Why Stuart Skinner struggled in the playoffs

The Edmonton Oilers 2022–23 season is done. After finishing the regular season in second place in the Pacific Division with a 50–23–9 record and dispatching the Los Angeles Kings in six games during the first round, the Oilers were eliminated in six games by the Vegas Golden Knights.

After what seemed to be a promising lead in into the playoffs, with a 18–2–1 record from March 1 on in the regular season, the Oilers just never looked right come postseason time. Connor McDavid and Leon Draisaitl once again played at record setting levels, while the Oilers power play continued to produce at the highest rate all time. However, that was more than offset by the secondary scoring drying up, and massive defensive breakdowns. But the biggest difference between the hot streak down the stretch and the playoff struggles points to goaltending.

It was obvious to start Skinner for the playoffs

The Oilers entered the postseason with Stuart Skinner as their clear cut starter. It wasn’t supposed to be that way for Ken Holland’s squad heading into the season, with former Toronto Maple Leaf Jack Campbell freshly signed to a five-year $25M contract to presumptively be the guy for the Oilers heading into the last few years of the McDavid and Draisaitl contracts.

Anything but that happened during the regular season. Campbell finished fourth last in the league with -18.81 goals saved above expected during the regular season according to Evolving Hockey. According to a methodology outlined here, Campbell lost the Oilers seven standings points on the year, before firmly losing the starting job towards the end of March.

On the other hand, Skinner surprised this year, taking over the starting role by the end of the year when the season started with questions if he would be able to handle the back up job. By the end of the season, Skinner had earned the Oilers 13 points due to his strong play. He was even playing his best hockey heading into the playoffs, finishing off his regular season campaign with six straight positive starts. Here is how he looked this season according to Top Down Hockey’s model.

So after such a strong regular season, what happened in the playoffs? Well Skinner indisputably struggled. He posted a 5–6 record, 3.68 GAA, and .883 SV%, finishing with the second worst -6.8 goals saved above expected. Now I don’t think he was as bad as the stat line reads, but with the margins being as tight in the playoffs, he was clearly on the wrong side of the ledger. I also think it should be important to note that the majority of these reasons were outside of Skinner’s control, so while you can point to his play as a cause for the second round elimination, it’s tough to put the blame on him. Here are some of the underlying issues why Skinner struggled.

Skinner had a huge workload

Significant advancements over recent years have been made in sport science across all major sports. Arguably the most noticable one has been in rest management. The NBA sees their stars rest regularly during the regular season and routinely come out of games early once the result is self evident. The NHL has started to employ this with their goaltending decisions. Now Carolina Hurricanes Assistant GM Eric Tulsky was one of the first to challenge the conventional wisdom to ride your starter, as his article on Broad Street Hockey showed decreases in performance when goalies were deployed in both halves of back-to-backs.

Since then, this practice has continued to grow. The volume of games that each starter has has gone down significantly on average, with the teams making use of two or sometimes three goaltending options regularly. Even if the downfall isn’t always immediate, an accumulation of games can lead to a reduction in performance in the future, as the Oilers saw with Cam Talbot’s struggles following his immensely successful 2016–17 campaign, where he played nearly every game.

On March 10, when it became clear Skinner was going to be the clear cut starter the rest of the way with Jack Campbell being borderline unplayable, I wrote this article outlining how the Oilers could deploy their goalies moving forward. Despite being in the thick of a playoff race, it was clear that an over reliance on Skinner could push his late season workload into territories few goalies see in any given year. When all was settled, Skinner started 16 games from March 1 on, which would be a 68 start pace over the entire regular season, which would have been first in the league. Skinner was successful during this stretch, picking up the March Rookie of the Month award, but it was a lot of stress on a goalie that had not played to this extent since junior.

The playoffs became even worse, with Skinner starting all 12 playoff games for the Oilers, despite never starting more than six consecutive previously. This brought the rookie goaltender to 28 games over 78 days, which would be a 72 start pace over a full season, leading to a crazy workload that he ultimately could not keep up with.

Not only was there the on-ice workload, but Skinner was dealing with a new born from January onwards. Ancedotally, this hasn’t affected goaltender performance right away, but rather starts to make a bigger impact when the new kids are toddlers.

To my eye, fatigue from the workload was likely the primary cause of Skinner’s downfall. Looking over each goal that he gave up, there wasn’t any new trends that were exposed in his game during the playoffs. Yes, teams did successfully target Skinner’s glove hand, but that was a trend all year and Skinner was still able to post the results he did. The excess goals against were more so coming from self inflicted mistakes that the Oilers’ netminder hadn’t made all year.

It’s possible the Oilers recognized this, but the next reason ultimately meant that they decided not to change it.

Lack of Confidence in Campbell

It’s hard to understate just how bad Campbell was during the regular season. His record looked good, but a large part of that was due to the Oilers’ scoring prowess. His personal performance was such that he was cumulatively ranked the fourth worst goalie in the league over the season by Evolving Hockey’s goals saved above expected. This wasn’t some anomaly either, as private firm Clear Sight Analytics ranked him as the fifth worst goaltender, with -18.88 goals saved above expected.

As the season wore on, it became increasingly obvious that Jay Woodcroft did not trust the 31-year-old goaltender in any game of importance. After a debacle in Winnipeg, where he was the primary contributor for the loss for the fifth consecutive start, it became clear that the coaching staff did not trust the veteran. His next time out might have been worse, where he “only” allowed four in an overtime win against San Jose, but was helped by three overturned goals and another two posts that beat him clean.

Campbell next played against the Arizona Coyotes with mixed results before his last two starts of the season came against the 32nd ranked Anaheim Ducks. It was in these pair of games where there started to be some noticeable technical changes to his game, which ultimately paid off, giving up a combined two goals over the two games. However, it would be tough for the coaching staff to put a lot of weight in games against the last place team in the league.

This complete loss of confidence was ultimately the downfall for both Campbell and Skinner in Woodcroft and Dustin Schwartz’s decision making process. When Campbell stole Game 4 of the first round in relief, the Oilers were hesitant to go back with him as it had been his first net positive appearance against a playoff team in over a month. This bet ultimately paid off, with Skinner having a strong Game 5 and 6 in the L.A. series.

Things started to go off the rails in the Vegas series. Skinner struggled in Game 1 but played the whole game before being relieved halfway through in Game 3. Vegas was up big at this point, and fell back into a trap, not attacking much, but Campbell once again looked good. Even with Skinner’s evident decline in play and heavy usage, the Oilers went back to him in Game 4. This decision was the real turning point, as it became clear that Skinner would be getting every start, regardless of performance, simply because the coaching staff did not trust the back up.

As a result, not only was Skinner playing likely his worst hockey of the year, but he was saddled with the extra pressure of knowing no matter what was happening, he would be the guy and wouldn’t be able to get a breather from the other half of the tandem. There’s been some commentary that Skinner couldn’t handle the pressure of the playoffs. Unless you believe the NHL playoffs present some distinct pressure that isn’t evident, this wouldn’t be the case as Skinner had been a clutch performer, winning championships at every other level he’s played at. It’s more likely that the specific situation with the tandem as things started to slide would have caused the extra pressure.

It also put Jay Woodcroft in a bind, as if Skinner struggled against before Round 2 was over, he would be faced with the decision between his struggling starting goalie and an untrustworthy backup that had not been tested much in the previous two months. This ultimately occurred, with Skinner being pulled once again in Game 5. Woodcroft stuck with his guy in Game 6, but ultimately Skinner was relieved for the third time in the series.

Multiple appearances against playoff opponents

The interaction between shooters and goalies is a real cat and mouse game. Skaters are trying to hide what they’re trying to do with deceptive releases and targeted body language, while goalies attempt to block everything down while monitoring other threats. Despite the mountains of video that are now available to teams, there is something to be said for experience that release from a first person point of view on both sides of the battle.

Initially, I’d believe that goalies would have an advantage. There are only so many effective ways to shoot a puck or disguise a release, so coaches and teammates can replicate most of what a goalie will see on a given night. Even then, there are now VR tools that can replicate what they don’t see in a simulated environment.

Goalies on the other hand are more unique in how each one approaches the game. There are fundamentals that each one relies upon, along with league wide trends as the position evolves, but each goalie is a bit different in their stance and all the movements that come from it. This would likely mean that with a larger sample against specific goalies, the pendulum would swing towards the skaters.

During the regular season, the Oilers ended up giving both the Kings and Golden Knights ample opportunity to see Stuart Skinner first hand. The rookie started every game against the two divisional opponents except for one in Vegas, while he was away for the birth of his son. Perhaps more importantly, both playoff opponents saw Skinner twice each in the last month, getting a last look at the opposing netminder as his game rounded into form.

It’s hard to pinpoint exactly what can be attributed in the playoffs to the extra looks that the shooters got, but there was definitely trends that hadn’t been evident through the regular season. The Kings attacked Skinner off his post play, while the Golden Knights targeted an already susceptible glove with inside-out releases that would pull Skinner away from that side.

Type of offence faced

Goaltending might be the toughest position in sports to properly evaluate. Where other roles can be evaluated on how they individually perform, goaltenders are constantly reacting to the on-ice environment in front of them, then evaluated on the result. As such, goalies might have the least amount of control over their individual metrics (particularly ones that aren’t adjusted for shot quality) out of anyone.

On the ice, goalies play each individual situation, which can result in some wonky stats that don’t necessarily reflect actual performance. The make up of the workload can have just as much of an effect on outcomes as performance. If the shots skew to the low danger side, the goalie simply has to not screw up, while higher numbers of higher danger chances can leave goaltenders repeatably in untenable situations. I think this is part of what happened to the Oilers during the playoffs.

In the first round, Todd McLellan was content to have the Kings sit back in the 1-3-1 neutral zone trap, waiting for the Oilers to come at them but ultimately making a concerted effort to limit shot quality. Due to them sitting back, if the Oilers made any mistake leading to an opening, L.A. would launch into a full scale attack, creating high danger chances without much support, leaving the goalie out to dry.

Come Round 2, Bruce Cassidy’s team used a different approach. During Vegas’ first round matchup, they made Connor Hellebuyck look extremely human, focusing on a combination of rush chances and tips/screens off point shots to beat of the league’s best goalies. This approach created a higher percentage of high danger chances, whether it be through broken plays, converted tips or slot line chances, which would theoretically tank a goalie’s traditional stats.

This continued into the second round against Edmonton. The Oilers were able to adjust their forecheck to limit rush chances later in the series but another issue arose for Dave Manson’s defensive group. Edmonton employed a man-on-man defensive structure. The Golden Knights were able to take advantage of this by drawing players up into the zone then attacking back down into the space that had just been left vacant. Jack Eichel’s line executed this the best for Vegas, while the pairing of Darnell Nurse and Cody Ceci, and Nick Bjugstad lines were victimized.

Overall, Skinner needs to be better

There is no doubt that Skinner did not perform at an acceptable level in his first time through the playoffs. Some team’s have won cups with mediocre goaltending in the playoffs (see Colorado in 2022), but it is extremely tough to do so, and the Oilers needed better from their rookie.

Skinner needs to be better, but there was a lot outside his control that ultimately led to the substandard playoff performance we saw. Since he took over the starting job, Skinner was given an unsustainable workload. The only reason he had the opportunity to take that job in the first place was because Campbell was deemed unstartable in any game that mattered, leaving Skinner as the only real option. Add in the fact that the opposing Kings and Golden Knights had seen Skinner recently and often this season, helping them create strong offensive game plans that would leave any goalie in a tough situation, and the fact that Skinner struggled becomes less of surprises.

Skinner starts a three-year contract with $2.6M AAV next year, in which we can expect to see him battle with Campbell for playing time. Based on his comments in his exit availability, I’d expect an even better version of the goalie we saw breakout this season after a summer of training.

---

Photo by Derek Cain/Icon Sportswire