August 28, 2024 by Ryan Lotsberg

Edmonton Oilers CEO Jeff Jackson made a smart move to beef up the Oilers’ analytics department last fall when he hired Michael Parkatti to be the team’s Senior Director of Analytics and Data.

Michael Parkatti has been named the #Oilers Senior Director, Data & Analytics! 📊

Working closely with the hockey operations department, Parkatti will lead the creation & operation of a new analytics department, providing insight-driven recommendations for the team. pic.twitter.com/2ITPw1DOeo

— Edmonton Oilers (@EdmontonOilers) September 21, 2023

Parkatti has spent his career in data and analytics in various sectors, but hockey is clearly a passion of his. He used to be a blogger, and his blog is still available to read at boysonthebus.com. Parkatti has a series of articles outlining the various steps he took to create a statistical model that would predict the likelihood of draft eligible CHL players having careers in the NHL. It’s a five-part series where each piece outlines a change that he made to the model that made it more accurate than it was in the previous piece. It was clear that the model was in the early stages of development when he wrote those pieces because he demonstrated that there was a lot of room for growth, but it was already reasonably accurate.

That’s just one example of how data can go beyond what is publicly available online. While I have no idea what sorts of projects Parkatti has been working on during his time with the Oilers, I have to believe that the information he provides to the Oilers is highly insightful and useful.

I can point to two moves made by the Oilers this summer that have Parkatti’s fingerprints all over them.

The first was the Oilers trading a conditional 2025 first round pick to the Philadelphia Flyers for the 32nd pick in the 2024 NHL draft, which was used to select London Knights centre Sam O’Reilly. 

Related: Oilers trade up to select Sam O’Reilly with final pick in the first round

They must’ve seen something they REALLY liked in O’Reilly to pull the trigger on that trade specifically to draft him. I’m not trying to put O’Reilly down here, but he wasn’t a featured player on the Knights this past season. He showed off his defensive acumen in a depth role on the Knights last season which is great, but it’s not like he popped off for 150 points and was a must-have prospect. As much as Director of Amatuer Scouting Rick Pracey liked what he saw in O’Reilly, my gut tells me that O’Reilly stood out as a smart player to draft based on Parkatti’s model that works to predict the likelihood of future success of CHL players.

The other move that stands out to me as a data driven move is the acquisition of Ty Emberson.

Related: The Ceci trade was more about Emberson than we realize

The Sharks got a third round pick and Cody Ceci, a veteran top four defenceman on a reasonable contract, and all they had to do was part with a 24-year old prospect that they scooped up on waivers for nothing.

That begs the question, what did the Oilers see in Emberson that made them willing to do this deal? Oilers head coach Kris Knoblauch coached Emberson with the Hartford Wolfpack in the 2022-23 season, and Emberson won the AHL Eastern Conference Defensive Defenceman of the Year award that season. I feel like there’s more to this one than the coach saying good things about the player though. It may have mostly been a cap dump, but the acquisition of Emberson is one that you have to dive into the data to truly appreciate from the Oilers’ perspective.

Emberson played on a Sharks team that ended up with the first overall pick in the most recent NHL draft, so a lot of the commonly analyzed underlying numbers don’t look great for Emberson. 

Comparing his performance to that of his Sharks teammates tells a more flattering tale about Emberson though. Let’s start with a comparison of how the Sharks performed at five-on-five with Emberson on the ice and without him on the ice using Natural Stat Trick’s Line Tool:

	SF%	GF%	xGF%	HDCF%	HDGF%
Emberson	46.21%	44.44%	43.26%	43.05%	50%
w/o Emberson	42.43%	33.33%	41.63%	41.11%	31.61%

The Sharks were better with Emberson on the ice than they were with him off the ice. While the expected goals for percentages are close, the actual goal share is wildly slanted in Emberson’s favour. The same can be said of the high danger chances and high danger goals percentages. 

It looks like PDO is a factor because Emberson benefitted from a 91.7% save percentage while he was on the ice, while the team’s save percentage dipped to 89.74% with Emberson off the ice. It’s also a small sample size for Emberson, just 498:43 compared to 3525:31 for his teammates.

PuckIQ tells us that Emberson got 32% of his ice time against elite competition, 43.2% against mid level competition, and 24.8% against low level competition. That’s on par with second pairing usage.

His best goal share came against elite competition (7-6, 53.8%), which is highly encouraging. It makes his 33.3% goal share against low level competition (4-8) immensely surprising. Once again, PDO played a role there as Emberson got a lowly 83.3% save percentage against low level competition compared to a 92.2% save percentage against elite competition. 

However, we can’t just use save percentage to explain why Emberson was so good against elite competition last season. PuckIQ analyzes Corsi and Dangerous Fenwick numbers relative to teammates (CF%RC and DFF%RC). In case you’re unfamiliar with the terms, Corsi is simply any shot attempt, and Dangerous Fenwick is a measure of unblocked shot attempts with the quality of those shot attempts based on the location of the shot attempt considered. Just for clarity’s sake, negative numbers are lower than teammates and positive numbers are higher than teammates. Higher numbers are better in these particular metrics.

	CF%RC	DFF%RC
All	-0.3	1.5
Elite	6.5	10.6
Mid	-5.3	-4.5
Gritensity (Low)	1.2	2

Emberson’s numbers against elite competition in comparison to teammates in these two metrics LEAP off the page!

If we only look at Dangerous Fenwick against relative to his teammates (DFA%RC) against elite competition, we see a figure of -23.05 for Emberson. That is absolutely ridiculous! His strong DF metrics against elite competition suggest that Emberson understands how to defend the dangerous scoring areas in the slot and near the net.

The big asterisk here is that these numbers are in comparison to his former teammates on the Sharks, who were the worst teammates in the league last season. This information tells us nothing about how he would’ve compared to defencemen on a Stanley Cup contending team.

So, we move on to individual stats. 

NHL_Sid (@NHL_Sid) from Oilersnation wrote a piece about the options to fix the team’s hole on the right side of the second pairing, and he shared some interesting numbers on Emberson.

He shared that Evolving Hockey says that Emberson ranks in the 89th percentile in defensive impact on expected goals at even strength. The fact that he ranked so highly in defensive impact on expected goals while playing a second pairing role on a terrible team is impressive.

NHL_Sid also said that data from AllThreeZones shows that Emberson ranks in the 75th percentile in puck retrievals, which is pretty good. He ranks in the 29th percentile in entry defence, and in the 24th percentile in controlled zone exits. That tells me that Emberson doesn’t excel at defending because he prevents the other team from spending time in his team’s zone. It tells me that his impact in the defensive zone comes from reading plays, defending the slot, and breaking up scoring chances. That sounds like a player that would thrive in Knoblauch’s zone defence scheme.

Emberson racked up ten points (1G, 9A) in his 30 games last season. Nine assists in 30 games is pretty good for a defensive defenceman whose team only shot 7.73% while he was on the ice. That level of production isn’t expected to continue, but there’s a chance that it might if he gets time with the Oilers’ scoring lines.

I’m certain that Parkatti and the Oilers have access to more data that I can’t see; but based on the data that I can see, Emberson has the making of a potential shutdown defenceman. We can expect more data-driven moves similar to the ones made for O’Reilly and Emberson in the coming years because the Oilers are going to be tight on cap space for a while and the team’s prospect pool is quite weak. That makes Parkatti a critical part of the organization.