Tracking Project Complete Debrief -- Way Too Many Charts
Total Contributions, Season Total Stats, Offensive Contributions by Type (Rush/Forecheck/In-Zone), Deep Dive on Tipping and Rebounding, Individual Contribution Conversion Rates in Higher Danger Areas
Though the Blue Jackets season is over and the stories largely told, I have now had time to organize and clean up the tracked data. Though there were snippets in the season end review, I think there’s plenty more to learn from deep diving this data. That will serve as the primary focus of this breakdown, to see how far we can probe into the data and see what patterns might emerge.
It’s going to be a ton of graphs, some more experimental than others, and maybe light on strong conclusions. Simply displaying this information comes with some analytical risks. Largely, the NHL PxP has been mined for significantly more powerful information and refined to create clarity and certainty through true analytics research.
Most of these charts and data points aren’t adjusted for deployment or role asks and certainly not validated or tested for any sort of efficacy. They should be treated as informative but not necessarily definitive in any way. Just because a player is leading in one category doesn’t necessarily mean they were the best without further verification or corroboration from other metrics, which I won’t be doing here.
Furthermore, these stats are mostly all created exclusively within the Blue Jackets context. We can do some work to compare how the Blue Jackets players compare with that players did against the Blue Jackets but without full data from all teams it’s difficult to know whether a players’ performance is good relative to the Blue Jackets or good relative to the rest of the league. For all we know, the Blue Jackets could be elite creators of certain circumstances as a team and therefore an average player actually good relative to potentially available replacements.
What we’ll have, which can be dangerous, are our instincts from watching closely all season. Perhaps this data reinforces those notions or asks us to think differently, perhaps it leads us astray.
As with all DataWrapper visualizations, the links will be completely interactive. Ideally, you can hover within this window and examine the color blocks to you hearts desire. If you’re on mobile, the graphs are likely a bit more difficult. You might have to click through to the link and the proportions might be unkind either way. Apologize for that in advance, there are certainly going to be some proportion and legibility crimes committed today.
Let’s jump in.
Bird’s Eye Overview
This chart is perhaps familiar to you if you read the season review. Conclusions? Shooting close to the net is where the Blue Jackets scored more goals. Tremendous insight.
I say that tongue in cheek but it’s still worth mentioning that their relative rarity was still compensated for by their threat. Said differently, though middle-distance shooting is a powerful tool the decreased difficulty in scoring but greater volume did not create more goals on the balance of the season.
The idea of the tracking project was, in part, to tell the story of how each team and/or player creates their offense. The theory was that certain types of offense are more likely to result in goals and coaches might design systems and structures to emphasize or prevent these situations.
If I were a coach, I’d conclude that getting to the most dangerous area is still critical. I’d be very interested to examine how a team like Seattle, who seems to emphasize middle distance shooting, performs on the balance of their season as well.
Without the context of the league, it’s hard to tell how the Blue Jackets may or may not be different, but we can check what they conceded to take a first stab at the differences.
First, we can observe that in-zone, for the Blue Jackets, was the wealth of volume. That forecheck was the least voluminous. Though it felt like the emphasis was certainly on forechecking, they created far more shots/chances/goals from rush situations. In fact, rush situations look more dangerous primarily because shots are more likely to come from dangerous spots on the ice and they get on net more as well.
Here we can start to see some of the differences between the offense created by the Blue Jackets and the offense conceded by them.
Primarily, the difference in point shots and tips in-zone. The Blue Jackets, for the most part, shot from “exterior” over the point, seeking to improve the condition of the puck and scoring more often in both point and exterior situations, while also tipping more and scoring even more from tips.
In nearly all situations outside of those, the Blue Jackets had less volume but scored more goals. We know this to be true because we know the Blue Jackets had fantastic finishing and the most 5v5 goals in the league.
Keep in mind that the percentages listed by goals is not Shooting % which tracks how many goals you score relative to the shots on net. These are the percentage of shot attempts that end in goals.
Shots closer to the net score more but also end up on net more. Rush shots are almost always better at getting to the net. The Blue Jackets shot about the same from the homeplate area but did have a small uptick in forecheck home plate shots.
For the most part, it’s easy to justify these outcomes. Goalies are challenged more and have less reaction time from in close, there are fewer and/or no defenders between shooters in rush situations, forecheck turnovers might lead to quick change of sides and or below the net passes which increases middle distance efficacy.
The reality of each situation is that there are plenty of other factors, pre-shot movement, passing, screens, etc that influence shot outcomes. Rush and forecheck turnovers are shorthands for these sorts of situations (though I chose them to help enhance analysis of tactics and the interactions of systems, moreso than pure scoring situations) and not complete descriptors of the danger of a situation.
It appears, at least, that shot location still holds plenty of the requisite information.
Inner slot attempts finishing at 18% seems really high but without more context it’s really hard to say. What’s perhaps interesting is that the Blue Jackets didn’t finish their “exterior” shots at a greater rate, in fact lower shooting%, than their point shots. They did, however, get them on net more often which opens up plenty of follow up questions.
It’s important to not think that this type of analysis is the end of evaluation for any shot attempt. For each of these, there’s a certain cost/benefit analysis that would be required to truly plunge the depths of any offensive decision.
Perhaps exterior shots, ones from closer to the circles or any that are outside of the dots/homplate area (dead angle from the wall being particularly common), resulted in more rebounds and were therefore more worth it. Perhaps they have clearer sight lines for goaltenders, or the Blue Jackets were unusually good at screening point shots with their beefy forwards, and so they are actually worse. Perhaps again exterior shots tend to be on net and are recovered more where point shots are blocked and lead to counterattacks more often.
I have the data to answer many of those questions, in theory, but that will be a task for later this summer.
So, the comparisons to this point opens up some specific research questions: Why do the Blue Jackets shoot more from the exterior than the point? Why do the Blue Jackets tip so much in-zone and score from these situations more frequently? Why do the Blue Jackets score so many rush shot inner slot goals and why more efficient scoring from tipping off the rush than in other situations?
In-Zone Offense
While I now intend to deep dive specific situations, I also intend to make each chart multi-purpose. While evaluating shots is the primary method of evaluating teams, so as to not exaggerate the importance of a shot and an assist combined, for players I think it’s important to value shooters and playmakers at the same time. We don’t really care whether a person passed to the slot or shot from there but rather that a slot shot was created in the first place.
Utilizing contributions and rates does depart us from the use case of examining “why the Blue Jackets shoot more from the exterior than the point, and why they tip more than other teams. Theoretically, I should just use shot volume over the course of the whole season to make that point specifically. Instead I’ll not do that and bring insights from it anyway.
The clearest answer to the first question, why exterior shots over point shots, is Zach Werenski. He’s the only defenseman with that relationship and, knowing what we know, played an astronomical amount of minutes. His habits of activating down the wall to shoot from the half wall exchanged the position of the shots though whether or not those were better is yet to be determined.
For the most part, defensemen shoot further away and, in in-zone situations, do bring the relative volume that any individual forward does. Where they differ is in their access to the slot. Before we go further, important to acknowledge that 1. Zach Werenski creates chance danger like a forward does, 2. Damon Severson and Denton Mateychuk are steps ahead of the rest of the D corps. Where Mateychuk falls behind, and Provorov shines, is in creating inner slot offense.
For the second research question, it appears we have at least the beginnings of a satisfactory answer. The Blue Jackets, especially in-zone, got incredible tip contributions from Boone Jenner and Dmitri Voronkov. Jenner’s rate performance is likely distorted by his relative small sample (as will be LDBB and Kunin who have the lowest samples by far) but his inner-slot + tip contribution volume exceed most other forwards’ full homplate + contributions. In-zone, these two’s net-front skillset is unmatched on the roster.
The rest of the “significant tippers” are James van Riemsdyk, Zach Aston-Reese, Justin Danforth and Mathieu Olivier. That’s three members of the late season “identity line” and a very specialized member of the Blue Jackets.
If you’d like to know why the Blue Jackets were relatively great tippers, it might just be that in-zone they had a strong identity that they played to in specific situations. Perhaps the Blue Jackets overperformed in part because 1. in-zone offense is fake and doesn’t contribute to that many goals and therefore their struggle in this specific area distorts plenty of their metrics and 2. they had good role players who were playing to their strengths.
I don’t fully ascribe to either of those conclusions but I don’t reject them outright either. If the Blue Jackets want to take a step, and especially if they want to win a Stanley Cup, they’ll need to attack dangerous areas from controlled situations even better. They’ll need their tip and net-front players but also their high danger playmakers who can play through possession.
From that perspective, Kirill Marchenko certainly looks ready to go. Sean Monahan, Adam Fantilli and Kent Johnson performed well in terms of distributed danger, performing on par with JvR, but Marchenko stands alone in his capacity to threaten dangerous areas. Perhaps he’s less efficient, and therefore not as judicious, but it certainly looks like he’s capable of carrying the load in in-zone cycle situations.
Where Kevin Labanc and Yegor Chinakhov stand apart is their desire to pump shot volume but inability to access the most dangerous spots in-zone. If placed alongside skilled players with a desire to work possession into dangerous areas, there’s perhaps some tension. If placed alongside players who are good at fishing rebounds, perhaps there’s some synergy.
Deeper into “Tips” and Rebounds
One of the problems with “tips” is my intentional binning of the data. The goal was to appropriately separate these types of “shots” from their position on ice. From a rigorous standpoint, defining the “shot-types” or in thinking through the lens of trying to create a more perfect xG model, separating out “high danger chances” from “pad-stuff attempts” is difficult even if we have full player and puck tracking.
Fundamentally, the shooting % turns out to support the definition but that’s not necessarily the justification I was chasing. I wanted, and this explains why I tracked only 5v5 and not necessarily 3v3 or 4v4, to understand the tactical choices available or how shots interacted with personnel and potential systems. We’re a bit away from that, but I think the next part is important to understanding why binning this all together may not have been a great idea.
Tips is perhaps a bad word and one that I will change provided I continue with this type of binning system. There are some tips and deflections and even differences between those. A point shot with a stick swung at it is certainly a tip but should that qualify differently as a pass through the crease banked off a stick for a backdoor tap-in? If we use the same word, we lose a lot of information.
Furthermore, I wanted to count “putbacks” and high danger chances off rebounds as different things. I wanted to find a way to account for attempts that were purely “pad stuffing” with little intention put into their taking. Jamming a puck into a sealed post is sometimes categorized as one of the most dangerous shots possible to take while clearly having limited probability in becoming a goal (primarily because they are classified as “snap shots” and location is not always tracked well by official NHL shot trackers).
At the end of the day, I was left with a subjective measure that I feel reasonably confident in but one that, like with plenty of other aspects, should be improved next season. Part of the objective was certainly to capture enough information but to keep it simple and digestible without too many data points either. The line between the two will be something I continue to work out.
So, to bring it back to the topic at hand, I do have a neat little way to separate out “putbacks” from the rest of the types of tips, deflections, pad-stuffs or what have you. I tracked the context and game-situation for all of these goals but also separated out rebounds from the rest. Therefore, it’s highly unlikely that a “deflection” from came post-rebound and therefore most of these situations can be classified as “putbacks” (noting that rebounds aren’t always exclusively tips but can also be inner slot shots or homeplate shots more rarely).
Here are the percentage distributions of tips vs putbacks in in-zone situations. The colored bars represent goals and the grey bars represent the rest of attempts whether missed/blocked or on-net. Putbacks and pad-stuffs can really only ever be on net. If my subjective tracking is perfect, each of these “putback” goals represent how often simply tapping the puck at the net without intention results in goals.
At this point, we have now reached the depths. The data can’t really get more fragmented or specific, so let’s proceed with caution. The players are ordered by “in-zone tips/60” so the higher up you go the more often they tipped the puck per their minutes. We’ll get into totals shortly.
Kunin and Del Bel Belluz have the smallest sample sizes, so their more extreme results make sense. The highest volume tippers generate the most away from rebounds though Zach Aston-Reese veers toward the highest stable percentage coming from rebound putbacks.
Otherwise, a hard to tell too much from this single graph. Players, outside of Sean Kuraly who doesn’t in-zone tip all that often, generally tip/deflect/wraparound more than they get rebound putbacks.
Justin Danforth appears to be the most voluminous “rebound putback” player and, perhaps, worst actual-tipper on CBJ. Does this mean his individual xG is distorted? That his on-ice metrics don’t reflect reality?
In-zone contributions, especially drilled down to the level of exclusively tips, wouldn’t be enough to make that sort of claim but it is worth thinking about especially given the way Olivier and Zach Aston-Reese were deployed alongside him. It’s also worth mentioning that rebound putbacks are only the “worst case” or “low skill” representation of shot attempts that come from rebounds. If I judged that the player put some intention into the “shot”, most commonly by elevating the puck, it wouldn’t show up in this data.
The most interesting stat, perhaps, is that Kent Johnson scored on rebound putbacks and those were his only attempts. Either he’s doing something special or he’s really lucky. What I can say for sure is that he is indeed good at actually tipping and/or deflecting pucks into the net. He has incredible hand-eye coordination but doesn’t spend much time at the netfront.
Otherwise, Olivier, JvR and Monahan were the most efficient actual-tippers among the team. This would suggest there is at least some degree of acuity in their net-front games.
This data represents that actual totals and where it perhaps serves to start breaking apart some of the more team-wide statistics. The players are still rank-ordered by their “rates” but now their contributions can be more accurately judged based on how they contribute to the full-season. I think it’s an adequate visualization of the sample sizes and fragmentation that we’re working with. Even the best player on the team at scoring tipped goals only scored 5.
In total, the Blue Jackets scored on 8.5% of their in-zone putback attempts. 7 putback goals, 82 putback attempts. As far as non-rebound tip-attempts, 20 goals and 233 attempts. A scoring rate of 8.5%. Perhaps my instincts to bin the “low quality” inner slot attempts was correct after all.
Scoring on 8.5% of attempts is still extremely good, especially for in-zone attempts, and grades out as still the second most dangerous type of shot a player can put on-goal.
Mathieu Olivier and James van Riemsdyk grade out as the most voluminous non-rebound scorers, with plenty of players finding 2 goals. Boone Jenner only played for the last quarter of the season but still contributed to the fifth most in-zone tipping on the team. He has a clear identity and clear strengths on the ice.
Defensemen don’t really contribute to taking tip or rebound chances in-zone (though now I’d imagine it’s interesting to see how many they set-up). Perhaps this is one of the most critical reasons that defensemen can’t produce offensively like forwards, they don’t hang out in places where even “low-skill” shots on net result in a goal on nearly 10% of attempts.
Zooming Out with Tipping and Rebounding
Now, I think it will serve us to split out rebounds and move away from exclusively in-zone variants. These other situations tend to be more dangerous as a result of more broken defensive structures. In each of these situations, finishing from tipping was greater (though only slightly on the forecheck).
Some of the same patterns emerge. These are shots only without passing or other contribution data. Rebounds, in particular, aren’t association with passing data (unless there was a pass off the rebound that resulted in a shot).
The data expresses largely the same sort of information save for a few differences.
Dmitri Voronkov is king “poor” rebounder but also the best rush tipper. In fact, the only rush tipper with any real margin. Danforth, Monahan and Olivier grade out here as well.
Jenner, and Kuraly, Danforth and Aston-Reese find plenty of tip situations off of forecheck situations, perhaps these are off-wall net jams but I believe plenty are from defenseman keeps into actual slot tip chances. This specific forecheck tip play, of an exit turnover into netfront attack, is critical to the CBJ tactical gameplan.
It was quite interesting to me that Kuraly grades out so poorly on in-zone tips but is perfectly fine in rebounds and forecheck tips. This, to me, suggests that his tipping comes from putbacks or stuff attempts rather than actual tips/deflections.
Jenner, Voronkov and Zach Aston-Reese have the most potential to upgrade their net-front scoring threat. Getting these high xG opportunities is already good but converting them into “inner slot shots” with a little more presence of mind could dramatically increase their offensive effectiveness (or not, we’ll see when we move into rebounds).
I think this also paints the young players in a particular light as well. Fantilli, Kent Johnson and Cole Sillinger are all on the bottom of the team in this particular style of shooting. Perhaps it’s just not in their bag or perhaps they upgrade their style of shooting more frequently and that’s why they feel like higher-end players. It’s worth mentioning that Sean Monahan and Kirill Marchenko were all indeed better than those three but also certainly worse than the rest of the “veterans and Voronkov”.
Now I’ll have to leave the prettier original CBJ color scheme behind and commit some eye crimes to continue working with variations of the electric green.
Instead of continuing forward with tips, and perhaps examining why the rush finishing rate was so much better (I’m guessing it’s sample size) we’ll move forward to rebound “quality”.
If we operated under the assumption that the higher skill cap players in Fantilli, Kent Johnson, Monahan, Marchenko converted more of their rebounds into “high danger” chances we would be at least partially correct. Each player had more “high danger” than they did “tips” save for Cole Sillinger. Yegor Chinakhov falls into the more high danger bucket as well.
Kent Johnson was one of the best “high danger” rebounders and, to be honest, could have earned even more production from this space(or perhaps this is some tracker bias projecting skill onto equal attempts). His net crashing habits are great and he often wasn’t rewarded on these timely runs. I don’t want to turn this into a Kent Johnson breakdown but I’d love to see some more zone-time and more motion from him that helps him get back to some flash screening and tipping, precisely because of how intelligent his approach in this area is.
Boone Jenner serves as the ultimate outlier. It’s difficult to say too much with his relatively small sample but his net-front tenacity is clear and fits strongly with what we’ve seen from him for years. He’s a shortcut to dangerous offense and he’s endeared himself to coaches and managers alike for exactly this reason.
Danforth, Kuraly, Olivier have a remarkably low number of true high danger rebounds. They crash the net, or peel the puck off the back wall, and jam like mad.
Sillinger, Pyyhtia, Lebanc and Kunin really bring little to a rebounding game. It’s one area where you’d really hope to see Sillinger grow given his stature and likely role fit going forward. You can’t have a contending 3C get out-rebounded by a defenseman as unique as Werenski ultimately is.
Now, we’ll slide on over to full rebounds outside of the context in which they were created so that we can evaluate how dangerous they are on their own. I’ve organized the players by rates but displayed totals. I hope this helps illustrate why I have been binning the outcomes into goals and non-goals but for the purposes of this exercise I think it’s important to include the information despite it making the graph hard to parse.
First, let’s acknowledge that the full season shot volumes are still quite low. The highest volume of rebounds comes from Dmitri Voronkov with 27 over the full season. Far less than 1 per game.
The Blue Jackets, teamwide, scored on 19.7% of their high danger rebound attempts, with an actual shooting percentage of 33.3%. If we were to translate these attempts into “expected goals” (a very bad idea considering the hyper specific context of this team) that would mean the average high danger rebound was worth 0.33 xG (xG are calculated on fenwick and I only tracked on-net or a combination of blocked and missed so this isn’t exactly correct either). Putbacks, however, scored on 9.9% of all attempts with a 13.2 shooting percentage.
Without diving deeper, this suggests that a rebound attempt with intention added (puck elevation or specific placement that I could see) is one of the most dangerous shots that the Blue Jackets created this past season.
Dmitri Voronkov and Zach Aston-Reese, the most voluminous rebounders, didn’t score on any of these high danger attempts. ZAR, in fact, scored on exactly none of them. Unlucky? Perhaps.
Sean Monahan, who had the highest rate of high danger vs putbacks, also scored on most of them. I’d like to think that his presence of mind in small space and under pressure contribute to his stellar ability to score on these situations but it’s also possible that his scoring lead to biased tracking.
Kent Johnson, on the other had, didn’t find a way to score on any of the high danger attempts but did score on all of the putback attempts. Perhaps not score induced bias or perhaps I’m missing some of the on-ice brilliance of Kent Johnson.
I think it’s important to consider that, although the skilled youth lack volume of rebound attempts, Kent Johnson and Adam Fantilli were above the high volume rebounders if we consider their high danger attempts exclusively. This isn’t necessarily a good thing, even those putback attempts turn into goals, but it might illustrate that they don’t need the volume to find the results either. Perhaps their open-ice play, instead of being constrained to a more stationary netfront, results in higher quality than quantity. Perhaps they simply score on the first shot enough that there aren’t as many low-quality attempts available.
I think there’s tremendous potential to get deep and analyze the nature of rebounds/quality/finishing and how they all effect eachother and on-ice chance shares but that’s certainly outside the scope of this project.
Goal-scoring Through the Lens of Rebounding
Having gotten a bit deep on rebounding and it’s relative danger, I thought it wise to offer a full team perspective. When I displayed the stats post-game, I offered rebounds as their own breakout metric but largely included them within the game-state they were created. I think a rebound from a rush shot should also count as rush offense. Now, at the end of the season, perhaps it makes a bit of sense to break it out entirely and see how that effects goalscoring.
What I’m going to do with the next chart will likely not be friendly to anyone on mobile, so I apologize for that. DataWrapper really didn’t like the way I did it anyway but I think it’s a fun timeline and worth looking at at least on it’s own.
So, first things first, I didn’t start breaking out rebounds until after 10 games. That’s going to give us a distorted look at the nature of game-state creation. Second, it’s really tough to tell much given game to game noise.
The reason I created this timeline was to examine whether or not “end of season” hockey looked fundamentally different from beginning of season hockey. We always hear about how important it is to get to the dirty areas and theoretically defenses should be dialed up and structure therefore made stronger but it doesn’t look like there are any large changes in shot generation as the season wore on.
Charting only “percentage of shots” of a given offense type does introduce some difficulty in viewing trends, perhaps the team recovered more shots and beat a team that way, but it doesn’t look like the Blue Jackets really changed their performance in an obvious way. Fewer rebounds through the last 20 games, maybe a little more forechecking? I should probably graph each of these separately or as a smoothed line graph to look for stronger trends.
From a similar concept, we can examine 5v5 goalscoring.
First, a little stats dump on the importance of rebounds. In-Zone shots represent 46.8% of the total shot attempts the Blue Jackets took this season, Rush Shots 31%, Forecheck Shots 15.1% and Rebounds 7.1%. In terms of goals, In-Zone offense accounted for 37.5%, Rush for 29.5%, forecheck for 18% and Rebounds for 15%.
Although rebounds accounted for only 7.1% of the shot attempts the Blue Jackets took, they accounted for 15% of the goals scored. That certainly explains the emphasis on getting to rebounds from plenty of coaches.
In observing the timeline of goals scored, there don’t appear to be any dramatic trends to observe. I would contend that the Blue Jackets scored more forecheck goals in the second half of the season, and perhaps dipped in-zone through the middle of the season, but otherwise rebounding and rush goals appear quite steady at first glance. On closer examination, rush goals may have decreased in volume over the course of the season. Funnily enough, the biggest and most dominant total 5v5 goalscoring games came at the very beginning and very end of the season.
The change in forechecking, if I could be permitted some at a glance analysis, looks like it corresponds with Monahan’s injury absence. I would suggest that the coaches and team really bought into forechecking structure without their top center and that might explain some of the increase. Perhaps it came from specific players (Fantilli? Boone Jenner? the Identity line?) but I think it was likely a team effect at the same time.
That general “middle of the year” trough in goalscoring roughly corresponds with the total shot volume. I believe we saw plenty of this trend in the end of season review, where some spurts in November looked pretty good and some re-emergence in February/March, but worth mentioning that the goals did follow the attempts/possession as well.
Back To Offense Types
I’ve been meandering quite a lot while exploring the different aspects of the tracked data but now I figure it time to get into the more interesting types of offense. The idea in differentiating “forecheck” and “rush” over in-zone or other types of offense was that these would represent shots taken against broken defensive structures. I don’t think it succeeded to the extent that I’d like, and I do have ideas on how to improve from there, but there’s still plenty of signal in differentiating the shots.
Forecheck
Given the importance of the forecheck as the season went on, I figured it would be important to start here. Forecheck offense accounted for only 18% of attempts (outside of the slower type of forechecking) but outperformed that in terms of goals.
We are, once again, fully back in the realm of forecheck contributions and not purely shots taken.
Luca Deb Bel Belluz forecheck god! His small sample certainly wildly distorts the rate metrics but there are a couple of other cool conclusions to pull.
Yegor Chinakhov, especially in his work early in the season, created a ton of forecheck chances. He was part of the most chance-share dominant iteration of the top-line alongside Monahan and Marchenko. His impact diminished with his injury and he never found his footing after his return but let’s not forget that he has the potential to be an absolutely special forechecker.
Otherwise, the top contributors were the usual suspects in Sean Kuraly, Justin Danforth and Zach Aston-Reese (all creating a large volume from tips). Interesting to note that top net-front tippers in Olivier, JvR and Voronkov all lagged far behind in direct forecheck contributions.
Marchenko, Labanc and Sillinger all created the most volume of homeplate area shooting chances. I suppose, at this point, it might be interesting to know whether they were taking those chances or passing but perhaps we get into that in a passing section specifically.
If we consider volume of “dangerous” chances we must conclude that Marchenko and Chinakhov were two of the most dangerous and structure breaking forecheckers on the roster.
It’s difficult to evaluate defenseman performance in specifically this context. Are they follow up shooters activating after a forward has stripped a puck (Severson goal from Marchenko and Monahan stripping Dahlin in the early Buffalo game comes to mind) or are these contributions coming just after a pinch or blue-line keep that they’re passing or skating into more dangerous areas?
In any case, it’s the same defensemen who stand out again. Zach Werenski is once again the best, Denton Mateychuk second in danger generation.
Rush
Rush offense is where we start to see some of the more skilled players and 5v5 producers differentiate themselves. I think, or at least postulate, that rush offense is so predictive of scoring primarily becase it is within the players’ control. If a player can reliably deceive or skate through the neutral zone, they can create open-ice in ways that have less to do with the defense than any other situations.
In any case, the top offensive performers are the top rush contributors: Chinakhov, Fantilli, Marchenko, Kent Johnson, Boone Jenner and Sean Monahan.
Early season Chinakhov was, quite literally, something different. He’s still a very specific and valuable type of player despite his offensive zone approach being a little bit warping. I think it’s possible that the front office moves on from Chinakhov but they should be careful in trying to find a player who can skate and create across the dangerous types of offense.
Fantilli, Marchenko, Johnson and Monahan all created nearly the same amount and danger of rush offense. I believe you’d expect Fantilli and Johnson to improve in this area specifically, I’ve outlined some of the things they coudl do better on numerous occasions, but it’s still good that they are so young and at the forefront of the Blue Jackets.
Boone Jenner was perhaps surprisingly good though he did struggle to get to the most dangerous areas. He has a tendency to create puck volatility and also to put pucks directly on net. Not necessarily ideal for building rush danger.
Kevin Labanc continued to be a very specific type of player which is perhaps one of the least valuable. He shoots a lot, from distance and doesn’t get it on net all that often. This is perhaps why he’s fallen down lineups and rosters after posting some impressive early seasons.
James van Riemsdyk is the best of the rest. You’d have to be a bit disappointed in Cole Sillinger, especially considering his time spent with Kent Johnson, but perhaps late season and time on lines with Olivier and Aston-Reese did much of the damage. He did indeed create off of the forecheck (and his rush assist on an Olivier goal was an incredibly good layer pass and surprising skilled play from both) which might suggest he was highly affected by linemates.
The elephant in the room is Zach Werenski. If he’s looked good in the other offense creation metrics, he looks truly worthy of his stature in his capacity to join the rush and create chances. He produced dangerous chances at a rate on the level of the top six forwards. I can’t really say enough about his rush brilliance.
Shots and Passes
One of the unique aspects of the tracking project is that every shot taken by the Blue Jackets is associated with a passer. What this means is that, theoretically, we could study the most prolific and common shooter and passer combinations.
It’s difficult, ultimately, to present this data while also accounting for their time-on-ice, at least given the tools that I have. Ideally, there are better ways to do this. Until I have more control over the visualizations the following complete crime of a table is all you’re getting.
The row is indicative of that players’ shot attempts from the column players’ passes. The table is sortable which means if you click or tap on a players’ column, you can see how many passes they completed that resulted in corresponding shots from the other players.
There are some notable interactions.
Primarily that Sean Monahan passed to Yegor Chinakhov more than Dmitri Voronkov. Voronkov passed to Adam Fantilli for shots more than double what he did for Sean Monahan. I imagined that Monahan and Voronkov had an excellent in-zone partnership at the net front but perhaps that was exaggerated. Perhaps it wasn’t and their chemistry was obscured by rebounds not coming with assists.
Mateychuk assisted on 14 of Fantilli’s shots but only 5 of the next most common forwards (tied with Marchenko, Kent Johnson, Mathieu Olivier, Justin Danforth). If those two find some chemistry that would be massive for the Blue Jackets developing multiple waves of attacking deployments.
Ivan Provorov assisted on 18 Zach Werenski shots, 17 Kirill Marchenko shots and less than 10 of any other player. He only played with Werenski until Fabbro joined the fold which is telling of how little chemistry he developed away from him.
Kent Johnson assisted on 31 Adam Fantilli shots, the most of any forward pair outside of Marchenko’s 37 to Fantilli and 35 to Voronkov or Monahan’s 31 to Marchenko.
Zach Werenski assisted on 50 Dante Fabbro shots, a lot of faceoff and point shots to contribute to high totals, but then 34 to Kirill Marchenko and 21 to Kent Johnson.
Feel free to click around and look for any relationships, I’ll get to work on finding a better way to illustrate this data and maybe find more as I go through this offseason.
High Danger Shots and Passes
Up to this point we’ve examined the finishing, etc, from a variety of types of offense and even taken deep dives across tips and rebounds. Now, we’ll hit the rest of the dangerous types of offense (and include passing), namely inner slot shots and homeplate chances.
Not only did Adam Fantilli shoot the most from the inner slot, he also scored the most! Those two things perhaps sound associated, and they are, we should commend Fantilli for his shooting prowess, scoring 17 goals at 5v5 from high danger areas is a fantastic total. A 20 year old who knows how to get open in the slot and score from there is a very good thing! It’s not represented here but Fantilli was particularly dominant in rush shooting situations. He scored the most goals on the team in this specific capacity and it’s a strength the team should certainly lean into.
Kirill Marchenko, on the other hand, was nearly a similar capacity shooter. He didn’t score as many rush goals as Fantilli but did score more in-zone goals and topped the team in that aspect. Playing time associated with Sean Monahan certainly helped, but we’ll get to that in a second.
Chinakhov and Monahan weren’t dominant in the inner slot the way Marchenko and Fantilli were but they got there in heavy volume. Perhaps that means a repeated performance in an alternate universe would have seen them finish better but maybe it means they just don’t have the pure finishing skill that Marchenko and Fantilli do.
Kent Johnson rounds out the top tier of inner slot shooters and his performance, though he did score plenty, leaves the most to be desired. He didn’t score as well and didn’t get pucks on net with the same proficiency. His deception is top tier but to improve he’ll need to make some better decisions or prepare the offense better prior to shooting.
The next tier of point creators for the Blue Jackets, James van Riemsdyk, Boone Jenner and Dmitri Voronkov, all created give or take the same amount of inner slot shots. I’ve covered their respective strengths plenty but worth mentioning that their skillsets allow them to create outside of purely dangerous shooting.
The rest of the Blue Jackets’ bottom six rounds out the bottom. Sillinger’s offensive acumen has been questionable in these metrics but it looks like we shouldn’t expect offensive breakouts from Olivier or Danforth, should they stay on the team, either.
Playmaking is something that’s a bit harder to evaluate but also share some of the effect on finishing. These passes to dangerous areas are only graded on their ultimate shot outcomes, whether in or on the net, which also doesn’t filter the passes that weren’t quite completed or were complete misses. There’s a layer of noise over top because passes require an additional step of execution to turn into goals.
Kirill Marchenko and Sean Monahan stand head and shoulders above the rest in their ability to pass to the inner slot for goals. Monahan was one of the best 5v5 primary playmakers in the league and it’s very easy to see why. In most cases we can’t be sure whether the pass quality resulted in goals or being on net but I think having watched Monahan we can say that his “setting”, or pass quality that leads to improved finishing, is extremely good. His poise under pressure and small space vision is stellar.
Kent Johnson, the next most voluminious inner slot playmaker, perhaps deserved a few more primary assists as his on-net vs miss percentage was still pretty high all things considered. Where he was ultimately more efficient he was surpassed in volume by Kirill Marchenko and then also passed in ultimate finishing of said passes. Truly evaluating “who was better” might require the tradeoff of said missed passes. Would you rather have the volume or the efficiency?
Zach Aston-Reese and Yegor Chinakhov grade out as similar inner slot passers. They fed the slot well and their teammates put the shots on net but they didn’t result in goals that often. An effect of variance, linemates or pass quality? Difficult to say.
While I think this, and perhaps some of the other metrics we’ve seen up to this point, paint a good picture for what Zach Aston-Reese brought and advocates for his continued inclusion on the roster, it’s worth comparing him to James van Riemsdyk who was one of the more efficient 5v5 rate scorers in the league. His slot-work, both shooting and passing, was much better than ZAR. He has a specialized identity that worked for the Blue Jackets and it’s at least intersting to find one of the underlying mechanics for his success borne out in data.
Where Adam Fantilli is dominant as an inner-slot shooter, he leaves plenty to be desired as a slot passer. His passes tended not to make it on goal though he still did finish with a better primary assist rate than Kent Johnson. Is he simply taking more chances? In order for him to take the next step, perhaps feeding him with the best slot passer CBJ can find is a good idea but his playmaking will need to improve if he wants to be a top of the league superstar.
Zach Werenski really stands out among defensemen, surprise, for his ability to make a volume of passses into the slot. Mateychuk and Severson, no surprise, grade out as the next best options.
To be perfectly honest, this visualization should probably have stayed on the cutting room floor. This is a complete representation of each inner-slot contribution based on the offense type. It completely requires the DataWrapper interactivity to make any sense of especially when the bar slivers are so small
Luke Kunin remains at the top because of his significant non-goal contributions in extremely limited minutes. Otherwise, the players are rank-ordered based on rates but their total season-long contributions at 5v5 shown.
Kirill Marchenko remains king of the inner slot, especially in-zone but even in forechecking too. Adam Fantilli, no surprise, is a dominant rush creator. Zach Werenski, especially as a defenseman, absolutely crushing high danger rush contributions is magnificent.
The most eagle-eyed of you might have noticed a new metric, if not on the actual chart, in the key. Rebound Inner Slot Assists. Most of the time, in fact nearly all of the time, rebounds don’t come with assists because they are created by shots and loose pucks are put back toward the goal.
In the special cases where a player had the presence of mind to find other options, literally four total all season, they passed the puck directly off of a rebound and into a shot. It’s a high level tactic that creates shots that are nearly impossible for a goalie to adequately react to. In any case, Dmitri Voronkov and Zach Werensk each had one that resulted in a goal and Ivan Provorov and Kent Johnson each had one that didn’t.
Homeplate Area Passing and Shooting
Since we’ve covered the rest of the dangerous areas, now I think it would be a good idea to examine the less dangerous, more distant shots, and see how players performed within this context.
Homeplate area shots, those outside of tips and the inner slot, were not commonly converted into goals at the same rate that either of the two closer shot types were. That said, these shots do come with the capacity to score and excellent wrist shooters can certainly take advantage and beat goalies clean. The best teams in the NHL create numerous inner slot chances but they also finish mid-range chances well, depending on system.
Yegor Chinakhov has been heavily involved in plenty of the shooting metrics but this one is perhaps where he shines the most. We know he has a powerful wrist shot, choosing blistering pace over quickness and placement in many cases, which sometimes puts him at odds with the best finishers in the NHL. Still, his capacity to score goals and create these chances means he was one of the best homeplate goalscorers on the team. It might be better to work to improve shot location, in which case Chinakhov’s approach might not be the best, but if you put him on a line that doesn’t want to build danger his mid-range threat could be an excellent option as a primary threat on the line.
Of the higher sample players, the skilled players once again represent well in terms of getting shots on net from distance but they’re each different in terms of actual scoring. Kent Johnson, the least voluminous of the skilled trio, scored the most. A signal of upcomign regression? Or a sign that he is extremely deceptive, perceptive and intelligent with his read of goalies and in getting pucks in places they can’t stop? Likely the first option but I wouldn’t rule too much out.
Kirill Marchenko and Adam Fantilli are the two other significant forwards in terms of mid-range scoring. Fantilli generated more chances on goal but scored fewer. Perhaps, as difficult as it seems to imagine, he’ll find a way to score even more next year. As much as Fantilli is a lethal goal scorer, I think his advantage comes from hands in-tight more than it does an overpowering shot. With improved shot selection, we might see a boost here too. If he finds chemistry with Kent Johnson, perhaps he can unlock his one-timer which has been mostly absent outside of extra man situations.
Kirill Marchenko, king of danger and volume, continues to put himself in great positions to score good goals. Different from some other players, however, was his role as the focus of plenty of faceoff set plays. Pretty much all of these count as chances and might explain some of his boosted volume. Many were blocked, so perhaps that explains Fantilli’s increased ability to get pucks on net.
The most surprising information, at least to me, was Werenski and Mateychuk’s mid-range shooting performance. Especially surprising considering Mateychuk outperformed Werenski as a finisher. Werenski is still clearly a centerpiece but I have to wonder if his raw finishing dominance was hampered by recent shoulder surgeries or perhaps fatigue late into the season.
Both he and Mateychuk (and Fabbro for that matter) finished better (from an attempt scoring %) from this distance than all but the top offensive forwards. While it’s difficult to say without further diving, I’d guess that this represents a structural pattern. CBJ loved to collapse space, work the puck high and have their weakside defensemen short the zone. Is this the full explanation for great middle distance shooting? I wouldn’t be comfortable saying.
This type of shooting will remain a key part of any Dean Evason team and it’s nice to see the most important future members of the Blue Jackets have the skills to thrive. Perhaps, mid-range shooting (or high zone playmaking to unlock a partners’ shooting) remains a key consideration for the offensive component of any defenseman addition.
Similar to the last passing metric, I’d imagine this one is even more vulnerable to teammate or situational effects. I think that getting shots on-net from these passes might be more signal of pass quality, or in hitting a player with enough space to find a shooting lane, but I can’t really say for sure without a bunch more work.
Sean Monahan reigns supreme as a playmaker. There’s not too much to say about it, his combination of ability, presence of mind and good linemates makes him an excellent offensive creator.
What is surprising perhaps, is the next tier of players in terms of pass volume. Jenner can perhaps be excused as a small sample, but James van Riemsdyk and Cole Sillinger still stand out relative to their offensive contributions elsewhere. JvR, if my reasoning that passing to on-net shots has more signal, grades out very well relative to the other perhaps less efficient passers. He is great at sliding pucks through space and onto sticks and I can say, at least anecdotally, that Marchenko did tend to force passes and his ultimate execution wasn’t particularly fantastic.
Sillinger, though he’s had a difficult time transitioning to the NHL for a variety of reasons, did see a re-emergence of his playmaking game this season. He certainly has ideas but his execution is still volatile. If anything, perhaps this was a proof of concept season and there’s a nice playmaking breakout coming in 2025-26.
Kent Johnson’s primary assist game at 5v5 was certainly lackluster. Given this data, alongside his inner slot passing, perhaps he was owed a few more assists. His capacity to pass to shots that were on-net was second to only Sean Monahan. He certainly put a stamp on his dual-threat abilities this season, though perhaps they have always been there given the distribution of goals and assists since his time at Michigan, and his late season play looks ripe with potential.
Adam Fantilli has a tremendous amount of room to improve as playmaker. Perhaps that’s not what we should be focusing on given his shooting prowess, but at the very least you’d like to see fewer passes on missed shots. I can say for sure that Adam Fantilli had a tendency to force passes to the slot without considering the likelihood of it being completed, which wouldn’t show up in this data, though he did eliminate the most egregious decisions as the season went on.
What stands out among defensemen here is that Werenski, Mateychuk, Harris(!) and Severson all created about the same amount of shots from mid-range passes. There wasn’t much to differentiate the players from eachother but neither Mateychuk or Harris were graced with any finishing on the other end.
Though they are obviously specific role players, it’s worht mentioning the struggles of Dmitri Voronkov and Mathieu Olivier. I personally would have expected more from Voronkov who, at times, can be an excellent spatial passer in build-up but certainly lost some of his on-puck rhythm as this season went on.
I wouldn’t say that we do expect Olivier to create a ton of offense through passing, his tipping and back wall work are certainly good enough, but his lack of passing contributions are stark. He’s been at the bottom of many of the other offensive metrics so it shouldn’t be surprising if his contract-year scoring outburst isn’t repeated. If his back wall into netfront jam playstyle is indeed valid, maybe he’ll still keep scoring goals.
In sinning again with the full volume of homeplate chances and passes, I think we get a good idea of how rarely some of these shots and passes turn into points and goals. Few players put up more than one of any one instance with Adam Fantilli and Kent Johnson’s 3 rush goals each being the most of any single count.
What’s notable, however, are Kirill Marchenko and Denton Mateychuk’s scoring off of a rebound from this distance (and James van Reimsdyk passing to Mateychuk for that goal as well). Incredibly few attempts were made at this distance but it’s a bit neat nonetheless.
Back to the Beginning
At this point, we’ve covered pretty much every aspect of nearly every shot that the Blue Jackets have taken this season. There was deeper context that I tracked specifically the style of each type of offense and how the Blue Jackets got that puck in the first place. Given how deep we’ve already been, I don’t have that much appetite to comb through this specific data with a fine-tooth comb but there’s still plenty of interesting information.
Here, we’re looking at the different types of rush and/or forecheck offense and re-combining them into “controlled” vs “takeaway” oriented offense. Corey Sznajder wrote recently about the changing landscape of rush offense and how neutral zone defenses, and defenders, are becoming talented and structure to the point that slower more controlled rush offense is dying. Teams, in many cases, find different ways to produce rush offenses based on their tactics but many of the defense zone structures are organized intentionally around creating counterattacks. (This was one of the central thesis' of Vegas’ shot fronting, collapsed slot defensive zone strategies and one that Pascal Vincent tried to emulate, albeit poorly).
I think the easiest ways to think about Takeaway vs Controlled is that Takeaway type offenses are more likely to result in truly broken structure. I would postulate that the best teams in the NHL have players who can score against set structures (I’m thinking of the way Kucherov or Eichel or Reinhart or Robertson or Rantanen or Tkachuk or Aho create through layers and consistently find scoring opportunities) and I think it’s clear that the Blue Jackets still have a little bit of work to do in that regard.
It turns out, not that we’re surprised, that the Blue Jackets created a ton of offense off of turnovers and takeaways. Defensive zone counter offense created more goals with fewer attempts than any other type of offense. Now, the true analysis goes in both directions. Being more efficient in one area is good but if you can supercede that efficiency with volume, or with safety or ease of creation, that doesn’t mean the efficiency is better.
Still, if we want to analyze quality or proficiency in a certain dimension, scoring efficiency is probably the best way. To truly examine the Blue Jackets’ ability we have to place it in context relative to the rest of the league. If every team is better from defensive zone counter offense, then we aren’t learning too much about the Blue Jackets. Unfortunately, I don’t have all of the teams data for the full season but we can hope that their performance against the Blue Jackets is at least a relative sample.
The Blue Jackets featured more controlled offense and less takeaway offense compared to what they conceded.
If you were an opposition team, perhaps you were okay letting the Blue Jacket have the puck and forcing them to work through set defenses. I think this might explain some of the issues we observed with struggles against structured clubs and the ease of goalscoring against sloppy teams.
This might also have something to do with some relative weaknesses of the Blue Jackets defense, dating back to last season if Waddell’s early season remarks still hold weight, and the relative inefficiency of youth.
The Blue Jackets conceded more goals from forecheck exit turnovers, which were given in volume as part of the retrieval rims preferred by Evason but also by the struggles of the wings lower in the lineup (ZAR, Sillinger, Voronkov, Olivier). I would say that perhaps there was a tradeoff somewhere at play but the Blue Jackets also conceded more retrieval turnovers than they created (though not goals).
In order to evaluate both together, I have combined the graphs and showed only the percentages. The totals next to the bars still give us an idea of volume but without the easy analysis of visual weight.
The Blue Jackets scored a ton of goals from defensive zone counters and got a ton of those shots on net. It’s safe to say that something that the Blue Jackets did, or players that they have, were in some way exceptional on the counterattack. The Blue Jackets were also more efficient at winning dump-in races and exploiting the opposition than they were conceding them.
If the Blue Jackets were trading rimmed retrievals high in the zone (graded as exit turnovers) for risks of eating a puck or making a play to the defense partner, it doesn’t appear to have been a worthwhile tradeoff. The opposition scored as efficiently in either scenario.
The Blue Jackets were equal in terms of defensive zone breakouts to what they conceded suggesting, at least to a shallow level, that their structure and capacity to attack was more or less average.
Where the Blue Jackets struggled, it appears, was in scoring off of deep regroup offense. This type of offense comes from cleared pucks that pass into the Blue Jackets’ zone where the blue-line was not contested. This does also include set breakouts where defensemen stayed at the dots before moving ahead. The Blue Jackets got shots on net but they did not score goals.
They looked better at neutral zone re-entry offense which is offense on zone clears or forecheck turnovers that result in loose pucks in the neutral zone (rather than a forecheck exit which is kept in-zone) and where the offense attacks quickly after retrieving the puck. Personally, I’d assign a lot of credit to Zach Werenski and Adam Fantilli but I suppose we can always dive into the numbers.
Once again ranking by rates but displaying volume, we can take a look into the foremost creators of dz counter offense which looks to be the Blue Jackets’ most dominant dimension.
To be honest, the results are slightly surprising. Adam Fantilli, to no surprise, was the Blue Jackets most dangerous counterattack scorer scoring 8 goals from this context exclusively. Where I was surprised, however, was in both Kent Johnson and Cole Sillinger(!) leading the rate contributions. That James van Riemsdyk and Kirill Marchenko follow, while being different stylistic contributors, is also fascinating.
Not to be forgotten is Zach Werenski who, at times early in the season, was a solo counter attacking demon. Under Evason he has been allowed to follow a forward high in the zone where his terrifying stick work created numerous offensive chances a la Roman Josi.
Johnson, Marchenko and Werenski were all the best playmakers. Johnson’s resulted in goals the most often but he and Marchenko still have some room to improve their passing efficiency converting some of those misses into shots on net. Cole Sillinger, especially, appears either hard done by finishing luck or as a passer who finds the ideas but struggles with the ultimate execution.
In any case, that these young and/or offense forward players thrive in creating a volume of counter attack shots bodes well for their futures under an aggressive system.
I don’t want to spend too much time on defensive zone breakouts but I think there’s plenty of fruit left on the vine for the top Blue Jackets players. Sean Monahan and Yegor Chinakhov grade out unsurprisingly well. Their middle conduit and weakside speed combination was lethal early in the season and a highly instructive demonstration of the system concepts.
Adam Fantilli, still, scored the most off the rush despite having the least ice-time. It’s worth mentioning just how efficacious he was there despite the very real frustrations with his inefficiency. Similarly, you’d hope to see more from Kent Johnson. He did have 24 breakout assists on net but perhaps grading by danger would give a better idea of luck vs execution. Adam Fantilli on the other hand grades out as an atrocious passer with 2 scored, 8 on net and 20 missed.
From the d-men, Werenski stands out as usual. What’s interesting here is that Mateychuk and Severson grade out poorly in their final shot/pass involvement where Christiansen and Provorov are more involved.
The Blue Jackets were not a deep regroup team, they weren’t particularly good at scoring or creating volume which is perhaps evidenced by the very small amount of red.
That said, Kent Johnson looks like a good player, especially in terms of converting these regroups to shots on-net. He is, perhaps, uniquely prone to creating deep regroup offense as he consistently diffuses forechecks by holding the puck, allowing them to line change, and then beginning his attack. He didn’t create as many goals as you’d like perhaps but he was effective in getting puck on net.
The usually offensive creators show up here again with a good effort from Boone Jenner in his limited sample. Monahan and Sillinger fall away in terms of volume as does Yegor Chinakhov.
The Blue Jackets were indeed a good re-entry team and it looks like the credit there goes to Adam Fantilli, Zach Werenski, a surprise Mikael Pyyhtia and both Chinahkov and Marchenko. It’s not listed here but Zach Werenski was indeed the highest rate contributor suggesting his performance in re-entry offense was indeed critical to the Blue jackets.
Fantilli, different from all other situations, featured primarily as a playmaker. On one hand, I think that makes sense. He’s likely to be the player getting the puck in the middle and carrying it over the blue-line which means he has to set up the final shot rather than fully-carry and shoot himself. I think his skating-on-arcs and excellent puck handling are built for exactly this type of offense.
Pyyhtia and Chinakhov might stand out as surprises but considering all of the top contributors of this offense are excellent skaters, perhaps it shouldn’t be. I think that carries over to the defense as well where Harris, Mateychuk and Christiansen round out the next group as well. These contributions are difficult to fully deliniate given the importance of the first few passes in build-up, and in the forechecking required to create them in the first place, but neat information nonetheless.
In terms of forecheck creation, the data is extremely interesting. The players who are associated with creating the most forechecks/60 look like they’re ultimately the least successful. Perhaps, then, forechecking as an intentional act breaks less structure or perhaps the more opportunistic forecheckers are also the ones with the skill to convert quickly.
Dmitri Voronkov struggled to get involved in retrieval turnovers but when he did they often ended in the back of the net. He’s a special player with a special skillset but not without weaknesses at the same time. If he comes back with a little more intensity or stamina, he could get involved far more often.
The defensemen, in partuclar, look like their involvement is much more random. Only Provorov and Werenski converted the retrieval shots into goals.
The forecheck offense looks a little more aligned in terms of volume → goals. Some of the same offensive players make an appearence here as well. Fantilli and Marchenko scored a ton of goals off of forecheck exits, suggesting some clear quick-strike ability, where Kent Johnson showed excellent capacity to set up these situations. If he’s high in the zone, he can find the next play, making him an excellent complement to their skillsets. James van Riemsdyk was the next best scorer of the bunch making up for his low volume with great efficiency.
Sean Monahan and Sean Kuraly were the other two excellent exit takeaway playmakers. It’s easy to see an alignment with the center position here, perhaps their position as F3 enables them the first and best offensive touch on the puck where other players are more involved in creating the takeaways in the first place.
Wrap Up
I hope that this series of charts was informative and entertaining. I think, after reviewing everything in depth, I’ve come away with a renewed appreciation for certain players (Adam Fantilli’s offensive prowess, Denton Mateychuk’s offensive contributions) and deepened understanding of the relationships between certain aspects of Evason’s scheme. I didn’t expect tipping to be such a great tool but the conclusions based on other forms of in-zone offense was still interesting.
I think, in particular, breaking out certain types of offense and shots for the tracking project next year is put into better clarity. Breaking deep-regroups out into set-breakouts, adding dead-angle shots, accounting for rebound danger, even separating blocks and misses and perhaps finding a way to account for hard-rimmed retrieval turnovers all stand out in particular. I think adding the player responsible for beginning possession could help inform some of the most specific play-driving aspects.
The part I struggle with is that the bins that were previously formed feel like an easily understandable depth of detail and breaking the danger into more bins might fragment the data even more. I don’t have any strong opinions on how I’ll move forward next season, though breaking it out completely between xG and xA1 (should I decide that NHL PxP is satisfactory for pure location data) and perhaps pass-weight Chance Grades (‘A’, ‘B’, ‘C’, etc) while keeping the detail for more specific dives later in the season.
This isn’t the end of my work with this data specifically. This offseason, I hope to develop the skills to marry my hand-tracked data with NHL PxP thereby forming on-ice metrics and being able to graph both for- and against metrics, which may be instructive for defensemen and forward play-driving capacities. Ideally, I can develop some ability to construct my own viz, unfortunately sacrificing the interactivity that I think is critical to fun exploration of data, but that way I can combine some charts by bringing more contextual information and different colors/textures to the bars or scatterplots.
Most importantly, I think, will be some actual statistical analysis to determine the best deployment of any of these metrics or statistics. Does including blocked shots by danger help improve “chances'“? Is there any single combination of danger metrics or shot types or what have you that makes the most sense to optimize for? Which metrics, chance assist, high danger contributions, all shot assists predicts the future acutal offensive production most quickly? These sorts of questions are things I hope to be able to answer
As I learn, my plan is to update this space with whatever new metrics or information I develop the capacity to share, so stay tuned! In particular, I already have some single player “cards” or charts that express finishing by situation and danger (or even who set them up/finished passes they created in a given situation). I may debut that with an Adam Fantilli specific film/stats project coming up, but if not I may simply publish a single post of just graphs and charts. Otherwise, I hope to wrap up some “roster holes and targets” and may revisit my publishing schedule closer to draft/free-agency if I’m feeling the itch or find some free-time amidst moving.
As always, thanks for joining in the journey! I appreciate any who reads this deep into anyting I’m doing and hope I can keep moving forward and creating better data and better analysis!