallow.

The Edmonton Oilers at a critical juncture in their history, with the best player in the world and pressure mounting to deliver a championship went down a very predictable path to find a new general manager. They interviewed a number of assistant general managers, sought out the ones with potential and those that had connections to Hockey Canada. And while Ken Holland has the experience, the credibility, the knowledge and the professional network to handle the assigned tasks, his hiring isn’t all that exciting or inspiring. Going with Holland perfectly exemplified the Oilers risk-averse nature, their conservative approach to building a championship contender, and their lack of innovation and creativity when it comes to finding any competitive advantages in the modern era.

Now Holland does have a history at finding inefficiencies in the market when constructing a roster, being one of the first to tap into Europe for players and discarding one-dimensional enforcers from his teams. The problem is that the rest of the league caught up to him fairly quickly, and in the salary cap era he hasn’t done anything that sets him apart from his peers. He’s also shown a lack of understanding when it comes to player’s prime ages, giving out a lot of bad contracts with no-trade and no-movement clauses to players who are well past their primes, but he feels a deep connection with. Holland doesn’t appear to have that ruthless nature that’s needed to get the Oilers out of their current cap and roster issues. And we know that in the modern era one of a general manager’s key tasks is to find roster inefficiencies and squeeze out as much production and value from all corners of the club.

The one other issue I have with Holland is his “over-ripening” philosophy when it comes to developing prospects. While I do agree that prospects need time to develop and that a strong AHL program is critical to support the regular influx of talent to the NHL roster – management needs to be able to leverage a player’s entry-level deal and get production from players especially on the third and fourth lines. This would require identifying those players earlier and if appropriate, take on some risk by signing them to team-friendly deals. Again, it’s fine to allow players to develop in the minors, but managers need to be very strategic so that they know what they’re paying for when a player completes their entry-level deal and needs a new contract. Highly recommend checking out Iyer Prashanth’s piece from 2016 on entry-level contracts and asset management related to Holland’s time in Detroit.

Coming to Edmonton, Holland will have a lot of work to do including assembling a coaching staff, improving the roster and addressing the team’s scoring issues when McDavid isn’t on the ice. He’ll need to add a goalie and possibly add more skill to the blueline. And he won’t be able to do any of this without shedding some salary and also getting a better sense of the prospect pool and which players may be ready to contribute at the NHL level.

For me though, the biggest area Holland needs to address is the overall decision-making strategy and processes within hockey operations. This of course will depend on the management group he surrounds himself with, the professional and amateur scouts and (hopefully) a well supported and integrated analytics department. But Holland needs to focus on implementing the right processes that leverages all of that information that’s going to inform the roster construction. Hopefully he can do that and have the support from ownership to get all the resources required including people, technology and infrastructure – but we’ll have to wait and see.

While I do hope that the hiring of Holland is what turns this team around and start competing for championships, I remain fairly skeptical that the Oilers gained any sort of competitive advantage through this hiring.

For one, I don’t have a lot of confidence in the Oilers owner and his ability to allocate the right infrastructure and resources to support the general manager. We’re hearing that Holland will have a lot of control overseeing the entire hockey operations, but it remains to be seen what changes he’ll make to the people and processes in areas such as scouting and player development. Secondly, I don’t have a lot of faith in Bob Nicholson who facilitated the hiring process, had the opportunity to talk to anyone in hockey, but yet picked someone from his own history through Hockey Canada. In his time with the Oilers, the team took a reactive approach to several issues, and he’s a big reason why the Oilers are in the mess they’re in.

Lastly, I have little faith in the current hockey management talent pool. There’s clearly a need in hockey for common sense business acumen and a better understanding of risk management. The fact that the same names from the same demographic with the same type of backgrounds come up again and again makes it obvious that hockey management isn’t progressing. And that an opportunity exists for a team if they want to exploit an inefficiency in their management structure and find a significant competitive advantage.


Thoughts on the Oilers’ GM search


One thing that following the Oilers has done to me is change my overall perception of general managers and the value I place on them.

It’s obvious that they have a significant role in the game, overseeing roster construction, contracts and the draft. But unless the organization itself is set-up properly and has instilled a shared set of goals and values, with strong support from ownership, the general manager’s true abilities and their input are completely useless. And on the flip-side, if an organization is well structured, then you don’t need your typical hockey executive to fill the general manager’s chair – you could even get by with a non-hockey person with perhaps a background in finance or risk-management to give your front office a different element.

That’s a big reason why the Oilers current general manager search has very little appeal to me. The names that have been reported on and speculated about all seem fine, but it’s all but guaranteed the Oilers will go with a safe, conservative option – mainly because we know how this franchise operates under the current owner. In an era where professional teams need to be proactive, and progressive, and develop innovative methods to take calculated risks, going with a safe option just isn’t good enough.

Unless the team goes with a proper re-organization and constructs a front office that relies on solid, well thought-out business practices, a new general manager isn’t going to make any difference for the Oilers. Knowing which individuals are in the running for the Oilers general manager position, it’s likely going to be another four to five-year term where there’ll be some highs and lows for the franchise and maybe a playoff run if they’re lucky. But in terms of building a long-term, sustainable championship contender, the Oilers won’t be any closer.

The one thing I hope the Oilers do following the hiring of a general manager is to get a completely outside perspective of their organization as a whole, and receive some advice on how best to restructure the organization. Multi-million dollar organizations do this all the time to stay competitive in their industries, so it’s nothing ground-breaking. Ideally, this sort of consultation process would have started long ago and would have helped inform the role of the general manager and who best to hire. Because the Oilers demonstrated their usual complacency, and the fact that the franchise is heading into a critical time of their off-season, they’ll be taking yet another reactive approach.



CBC Edmonton News (TV): Re-capping round one of the Stanley Cup playoffs

cbc edmonton logoI joined host Nancy Carlson on the CBC Edmonton News to talk about the first round of the playoffs. Segment is here and starts at the 7:50 mark: CBC Edmonton News (2019, April 25)

Topics we covered:

  • Observations from the first round, and how unpredictable the tournament has been.
  • The early departure of the Flames, Leafs and Jets from the Stanley Cup playoffs and the reasons for their failures.
  • What lessons the Edmonton Oilers can take away from the first round.
  • Edmonton Oil Kings in action against the Prince Alberta Raiders in the WHL conference finals, and what to expect.


Next steps


After such a disastrous season, there’s going to be plenty to dig into in an attempt to chart out the necessary course(s) of action for the Edmonton Oilers.

What’s important before looking for solutions and making roster and salary cap related decisions is to look at the overall results and find where exactly the deficiencies were that prevented positive outcomes. With the end goal of winning championships in mind, it’s become even more imperative for the Oilers to make evidence-based decisions, and it needs to begin immediately.

The number to start any analysis with is the Oilers 2018/19 season goal-differential of -42, which was the eighth worst in the league.

Stripping out special teams and empty-net goals and focusing on even-strength (5v5), the Oilers had a goal-differential of -32 and a goal-share of 45.06% – both of which ranked third worst in the league and only ahead of New Jersey and Ottawa. The Oilers results shouldn’t be too surprising considering they were a poor possession team all season, finishing with a 23rd ranked Corsi-for percentage of 47.53%, and a 25th ranked Fenwick-for percentage (a proxy for scoring chances) of 47.34%.

Especially troubling is the fact that even with McDavid on the ice, the team struggled to out-shoot and out-chance opponents.

Metric (5v5) With McDavid Without McDavid
Corsi For% 49.32% 46.52%
Fenwick For% 48.81% 46.47%
Shots For% 48.04% 46.64%
Scoring Chances For% 49.70% 45.81%
High Danger Scoring Chances For% 46.46% 45.69%

What’s worth noting is that it was really the Oilers’s offence that struggled, with the team only generating 27.99 shots per hour, one of the worst rates in the league. Defensively, they weren’t great, but they were just below league average allowing 31.36 shots per hour – 19th in the league. This isn’t to absolve the Oilers defence for the team’s struggles; they were a big reason why the team couldn’t generate offence. But when it came to limiting shots, they were better than what they got credit for.

The other factor in the team’s even-strength results is their shooting percentage, which was 7.68% (21st in the league, league average being 8.05%) and save percentage, which was 91.51% (25th in the league. league average being 91.94%). Had the Oilers finished with a league average shooting percentage, they would have scored approximately 153 goals, a seven goal improvement from the 146 they actually finished with. And if they had league average goaltending, they would have allowed approximately 169 goals instead of 178. So instead of a -32 goal differential at even-strength, they would have finished with a -15 goal differential – about three and a half more wins in the standings. Had the coaching staff figured out a way to generate a league average rate of shots on goal, that goal differential could have been a lot better and the team much closer to playoff contention.

And while the Oilers powerplay was good this season, and probably could have been one of the best had they received better production from the second unit, the penalty kill was atrocious. The team allowed a rate of 9.21 goals against per hour shorthanded, second worst in the league, due in large part to their poor goaltending and their inconsistency in preventing shots and high danger scoring chances. Had the Oilers received league average goaltending of 86.26% instead of the 31st ranked 82.68% they did receive, the Oilers would have allowed approximately 49 goals instead of 62 – a difference of 13 goals.

Now going back to the overall goal differential of -42.

Had the Oilers been average at even-strength (better by 17 goals), had received better production from the second powerplay unit (additional four goals), and had an average penalty kill (13 fewer goals). their overall goal differential would have been -8 instead of -42. Still not good enough to be a contender, but at least closer to being a wild card playoff team.

And that right there should be enough to trigger the Oilers management to make changes – obviously to the roster but also to the coaching staff. The team, even playing to its abilities, isn’t good enough. A -8 goal differential would not have been good enough. There’s an obvious need to improve the scoring talent up front to play in the top six, but it’s also worth looking into depth players who can contribute at even-strength and also have an impact both on the penalty kill and second powerplay. The defence could absolutely use more skill and offensive talent- but again I think the team could focus on depth players who could contribute as complementary players at even-strength, but also play a feature role on special teams. That’s going to be up to the front office to uncover these undervalued assets and ensure they get a fair chance from the coaching staff.

Speaking of which, I think what’s getting overlooked because of the general manager vacancy is the importance of the next head coach, who will have to figure out how to maximize what could again be a below-average roster behind McDavid. The Oilers may have to take a conservative approach this summer and build through the draft, putting greater importance on improving the underlying numbers at even-strength and special teams – areas that coaches can have a direct impact on. The Oilers do not have the assets to give up in a trade, which will force the team to rely on tactics and the exploitation of other inefficiencies rather than actual talent and skill to win hockey games.

Data: Natural Stat Trick

CBC Edmonton News (TV): Oilers season post-mortem, GM search and off-season approach

cbc edmonton logoI joined Nancy Carlson, the new host of the CBC Edmonton News, for my weekly television segment to discuss all things Oilers. Clip is here and starts at the 19:30 mark: CBC Edmonton News (2019, April 8)

Topics we covered:

  • The key reasons why the 2018/19 season was such a disaster. Included on the list: poor roster construction, lack of offensive production, poor possession numbers, horrible penalty kill, inconsistent goaltending – just to name a few.
  • Key takeaways from the Oilers press conference this morning: lack of urgency among executives when it comes to long-term strategic planning, the same messaging we’ve been hearing for a while (and little action), and the overall, passive approach the Oilers are taking with their general manager search.
  • Course of action this off-season, which should include being ruthless with their cap situation, and creating a market for the players whose value might be perceived as higher than it actually is.

Thanks again to the team at CBC for making it all happen. It was cool to be the first guest on the new set with Nancy!