2023-2024 Debrief: Microstats in 2024
** The graphs might require desktop viewing for the best experience. If you’re on mobile or in email, make sure to click on the plots to have them open in a new tab to make sure you’re getting the fully updated and complete information**
The primary purpose of this series of analysis was to learn lessons about the Blue Jackets when they were playing their best. The idea was to observe how they did it, what worked, what didn’t and then apply all of that information so that we can construct a better team. Ideally, one that plays to each player’s or the team’s-as-a-whole strengths as much as possible.
The best way to do that is through the AllThreeZones hand-tracked microstats database. It’s the best resource for describing in detail the what and why of events on the ice.
I have had to make my own graphs because the data on the AllThreeZones website doesn’t have the possibility of filtering the data for time but also because the visualizations haven’t been updated to reflect data after the initial wave of tracked games.
Sample Size
Unfortunately, it’s also a limited tool. All of the tracking comes from a single person, Corey Sznajder, who puts in a Herculean effort to get an absurd number of games tracked in detail.
As it pertains to this specific analysis we’re then presented with a problem. A vast majority of the games were tracked during November with only 5 total games during “the Good Stretch.” If we’d like to know the difference between the early season and that good stretch, we’d have a highly fractured data set.
So, I’ve had to pivot for the use of AllThreeZones tracked data and even still we won’t be able to drive powerful conclusions. I have, instead of exclusively using “the Good Stretch,” used games from the entirety of the 2024 portion of the last season.
This leaves us with the following games in the tracked sample:
1/2 vs Boston
1/9 vs Winnipeg
1/13 vs Seattle
1/25 vs Calgary
3/7 vs Edmonton
3/30 vs Pittsburgh
4/13 vs Nashville
5 of these games are during the hot streak and the other 2 are after the team was largely destroyed by injuries.
For the following data, I have removed all players who have less than 40 minutes of 5v5 ice-time with the exception of Boone Jenner. I thought his performance was worth including as a sort of example to be wary of what extreme sample sizes can do.
What these new games do give us is a bigger sample size, at least relatively, of David Jiricek and a look at what Alexander Nylander brought to the table.
Data Warning
So, before we get into the actual data, I’ll offer the following warning. The following performances are extremely volative and subject to variance and noise in data. I’m sticking to the robust metrics to the extent that I can and any conclusions reached can’t be done with absolute certainty.
Many NHL teams with very high quality data will evaluate their team in 10 game benchmarks but more robust samples, of data with more variance, can require as many as 25 games.
AllThreeZones data, on the whole, is highly subject to coaching and role. It does not account for quality of competition, system asks or even score effects.
Graph Descriptors
All Data is from 5v5, garnered from AllThreeZones.com, and nearly all metrics are moved into rate based metrics. At times, they are Percent metrics and I will acknowledge each of these under each specific graph.
In each scatterplot, I have varied the size of the player indicator. Most of the time, I have tried to use metrics that correlate with “workload.” Often, that’s simply total 5v5 icetime tracked but in other metrics it is DZ Puck Touches or some other relative metric.
In each, the plot size, upper and lower bound, will be determined by the entirety of the NHL’s performance in the metrics. The dark lines indicate average performance and the larger bars indicate the upper and lower standard deviations across the sample.
Since the bars and analysis don’t require these hard numbers, I didn’t do any of the deeper math to weight averages/variance by ice-time and so these bars and averages shouldn’t be used in any rigorous way.
For these plots, the Blue Jackets players may look a little goofy with their orientation. If there’s a large amount of white space that’s because there are NHL players occupying some of the upper bounds. The shaded areas, those created by the standard deviations, should be used to give us ideas of the Blue Jackets’ performance relative to the NHL.
These means and standard deviations are also pulled exclusively from the 2024 portion of the season with any players who had less than 40 minutes of tracked ice-time removed.
In all cases, forwards are represented by dark blue and defensemen in red.
If the information is overwhelming at first, that’s part of the intention. In order to not have to offer a tedious number of graphs, I’ve stocked a lot of information on each one. You can always hover the point to get the full information and even hover the color to filter out opposing colors.
If anything isn’t working, let me know! These plots should update live as I make any corrections.
Offense
The first metric, and likely the most robust, is offense production volume. One of the major reasons I sought to include the noisy dual-ranges was so that we could have an idea of total contributions across different domains.
In this case, we can see that even very good defenseman still have below-average-forward-level impact on shots and shot assists. With that said, a couple of data points stand out.
Early in the season, the Blue Jackets were among the worst teams in the league at generating shots off of passes. In this small sample, Kent Johnson stands out as an excellent generator. His poor start, and even Johnny Gaudreau’s continued sub-optimal performance, have to be a focus for improvement in 2024-2025. Whatever is causing the playmaking left-wings to underperform needs to be addressed.
That being said, his flashes of playmaking in this small sample give us a glimmer of hope. I am not bringing these metrics to the graphs here but his center-lane playmaking and high-danger playmaking, as well as shooting from High Danger Passes, also stood out among all forwards in the small sample. It’s not a large sample but there’s still plenty of Kent Johnson to be uncovered.
Otherwise, nearly all Blue Jackets defensemen are above average in shot and/or shot assist generation. Adam Boqvist’s playmaking stands out the most but the rest, outside of Gudbranson, are tightly grouped as relatively high volume shooters.
Whatever the Blue Jackets did during this sample resulted in the D shooting quite a lot. This shouldn’t come as too much of a surprise considering the point generation numbers of the Blue Jackets D corps.
Zach Werenski obviously had an elite 5v5 offensive production season but it’s worth investigating whether their offense coming from defensemen was coming at the expense of higher value offense elsewhere.
Cole Sillinger really stands out. Early in this sample he got plenty of time with Johnny Gaudreau but ended the season in quite dramatic fashion as part of a bona-fide matchup line with Alexandre Texier and Kirill Marchenko. At the very least, Cole Sillinger at least showed the high-volume shooting that made him a high draft pick out of the USHL.
Gaudreau’s offense is also disappointing. Though he reclaimed a good chunk of his season and performance, and drove great underlying results during “the Good Stretch,” you’d expect him to be able to carry some of those late season games a little more.
That being said, he really tailed off towards the end of the year. His 5v5 icetime dropped as a result of placed on a line with Dmitri Voronkov and Alexander Nylander, both of whom greatly fell off after their hot starts.
Alexandre Texier has, perhaps, gone under the radar as an offensive force this entire season. He’s one of the best Blue Jackets plays along the wall and has taken a dramatic step forward in converting his dekey and tricky puckhandling into small areas off the wall. Unfortunately, he hasn’t left behind his propensity to turn over pucks and play an exclusive forecheck game.
If he had, while also improving his defensive awareness and discipline, perhaps that early Texier-Fantilli-Laine line would stuck together and he would have posted his breakout year. There’s room to ask whether all of those improvements are possible and the answers to that question could determine whether he gets a contract with the Blue Jackets for 2024-25.
The wall and retrieval skills, however, are huge reasons he “worked” with Cole Sillinger and Adam Fantilli. There’s much more to be said on the topic but now isn’t the time and here isn’t the place.
Here, “Offense” is defined as shots given the situation. Rush Offense is any shots coming within 4 second of a zone entry. Cycle and Forecheck are lumped together because they both come greater than 4 seconds after any entry. Of course, sustained possession and forecheck takeaways can offer dramatically different types of offense.
As far as the NHL sample goes, we can see that a primary differentiator between the type of offense that Forwards and Defensemen commonly produce lies within rush vs forecheck/cycle offense. Since defenseman are so often tasked with starting possession at or behind the net, they have significantly more work to do to get involved in rush offense.
The Blue Jackets were primarily a rush generation team. In fact, Alexander Nylander, Alexandre Texier, Kent Johnson, Cole Sillinger and Ivan Provorov were the only above average NHLers in terms of Cycle & Forecheck offense in the sample.
That Damon Severson also produced more than all but the above four forwards highlights a tremendous area of opportunity for the Blue Jackets. Their in-zone offense needs dramatic improvement. Largely, that tracks. Young players often play “one and done” style offense, especially true with Adam Fantilli in the back half, and have a lot of work to do to become quality shoot and retrieve players.
Despite Rush Offense being the primary driver of the Blue Jackets’ offense, notice the gargantuan amount of white space to the right of even Cole Sillinger here. The best offenses are driven by outright dynamic rush creators like Jack Hughes, Artemi Panarin, Jack Eichel or Nathan MacKinnon.
The top in-zone creators are Aleksander Barkov, Nikita Kucherov, Kirill Kaprizov and Sidney Crosby (who is also a top 10 rush creator). Either way, the Blue Jackets need Johnny Gaudreau to return to form and Adam Fantilli to reach his potential quickly if they’d like to become short term competitors.
I think we can see, at least partially, where both Johnny Gaudreau and Kent Johnson both struggled. They simply did not shoot off of the rush. Part of that is their pass-first style but additionally the lack of return passes or otherwise the volume that would result in their shooting.
In 2022-2023, Johnny Gaudreau generated 13.42 rush shots/60 and 12.7 cycle and forecheck shots/60. Kent Johnson was evenly balanced with around 8 shots/60 from each situation. Though he wasn’t included, 2022-2023 Patrik Laine generated 15 forecheck and cycle shots/60 and 11.78 rush shots/60. In the early season of 2023-2024, Laine generated 5.5 cycle and forecheck shots/60 but 13.07 rush shots/60.
In any case, true high quality rush offense didn’t really exist for the Blue Jackets.
Perhaps Kent Johnson’s ability to generate shots from cycle & forecheck is a sign of potential improvement as well. His ability to occupy dangerous areas, especially when paired with Dmitri Voronkov and Kirill Marchenko (or briefly with Sillinger and Nylander), was a hallmark of his second-half production.
Though his transition game largely deteriorated this season, more on that later, perhaps his return to effective neutral zone play will be the springboard for a breakout on top of some improved in-zone creation habits.
Werenski’s overall shot production remains a little disappointing but perhaps it’s a better indicator of the potential weakness of using a sample this small. Werenski was one of the best, and most dangerous, 5v5 shooters among all defensemen. It makes sense that a volume shooting approach doesn’t necessarily mean high quality offense in cycle and forecheck scenarios.
Similarly, shooting off the rush and playmaking from the blue line make a lot of sense for top defensemen. Prioritizng high quality shots and otherwise setting up teammates may look exactly like Werenski’s profile. Still, increased volume on top of both of those would still be preferable.
The top rush shooting defensemen: Cale Makar, Adam Fox, Erik Karlsson, Victor Hedman, Mackenzie Weegar, Josh Manson, Shea Theodore, Dmitri Orlov. The top cycle/forecheck shooting defensemen: Brent Burns, Kyle Burroughs, Quinn Hughes, Jared Spurgeon, Alex Pietrangelo, Dougie Hamilton, Tyler Myers.
Not huge differences, and there are very good defensemen in both cohorts, but the top rush shooters also tended to be quite high in all shots.
Transition
Carry% enumerates the share of a player’s Zone Entries/60 that the forward completed with control of the puck. Largely, volume of controlled entries is the most desirable transition trait in forwards because it often results in higher quality offense.
Here, once again, we can explain some of the perceived drop off of Johnny Gaudreau and Kent Johnson’s games. In 2022-23, they were starkly different transition players. Johnson completed a similar number of total entries but at a 62% controlled clip. Gaudreau too completed a similar volume but finished above 70% Carry%.
Improving their ability to impact the game in transition should be one of the higher priorities for Pascal Vincent, or whoever is coaching the Blue Jackets, in 2024-2025.
Adam Fantilli’s transition game, though a small sample, showcases a nearly perfectly translated transition ability. Prior to this sample, he was carrying the puck in at around 70% which put him in line with players like Connor McDavid (perhaps underperforming due to injury and prior to the coaching change), William Nylander, Martin Necas, Robert Thomas and Jesper Bratt.
If he can simply improve his volume he’ll be well on his way to becoming a dominant rush force (the pinnacle being Jack Hughes with 31.91 Entries/60 at an astounding 86.2% controlled rate or Logan Cooley with fewer entries at 12.51 Entries/60 but a 91.7% Carry%).
Cole Sillinger struggled to enter the zone with control but generated a massive volume of entries. His similar players in the larger AllThreeZones sample are Jordan Kyrou, Josh Anderson, Joel Farabee, Nils Hoglander, Trevor Moore, Mason Appleton and Nick Foligno. I am not sure what it means that these are primarily wingers but at the very least Sillinger is driving some type of entry and beginning to carve his identity.
As far as defensemen go, it’s nearly not worth including but I think it does reflect the gap in performance and expectation between the roles. It’s likely that, due to the Blue Jackets’ preference for neutral zone passes into middle drives from defensemen, their entry transition efforts are being undercounted in this sample. Said differently, they activate through the middle and take the strong side defenseman in a more selfless act of space creation.
Still, generating more total entries could be something to strive for for the best Blue Jackets puck movers, especially Zach Werenski. In previous years, Werenski was among the top defensemen in both Entries/60 and Carry% though not one that distinguished himself from the pack.
His peak season, the small sample in 22-23, included 9.33 Entries per 60 at a 63% Carry%. That’s slightly more than Ivan Provorov produced this past season.
The best defenseman in the league, Roman Josi, averages around 15 Entries/60. This past season, he was asked to change his approach away from full transition carries and it helped support his forwards much more.
This graph is a little bit different. This time, the dot size is not representative of Time On Ice but instead D Zone Puck Touches/60. I believe these are better indicators of total workload in this specific use case.
Of note are the large disparity between forwards and defense. Though forwards drive a near equal volume of exits they have a much higher success rate in the tracked data. The reason? It would be hard to know for sure but most likely it has something to do with the difficulty of pucks received and the way credit is attributed for exits.
Defenseman are often retrieving pucks and exiting from more difficult situations. Wingers, or even centers, are most of the time given pucks through a layer of defense and much closer to the blue line. Additionally, many wingers are given breakout pucks on the half wall in contested situations. If they fail, that’s probably just largely chalked up to a lost puck battle. If they succeed? Zone exit.
It’s hard to say for sure what exactly drives this tendency but it’s worth keeping in mind as we evaluate the results.
Of note, the only players that are above average in all categories with respect to their position, during this sample of course, are Kent Johnson, Johnny Gaudreau, Zach Werenski and David Jiricek. If these results hold over large samples, the Blue Jackets will at the very least have high quality puck movers beginning the attack. Perhaps building from the back, rather than pumping pucks north, then becomes a possibility.
I would also be curious if the increased exit burden played a part in Johnny Gaudreau and Kent Johnson’s relatively suppressed rush offense numbers.
Adam Fantilli is the only forward who shared a defenseman’s workload on zone exits. In fact, he was second to only Zach Werenski in terms of DZ Puck Touches/60. Perhaps that explains some of his difficult results in the back half of the year (it could also be an artifact of sample size).
Largely, I don’t think it’s necessarily a bad thing. His getting puck touches means he’s preparing to be a center who leads through the back and get more comfortable with the puck in these high leverage moments.
Perhaps Cole Sillinger and Alexandre Texier could stand to be more involved at transitioning the puck out of the zone but maybe it was their low total work volume that helped them pinch in successfully when they were called upon.
If the Blue Jackets seek to improve their league worst failed exit% it looks like Jake Bean, Adam Boqvist and Erik Gudbranson would be the place to start. Though Werenski and Jiricek only performed in small samples they both posted nearly elite puck-moving results. If Jiricek continues to shore up his consistency and reliability with the puck, the Blue Jackets could have two puck-moving anchors.
Then again, perhaps Boqvist and Gudbranson suffered because of role. Looking further into the data, though, and those two were by far the worst at completely failing to exit the zone (turnovers or icings) in the sample as well. Still, it’s easy to see how they would be the second option in any of their pairings and as such have limited opportunities to actually drive zone exits.
Defense
The Forechecking metrics are a little more nebulous. Dump-in Recoveries/60 is pretty easy to understand but Pressures are more nuanced.
The first off-puck events tracked relate to the forecheck. Starting with Forecheck Pressures & Exit Disruptions, which are how often a forechecking player forces the player exiting the zone to make a play with the puck. All pressures are counted as such & disruptions are credited when the forechecking player forces an uncontrolled exit (clears, turnovers, missed pass, icing etc.). These are shown because forcing turnovers or a failed exit is shown to lead to goals & chances, while forcing bad exits gives the defending team a higher chance of recovering the puck & attacking.
In the first half of the season, the Blue Jackets really struggled with their forecheck. Only Johnny Gaudreau, Sean Kuraly, Dmitri Voronkov and Justin Danforth were above average pressurers of the puck.
During this 2024 sample, their more skilled and younger players stepped up to the plate. Yegor Chinakhov grew into his retrieval role (when positioned with his Pressure data, comparing to large sample players like Brad Marchand, Matt Barzal, Roope Hintz and Brenden Gallagher) and Texier started to find his game as well.
Fantilli went from a puck retriever to a breakout pressurer which could very well be an artifact of sample size but could also indicate that he is growing as a forechecker/exit-killer. With the direction of the NHL’s rush offense moving towards neutral zone regroups, and his potential as a downhill attacker, this could be a significant development.
That Kirill Marchenko began to disrupt breakouts should be a point of focus for his game considering its synergies with his beneath dots shooting prowess. If he can learn to recover pucks and forecheck, he’ll become a nearly perfect player opposite Johnny Gaudreau.
Retrievals, like forechecking pressures, are also a bit more complicated to define. Largely, it’s any time a defenseman approaches a dump-in or loose puck in the neutral zone under pressure. Their Successful Retrieval% is how often they exit that maneuver with possession.
After that, we can develop some stats looking at how often each defenseman is at facilitating zone exits, one of which is “retrievals leading to exits” which are how often a defenseman’s successful retrieval leads to a successful exit (an exit with possession or a clear that the forechecking team doesn’t retrieve). Basically a way to show how good each defenseman is at evading pressure to solve problems with the puck, even if they aren’t the players exiting the zone. The start of the zone exit, if you will.
This time, the dot size is associated with total retrievals/60 as a way to demonstrate workload across their sample.
This graph, along with the zone exit graph, demonstrates Zach Werenski’s stature as an elite puck mover, at least during this sample. He started off 2023-24 poorly and graded out as one of the riskiest puck movers in the league. Once he got his timing, he greatly improved and proved his bona fide chops.
Repeating this performance is crucial to him maintaining his True #1 Defenseman status. His retrievals leading to exits per 60 would rank him higher, over the course of this small sample, than other dynamic puck movers like Miro Heiskanen, Chris Tanev and Victor Hedman. The small sample makes his performance not as robust, #1 defensemen aren’t just good in stretches, but it’s at least a 7 game sample to build from.
That no other defenseman on the Blue Jackets could perform above average in successful retrieval % demonstrates part of the issue facing this club. They simply did not perform under pressure.
Similarly, David Jiricek showed that he certainly has what it takes to be a talented, if not potentially elite, puck mover. He’s still failing retrievals and exits too much for an NHL coaches taste but once he cleans it up he could be special. The ceiling is there, the consistency isn’t.
Furthermore, we can see how much work Werenski did for Adam Boqvist. Perhaps this was an intended role dynamic but it’s hard to see how pairing Werenski with another retriever wouldn’t bring better results.
He’s not included in this sample, nor are any forwards, but Patrik Laine was by far the most prolific puck retrieving forward for the Blue Jackets in the early 2023-24 data. This extends his eager low-zone work from 22-23 and was likely the foundation of his transition to Center.
If he can return to the lineup and continue to perform at near defensemen levels of breakout work, he could be a significantly impactful addition to the lineup.
Few forwards have a strong impact on puck retrievals leading to exits but the most consistently high performing are a strong group: Connor McDavid, Ryan O’Reilly, Aleksander Barkov, Kirby Dach and Matty Beniers.
Entry Defense is also a little more complicated to explain, or at least to track. Corey Sznajder tracks any time the opposition attempts to enter the zone with control and measures how often that opponent carries the puck into the zone with control or is forced into a turnover which would then be registered as a denial.
In this plot, the size of the dot is related to number of entry targets per 60, the most appropriate measure of workload.
Many of the top defensive teams are excellent neutral zone deniers the pre-eminent two being Carolina and Vegas but early season overachievers like Philadelphia and Arizona popped up as well. High performing teams like Colorado and Vegas also feature defensemen heavily in this regard.
The Blue Jackets were horrific at generating blue line turnovers in 2022-2023 and didn’t fair all that much better this past season. That being said, aside from the very late games, “the Good Stretch” represents one of the better performing defensive samples of the season for the Blue Jackets.
From my perspective, the increased denying of Zach Werenski and Adam Boqvist are positive steps but they still struggle to cohesively deny zone entries evidenced by their very high Carry Against%. If Werenski wants to be a truly elite defenseman, he’s got to improve here.
Theoretically, Damon Severson was the perfect partner for Zach Werenski for exactly this reason. Early in 2023-2024, he was performing up to his previous NJD standards (45% Carry Against%, 15% Denial%). In this 2024 sample, Severson regressed hard.
Whatever happens next season, allowing Severson to reclaim his neutral zone aggression should be one of the first priorities.
It’s also not entirely out of the question that the Blue Jackets simply didn’t have the forward support to enable Severson’s game. Then, when he pulled back and was a more conservative defender, his defensive numbers improved. Without deeper study, this could all be an artifact of sample size as well.
Similarly, Jake Bean and David Jiricek had incredibly up and down neutral zone killing seasons. Pascal Vincent justified Jiricek’s demotion to the AHL as issues with “gap control” and perhaps it was this data that was more damning than any of the puck-risks he was taking. Jiricek, at least in European pro and World Juniors, was an aggressive neutral zone defender and talented entry killer.
In order for him to become a dominant neutral zone defender at the NHL level, he’ll need to drastically improve his skating or have near perfect reads (while probably also improving skating). With improved lower body and/or core strength, it’s possible some skating deficiencies iron themselves out. With improved skating, it’s likely his aggressive pinching and hitting become significantly more forceful and effective.
Jake Bean is an excellent aggressive neutral zone defenseman but this chart hides that he was also significantly risky. Though he did deny entries very well he also allowed passes and chances often after entries. Perhaps this came from being paired with Gudbranson and/or Jiricek but it also represents his struggles in overcoming his 3rd pair status.
Ivan Provorov continues his streak of being solid, if largely unspectacular across all metrics. His reputation was saved by partially excluding some of his failed exit and botched retrievals metrics but he certainly had a resurgent second half when played against reduced competition alongside Damon Severson.
Wrap Up
Next article, I’ll finally tie together all of the different data and go in depth on the potential lessons learned, questions asked and potential areas for improvement for the Blue Jackets.
Thankfully, Don Waddell has already given us some insight into his potential approach so we have some more angles to investigate.
After, and somewhat tied to the burning questions and lessons learned, I’ll do some season reviews and potential roadmap type content for individual players. At the moment, I’m looking primarily at Cole Sillinger, Adam Fantilli and Yegor Chinakhov as I believe they had the most information from which to draw information.
Perhaps looks into Kent Johnson, Kirill Marchenko and potentially Alexandre Texier or David Jiricek as we move further into the offseason. It would be fantastic to get back into video.
Then, it should be onto hand-tracked draft content (from Mitch Brown and Lassi Alanen) and a potential look at offseason roster targets based on what we learn about areas of priority for team improvement. Perhaps the plans will be ruined by early activity from Waddell, but in the meantime I’ll keep plugging along.
If I find time, perhaps some sort of armchair GM roundtable or livestream content.
Thanks for reading :)