If you told me this off-season that Edmonton would post a record of 18-14-2 in their first 34 games of the season, I would have been highly disappointed.
Over 82 games, that’s a pace of roughly 92 points. Edmonton ranks 4th in the Pacific and 9th in the Western Conference, just outside the 2nd wild card spot in terms of points percentage.
Saying this shouldn’t be acceptable would be a pretty massive understatement.
It’s crushing that the team still dwells in mediocrity in spite of four first overall picks and two elite talents on their current roster in McDavid and Draisaitl. We’re currently at the midway point of McDavid’s maximum term deal, and past the midway point of Drasiaitl’s maximum term deal, and the Oilers have a mere one playoff series win to show for it.
The expectations for a team with players such as them should be nothing short of a perennial contender. Edmonton should consistently be among the top five teams in the league, while possessing a legitimate possibility to win the Stanley Cup.
But it’s evident that the Oilers just aren’t good enough.
So, what are the major factors separating Edmonton from the league’s most successful teams? What are their specific areas of weaknesses, and should goaltending be treated as their most concerning issue?
*All stats via EvolvingHockey unless stated otherwise
Let’s talk about Edmonton’s goaltending
I’m merely stating the obvious, but it’s clearly evident that the Oilers require superior goaltending. However, should this be considered as Edmonton’s largest concern?
I won’t be using save percentage here. The primary flaw with SV% is that it only accounts for the quantity/number of shots against a goalie has faced. It doesn’t incorporate shot quality in any manner. Save percentage can often overrate goaltenders on prominent defensive teams, and undervalue goaltenders on poor defensive teams.
This is why I believe GSAx (Goals Saved Above Expected) is a superior evaluation method. GSAx is the difference between how many goals a goalie has allowed in comparison to how many goals they’re expected to allow. Expected goals combine both shot quantity and quality. These models achieve this by assigning the likelihood that a shot goes based on an in-depth formula using shot distance, shot location, shot angle, and game state.
This statistic isn’t perfect, but the vast majority of publicly available xG models have generally been fairly accurate in predicting future goals.
Going back to my initial point, the combined 5v5 GSAx of Edmonton’s goalies is -9.4. In different terms, Edmonton’s goaltenders have allowed roughly ~9-10 goals more than they should have. Quite obviously, this isn’t satisfactory by any means.
With that in mind, Edmonton has been out-scored 68/80 at 5v5. Even if you reduced their total goals against by 9-10 GA, they would still possess a negative goal share.
Furthermore, the vast majority of fans and media alike have specifically blasted Koskinen for Edmonton’s recent goaltending woes. But what would happen if Koskinen performed like an above-average starter this season? In a hypothetical scenario, let’s say Koskinen’s 5v5 GSAx ranked in the 85th percentile amongst NHL goaltenders, which is a GSAx of roughly 4.4. What would that mean?
Even if Koskinen’s totals were at this level, the Oilers would have allowed roughly 9 goals less at a maximum. Consequently, Edmonton would still be a net negative at 5v5.
My point still applies even if you have doubts about GSAx, and place more faith in the traditional/simple metric of SV%. Hypothetically, if Koskinen posted a 0.925 5v5 SV% this season, he would have allowed 10 goals less. Yet again, it wouldn’t be enough for Edmonton to sustain a positive 5v5 goal differential.
It’s fascinating that several xG models even have Koskinen performing close to the average goalie he is (at all-strengths). Micah Blake McCurdy’s (@IneffectiveMath) model has Koskinen at nearly 60 xGA, while he’s allowed 61 GA.
Personally, I feel that the goaltending issues, especially Koskinen, have been used as somewhat of a scapegoat covering Edmonton’s most significant concerns.
Of course, I understand the argument that Koskinen consistently allowing low-danger, early goals can be demoralizing to the team, and I thoroughly agree. Having said that, Edmonton has a 48.5% scoring chance share when the score is tied, and a 41.7% SCF% when the Oilers are up by one (per Natural Stat Trick). Even when Edmonton hasn’t allowed the first goal, they’ve struggled defensively.
I’ll clarify that my intention isn’t to justify Koskinen’s play. Both him and Mike Smith have to be better, and trading for a starting goalie at the deadline would definitely be beneficial. However, unless that goalie performs at a Vezina-worthy level, it wouldn’t be enough to turn Edmonton into a cup contender.
Defence, forward depth, and coaching are larger issues
As of the time I’m writing this piece, the top five teams in the NHL in terms of points percentage are Carolina, Florida, Toronto, Tampa, and Colorado respectively.
Here’s how Edmonton ranks in comparison to these teams in regards to defensive play:
It’s interesting that Florida and Carolina haven’t been exceptional defensively at 5v5, and a large portion of Carolina’s defensive value comes from their penalty kill. Nonetheless, they’re quite obviously superior to Edmonton in this facet, and so are the NHL’s other top teams.
League-wise, the Oilers rank 21st in 5v5 xGA/60 and 23rd in all-strengths xGA/60. Simply put, Edmonton has given their goaltenders a difficult workload.
Defence remains a huge issue, and there’s no question about that, but forward depth is in a similar boat. If you removed the top three scorers on Edmonton and the league’s top five teams, how would they compare?
That is an exceedingly large gap between Edmonton and the league’s top teams.
The argument can be made that Edmonton’s GA has worsened thanks to their goaltending, but it isn’t as if they’re generating much offence. I’ve mentioned this before, but Edmonton desperately needs strong finishers, as the bottom-six has scored 15 goals on roughly 19.1 expected goals.
None of the bottom-six forwards Holland acquired during the off-season excelled at putting the puck in the net. Foegele, Ryan, and Sceviour have historically scored at a rate less than they’re expected. Combine their inability to finish with Dave Tippett’s poor system and their putrid defensive structure, and the result is one of the league’s worst forward depth cores.
Speaking of Tippett’s system;
I’ve talked about this several times in the past, so I won’t go into full detail here, but the majority of Edmonton’s players have performed worse with Tippett as opposed to without.
In my opinion, a coaching change is necessary and could definitely improve the team. Their roster isn’t good enough to compete for a cup, but nonetheless, it’s evident that Tippett is making the team worse than they actually are.
Another huge difference between the Oilers and the NHL’s most successful teams? Edmonton’s lack of a proper analytics department
As the vast majority of NHL teams gradually begin to frequently accept analytics more and more, Edmonton decides not to adapt.
Edmonton currently has Justin Mahe employed as their manager of hockey analytics, alongside his brother, Shaun. However, Holland had clearly mentioned in a past interview that Justin Mahe is also in charge of handling/managing immigration.
If an organization has hired merely two people in analytically-based roles, one of them shouldn’t be having the time to also handle immigration/visas. This typically indicates that the team isn’t invested in analytics to a large extent. Furthermore, back in 2014, when asked a question about his use of analytics, Holland responded with, “the analytics we use are gut analytics.” This prior off-season, he was asked a similar question. Holland replied, “the analytics I like to use is 5v5 goal differential.”
Just based on their moves and these statements, it’s clear that the Oilers put little stock into analytics. In spite of his recent comments, Holland has acquired and re-signed numerous players with poor goal shares (e.g. Shore and Turris).
Meanwhile, the Tampa Bay Lightning, the double-reigning Stanley Cup champions, have utilized hockey statistics ever since 2009. That was the year in which they hired their current director of analytics, Michael Peterson (here’s an excellent piece about Peterson). Their use of statistics is evident in their moves, and a prime example is during the 2020 Trade Deadline.
Tampa acquired Blake Coleman and Barclay Goodrow, who had excellent underlying defensive numbers. Coleman and Goodrow formed a phenomenal checking line alongside Yanni Gourde, and the trio was integral to Tampa’s cup win that season.
The Colorado Avalanche also began a more analytically-inclined mindset four years ago. They’re currently first in the league in points percentage in the past three years. An example of this mindset is Devon Toews, who’s currently deployed on Colorado’s top pair. He’s probably the textbook definition of an analytical darling, posting excellent scoring chance numbers even when he was on the Islanders (funny enough, I had advocated for Toews back in the 2020 off-season as a replacement for Klefbom).
Carolina ranks 2nd in the league in points percentage in the past three seasons, and they’ve also been avid users of hockey analytics. They hired former nanotechnologist Eric Tulsky back in 2014-15, who began Carolina’s data department roughly six years ago. Tulsky has been an integral part of Carolina’s front office.
Another example? The Florida Panthers, who extended their analytics department ever since Bill Zito was hired in September of 2020. You could argue that Florida’s former GM, Dale Tallon, left Zito with a much larger mess than Chiarelli left with Holland.
Zito didn’t inherit Connor McDavid or Leon Draisaitl, and instead has Sergei Bobrovsky with a cap-hit of $10M.
In spite of this, Zito has managed to acquire Verhaeghe, Reinhart, Duclair, Forsling, and Hornqvist ever since, and Florida is currently second in the league in P% this season. To put more salt in the wound, Zito was one of the several candidates as Edmonton’s general manager, back in 2019.
I wonder where the Oilers would be today if Nicholson hired Zito instead. My prediction is near the top 2-3 teams in the NHL.
Ken Holland had $26M in cap space this past off-season.
With that much flexibility and cap room, he didn’t fix the goaltending. He didn’t find a 3C nor improve the forward depth. He also probably downgraded the defence.
Holland has been Edmonton’s GM for almost a thousand days, and nearly the exact same issues that persisted prior to his tenure are still huge concerns today.
While players like Sam Reinhart, Pavel Buchnevich, Viktor Arvidsson, and Jared McCann were all available throughout the off-season, Holland was too preoccupied with trading assets for Duncan Keith while not retaining a single dollar of his $5.5M cap-hit.
Even if you do feel that goaltending has held them back to a greater extent as opposed to any of their other roster issues, the blame should be entirely on Holland. He had plentiful time and cap-space to acquire an upgrade. He signed 39-year-old Mike Smith to a two-year contract before free agency had even begun. Some want to give him the benefit of the doubt by stating that he tried to search for an improvement, but that’s essentially the equivalent of granting him a participation award. With the exception of Zach Hyman (whose deal could age poorly in the long-term), it isn’t as if he fixed any of Edmonton’s other roster holes.
Holland is the highest-paid GM in the league; he shouldn’t get praise for merely trying. It‘s quite simply his job to improve the team and deliver results.
Edmonton’s goaltending remains a concerning issue, but it isn’t their most significant problem in my opinion. Management and coaching deserve the bulk of the blame for Edmonton’s mediocre record.
Find me on Twitter (@NHL_Sid)