• 77% YouTube Viewership Growth: Live streaming viewership on YouTube jumped by 77% season-over-season, fueled by fans tuning in from 154 countries worldwide (up from 106 the previous year).

• 1,116,497 Regular-Season Attendance: The league shattered records by surpassing 1 million single-season attendees for the first time, drawing over 1.1 million fans across 120 regular-season games.

• 9,304 Fans Per Game: Average attendance per game rose to 9,304—marking a 28% increase over the previous campaign (including playoffs) and a staggering 71% jump from the league’s inaugural season.

• 18,006 Peak U.S. Attendance Record: The league set a new U.S. record for professional women’s ice hockey when a sold-out crowd of 18,006 fans packed Madison Square Garden for a matchup between the New York Sirens and Seattle Torrent.

• 190% Post-Olympic Merchandise Surge: Following the Milano Cortina Winter Games, online merchandise sales experienced a 190% increase compared to the same period the prior year, alongside in-arena gear sales doubling overall.

It’s only fitting that as the viewership grows, so does the accessibility of data. I set out to build something that I hope does the league justice when it comes to publicly available analytics.

The Data

Surprisingly enough, the easiest task when building out the site was getting the data to do so. I’m fortunate to have access to one of the top proprietary firms when it comes to data from every league, but the issue there is they don’t actually have half of the metrics I’ve built. This is where logic and engineering come in. I had a huge file of over 1,000 events per game but how do I get the metrics that matter? I had to think realistically: in this play-by-play file, how do we identify a retrieval that was successful? Well, we have the breakout that occurred post-retrieval, the result of that entry, the location, and the player. From that, we can accurately say that when X occurs after Y action, that was a successful retrieval. One of the most exciting aspects of all this was combining source pages of data. I had about 10 different tabs that I had to go in and create custom extractions for somehow and then merge them all into one. One of the most time-consuming aspects of it but in the end 1000% worth it.

xG Model Architecture

I started with 7,249 shots from PWHL InStat play-by-play data, including 602 goals. For every shot, I looked at both where the shot came from and how the chance was created. Distance and angle to the net form the foundation, but the model also considers things like rebounds, rushes, cycle/forecheck possessions, tips, goalie movement, power plays, and other details surrounding the shot. The idea behind xG is actually pretty simple: if this exact type of chance happened 100 times, how many would we expect to become goals? A shot worth 0.25 xG is essentially saying we’d expect that type of chance to become a goal roughly 25% of the time.

I used logistic regression to turn all of those factors into a goal probability. The model was then tested using five separate validation samples, achieving an AUC of 0.944. In simple terms, the model does a very good job distinguishing dangerous chances from low-quality shots.

After accounting for location, shots following sustained cycle or forecheck pressure received one of the largest boosts in the model, while rebounds and uncontrolled rebounds were also extremely dangerous. One of the most interesting but logical findings to me was when a goalie is moving rather than set, the probability of a goal increases.

That’s ultimately what I wanted the model to capture. Not just where a shot happened, but what was happening around it. Two shots can come from almost exactly the same location while being completely different chances, and I wanted the model to recognize that.

Features

Once the underlying data was built, the next step was figuring out which metrics were actually worth presenting. I wanted the site to go beyond traditional box-score statistics and show how players and teams were creating results.

Some of my favourite metrics include:

Neutral Zone Risk

Neutral zone risk looks at how often a player puts their team in difficult situations when attempting to move the puck through the middle of the ice. Rather than simply counting turnovers, the goal was to add context to where those turnovers occurred and what happened afterward.

Entry Methods

Not all zone entries are created equally. This breaks down how players enter the offensive zone, whether that is through controlled carries, passes, dumps, or other methods. It gives a better picture of a player’s tendencies and how they prefer to attack the offensive zone.

Breakout Efficiency by Lane

Breakouts are another area where traditional statistics don’t tell the full story. This metric looks at how efficiently teams and players move the puck out of their defensive zone through different lanes, helping identify which options are actually producing successful exits.

These are just a few examples of the metrics available throughout the site. The larger goal was to build statistics that provide context rather than simply adding more numbers to a page.

Draft

With the draft occurring in the last few months, it felt fitting to include profiles on the draftees. All the profiles are the exact same as the individual players just with some different design work as well as needing to build my own API for all logos. It took some time but for the present and future it will certainly pay off since I was able to put it into one central API rather than finding random APIs for SDHL logos or other non-traditional leagues in the future. The biggest challenge was certainly teaching the system how to sift through NCAA data versus all the international tournaments the athletes play in. Overall, I’m happy with how it turned out. I plan to make some UI adjustments in the future but for now we have a strong baseline to build on.

PWHLe

The PWHL Equivalency projection tool works similarly to the traditional NHLe models you might be familiar with from the men’s game, but it is built specifically around the landscape of women’s hockey. Comparing a 40-point NCAA scorer to a 40-point player in Sweden’s SDHL isn’t as simple as looking at the box score. The leagues have different levels of competition, different scoring environments, and different levels of scoring compression, so the same number of points can mean very different things. The PWHLe tool attempts to account for that by calculating translation factors based on how players from different leagues have historically transitioned into the PWHL.

The differences are pretty significant. Historically, Sweden’s SDHL sits at the top of the model’s translation hierarchy, with a translation factor of .78, while Finland’s Liiga comes in at 0.74. NCAA D1 sits at 0.62, meaning a player’s college scoring is expected to decrease by roughly 38% when moving to the PWHL, while U SPORTS sits at 0.55. In other words, a point is not necessarily a point when you’re comparing players across leagues.

One of the more interesting parts of the model is that it doesn’t simply treat every player coming out of the same league identically. There is a reward for proving that you can move up. If a player goes from a weaker league into a stronger professional league, continues to produce, and then enters the PWHL, the model gives them a 4% to 15% boost to their most recent season’s translation. That matters because there is a big difference between dominating college hockey and proving that you can produce after already making the jump to a professional environment. The model is essentially saying: we’ve seen this player take a step up before, and they proved they belonged.

From there, the tool goes beyond simply projecting points. Players are also categorized into archetypes based on the balance of their offensive and defensive profiles. A forward with an elite scoring history can be classified as an Elite Finisher, while strong assist production and offensive scores can identify a Playmaker. Players with balanced offensive and defensive profiles fall into the Two-Way category, while players who don’t show significant production on either side are classified as Depth. Defensemen are treated differently, with the model distinguishing between Transition Defensemen, who show a strong offensive profile and the ability to drive play, and Shutdown Defensemen, whose defensive impact outweighs their offensive production.

The really interesting part is what happens next. The model doesn’t stop at predicting how many points a player might score. It uses that projected scoring rate to estimate what their game could actually look like on the ice. A forward’s expected points-per-game is used to estimate things like high-danger chances and zone entries. For example, a forward projected for 0.70 points per game would be expected to generate roughly 0.46 high-danger chances per game, while also being projected for around 4.2 zone entries per game. Higher-scoring forwards are expected to carry the puck more frequently rather than simply dumping it in and hoping for the best.

The model also recognizes that defensemen create offense differently. A high-end Transition D isn’t expected to produce in exactly the same way as a scoring forward. Instead, their offensive value is more closely tied to things like puck retrievals, transition play, and their ability to turn defensive possessions into offensive opportunities.

Ultimately, that’s what makes the PWHLe tool more interesting than simply taking a player’s points and multiplying them by a number. It tries to answer a much bigger question: if this player were playing in the PWHL right now, what would their game actually look like? It isn’t a crystal ball, and no model can perfectly account for things like defensive impact, physical development, role, or how a player adapts to a new level of competition. But it provides a common language for comparing players coming from completely different hockey environments—and gives us a much better starting point when trying to figure out who might be ready to make the jump to the PWHL.

Check it out here: https://project-u21dk.vercel.app

Leave a comment

Subscribe now

Microstats are the next step.

Corsi tells you what happened. Microstats try to explain how and why it happened.

1. Why did one line consistently create clean offensive-zone possessions while another spent the night chipping pucks off the glass?

2. Why does one defenseman suppress forecheck pressure while another gives up controlled entries despite identical xGA numbers?

3. Why does one player drive offense even when his linemates change?

The answer usually lives in the details:

• controlled entries and exits

• entry denial rate

• retrieval efficiency

• pass completion under pressure

• pre-shot movement

• slot pass generation

• puck recoveries

• transition involvement

Traditional analytics describe outcomes.
Microstats describe mechanisms.

That distinction matters.

Here are 10 principles for using microstats the right way.

1. Track Possession-Preserving Events, Not Just Touches

Not all touches are created equal.

A controlled zone entry is more valuable than a dump-in because possession actually survives the blue line.
A controlled exit is more valuable than a glass-and-out clear because it keeps the puck in play through the defensive zone.
A completed east–west pass through the slot is more valuable than a harmless perimeter touch because it forces the defense to move.

Modern tracking focuses on things like:

• controlled entry %

• controlled exit %

• retrieval-to-exit conversion rate

• offensive-zone possession retention

The game rewards teams that keep the puck.
Microstats measure how they keep it.

2. Treat Transition as the Core Engine of Hockey

The neutral zone isn’t empty space. It’s where modern hockey is often won.

Microstats show that transition ability drives territorial control:

• controlled entries create significantly more offense than dump-ins

• clean exits reduce time spent in the defensive zone

• entry denials suppress shot generation before it even starts

Elite transition players impact multiple layers of the game:

• retrieval

• support positioning

• puck transport

• entry creation

• offensive setup

This is why elite transition defensemen often look “quiet” to casual viewers. Their best skill is preventing chaos before it begins.

3. Use Rates and Context, Not Raw Totals

A player with 18 controlled entries isn’t automatically better than a player with 10.

Volume without efficiency doesn’t tell you much. Microstats become useful when you pair them with:

• success rates

• pressure context

• teammate quality

• usage patterns

• downstream impact

A defenseman retrieving pucks against an aggressive forecheck might post lower exit percentages than one playing behind a passive neutral-zone trap. That doesn’t mean he’s worse; it may just mean his environment is harder.

Raw counts encourage lazy analysis.
Rates and context lead to useful analysis.

4. Measure Defense Through Denial, Not Violence

Traditional hockey culture often measures defense with hits and blocks.
Microstats measure defense through denial.

Elite defenders:

• kill entries early

• angle attackers wide

• disrupt passing lanes

• recover dump-ins efficiently

• prevent controlled possessions from developing

Entry denial rate is often more valuable than hit totals.
A clean breakup at the blue line is worth more than a big hit after the opponent already has possession.

The best defenders reduce offensive Creation before it starts. Microstats finally let us quantify that.

5. Recognize That Passing Drives Offense

Shots finish plays.
Passes create them.

Modern microstat tracking is increasingly focused on:

• slot pass generation

• royal-road passes

• pre-shot movement

• pass completion under pressure

• offensive-zone connection sequences

Defenses break when they have to move laterally. A low-danger shot after three east–west passes is often more dangerous than a stationary shot from the slot.

Microstats help identify players who manipulate defensive structure—not just the ones who shoot.

6. Separate Retrieval Skill from Exit Skill

Retrievals and exits are not the same thing.

Some defensemen retrieve pucks well but panic under pressure.
Others struggle to win the puck but excel once they have it.

Elite transition players do both:

• recover the puck cleanly

• evade forecheck pressure

• complete possession-preserving exits

Retrieval microstats often reveal hidden defensive value. The player who consistently wins races to loose pucks and exits cleanly is quietly controlling the pace of the game—even if nobody notices it live.

7. Respect Sample Size and Repeatability

Some microstats stabilize quickly. Others take time.

Controlled exits and entries tend to become meaningful faster than goals or points because they happen more often.
Finishing talent is volatile.
Transition involvement is less volatile.

That’s why microstats are so useful in projection models. A player who consistently drives repeatable transition processes usually produces sustainable results over time. If the underlying mechanisms stay strong, the outcomes often follow eventually.

8. Understand That Systems Shape Metrics

Microstats are not system-neutral.

• An aggressive forecheck creates a different retrieval environment than a passive collapse.

• A rush-oriented offense creates different entry opportunities than a cycle-heavy structure.

• A defenseman told to prioritize glass-and-out clears may post terrible controlled-exit metrics even though he’s doing exactly what his coach wants.

Sometimes the numbers are measuring the system more than the player.
The best analysts understand tactics before they draw conclusions from data.

9. Treat Microstats as Decision Tracking

At their core, microstats are decision-tracking tools.

Every event reflects a choice:

• carry or dump

• pinch or retreat

• force the middle or go wide

• pass immediately or delay

• support low or stretch high

The best players consistently make decisions that improve possession value. That’s why elite play drivers tend to dominate microstats across multiple categories at once: retrieval efficiency, controlled exits, controlled entries, dangerous pass creation, offensive-zone recoveries, and transition involvement.

Microstats are behavioral analytics for hockey. They quantify decision-making under pressure.

10. Remember: Metrics Are Models, Not Reality

Every microstat is an abstraction.

• Tracking definitions vary.

• Pressure classifications vary.

• Event tagging varies.

• Data quality varies.

Two analysts can watch the same sequence and disagree on:

• whether possession was maintained

• whether pressure existed

• whether an exit was controlled

• whether a pass created a real advantage

Microstats are not objective truth. They’re structured attempts to quantify an extraordinarily fluid sport.

The goal isn’t perfect measurement.
The goal is better understanding.

Microstats aren’t meant to replace hockey knowledge.
They’re meant to deepen it.

Microstats are still in their infancy.
Computer vision is coming. Automated tracking is improving. NHL organizations are building increasingly sophisticated possession-value models behind closed doors.

Ten years from now, public hockey analytics may look completely different.

But one principle will probably survive:

The analysts who understand the micro-events that drive possession will typically outperform the analysts who only count the final result.

Leave a comment

Subscribe now

Expected goals, shot quality, and context-adjusted metrics now drive decision-making in front offices. But analytics are only powerful if you use them correctly. Misapply them, and you’re just the modern version of someone quoting plus-minus like it means anything.

These are the 10 commandments of doing it right.

1. Thou shalt not use save percentage to evaluate a goaltender’s shot-stopping ability.

Save percentage (SV%) assumes all shots are equal. They’re not. A backdoor one-timer and a harmless point wrister carry vastly different probabilities, yet SV% treats them the same.

A goalie facing a barrage of high-danger chances will almost always look worse than one seeing clean, perimeter shots.

Use expected goals against (xGA) and Goals Saved Above Expected (GSAx). These account for shot quality and tell you how many goals a goalie should have allowed. A .910 behind a defensive tire fire can be elite. A .925 behind a shot-suppressing machine might not be.

2. Thou shalt not use plus-minus to evaluate defense.

Plus-minus is a relic. It assigns credit or blame for goals without accounting for context—linemates, competition, or goaltending variance.

A shutdown defender on a bad team can be buried. A passenger on a stacked line can look elite.

Instead, look at xGA/60, on-ice suppression metrics, and transition data like zone denial and exit efficiency. Defense is about killing plays before they become shots—not just being on the ice when the puck goes in or out.

3. Thou shalt not judge scoring ability by small-sample production.

Goals and points are noisy, especially in small samples. Shooting percentage swings wildly over short stretches.

Twelve goals on 80 shots isn’t sustainable. Eight on 120 might be unlucky.

Focus on expected goals (xG) and individual xG rates (iXG/60). Players generating chances consistently tend to get rewarded over time. Finishing talent exists—but outside of the elite tier, regression usually wins.

4. Thou shalt not treat Corsi or Fenwick as the final word on possession.

Corsi and Fenwick measure volume, not quality. A team can dominate shot attempts while generating nothing dangerous.

Expected goals (xGF%) fixes this by weighting shots based on danger. The best teams don’t just outshoot—they outchance.

Shot quantity was a stepping stone. Shot quality is the destination.

5. Thou shalt not use hits, blocks, or penalty minutes as proxies for value.

“Grit” stats are seductive—and misleading.

Hits often mean you don’t have the puck. Blocks often mean you’re stuck defending. Penalty minutes can actively hurt your team.

The best defenders are quiet. They close gaps early, break plays up in transition, and limit dangerous chances before they happen.

If you want real value, look at chance suppression, puck recovery in dangerous areas, and clean exits.

6. Thou shalt not overvalue faceoff win percentage.

Faceoffs matter—but far less than people think.

Win percentage ignores context: zone starts, opponent strength, and what happens immediately after the draw. Plenty of “lost” faceoffs are recovered seconds later.

A center’s real value comes from what happens after the puck drops—transition play, forechecking, and defensive coverage.

Faceoffs are a piece of the puzzle, not the picture.

7. Thou shalt not judge teams by the standings alone.

The standings lie—at least in the short term.

One-goal games, shootouts, and hot goaltending can inflate records. Conversely, strong teams can underperform due to poor finishing or bad luck.

Expected goal share (xGF%) and underlying differentials tell you who’s actually driving play. Over time, results tend to follow process.

If a team looks “lucky,” they probably are. And it won’t last forever.

8. Thou shalt compare players using rate stats, not raw totals.

Raw totals ignore opportunity.

Ice time, role, and deployment drastically affect counting stats. Comparing a fourth-liner to a top-line player without adjusting for minutes is meaningless.

Use per-60 metrics—especially at 5-on-5. They level the playing field and reveal who’s actually driving results versus who’s just benefiting from volume.

9. Thou shalt fear small-sample shooting percentage and PDO.

Few stats are noisier than shooting percentage and PDO (shooting + save percentage).

Hot streaks create illusions. Cold streaks create panic. Most of it is variance.

A 25% shooting heater over 15 games doesn’t make someone a sniper. A PDO spike doesn’t mean a team “figured it out.”

Regression isn’t a theory—it’s a guarantee.

10. Thou shalt always account for context.

No stat exists in a vacuum.

Deployment, quality of competition (QoC), and quality of teammates (QoT) shape every number. A player’s results are heavily influenced by who they play with, who they face, and where they start.

Modern models—like RAPM or GAR—attempt to isolate individual impact by controlling for these variables.

The best analysts ask three questions before anything else:

Who did he play with?

Who did he play against?

And where did those minutes happen?

Leave a comment

The NHL Entry Draft is the primary mechanism for teams to acquire young talent, typically aged 18–20, from amateur and professional leagues worldwide. Understanding how drafted players progress to the NHL—considering height, weight, position, draft-year league, and draft position—can reveal optimal development pathways. This research project analyzes NHL draft classes from 2015 to 2024 (~2,240 players) to identify trends and determine the most effective routes to the NHL, defined as playing 100+ NHL games (with 200+ games as a secondary benchmark for established careers). By incorporating draft position, we account for varying expectations and development timelines, particularly the differences between high (1st round), mid (2nd–3rd round), and late (4th–7th round) picks.

Methodology

• Data Scope: NHL Entry Drafts from 2015 to 2024, covering approximately 2,240 players (7 rounds, ~32 teams, ~224 picks per year).

• Variables Analyzed:

  • Height and Weight: Sourced from NHL Central Scouting, Elite Prospects, and NHL.com.

  • Position: Forwards (centers and wingers), defensemen, goaltenders.

  • Draft Year Leagues: CHL (OHL, WHL, QMJHL), USHL, NCAA, European leagues (SHL, Liiga, KHL, etc.), and others (e.g., BCHL).

  • Draft Position: Categorized as 1st round (picks 1–32), 2nd–3rd rounds (picks 33–96), and 4th–7th rounds (picks 97–224).

  • Pathways: Post-draft leagues (AHL, CHL, NCAA, Europe) and time to NHL debut (measured in seasons post-draft).

• Success Metrics:

  • Primary: 100+ NHL games played, indicating a meaningful NHL career.

  • Secondary: 200+ games for established players.

  • Additional: Points per game (PPG) for forwards, plus/minus for defensemen, and save percentage (SV%) for goaltenders where applicable.

• Sources: Elite Prospects, NHL.com, Hockey-Reference.com, DobberProspects, Neutral Zone, The Hockey Writers, and analytical studies from Sportsnet and ESPN. Web searches and posts on X were used for real-time prospect data, particularly for 2023–2024 drafts.

NHL Draft Overview

The NHL Entry Draft occurs annually in late June, with 32 teams selecting players over seven rounds. Eligible players are 18 by September 15 of the draft year, up to 20 (North Americans) or 21 (non-North Americans). The draft order is set by a weighted lottery for non-playoff teams and regular-season standings, with trades impacting pick distribution. Drafted players typically sign three-year entry-level contracts (ELCs) and develop in junior leagues, the AHL, NCAA, or European leagues before reaching the NHL. First-round picks face higher expectations and often receive more development resources, while late-round picks must overcome greater odds to reach the NHL.

Detailed Analysis of Player Attributes and Pathways

1. Height and Weight Trends

• Height:

  • Forwards: Average height is 6’0.54” (184.24 cm). Smaller forwards (<5’11”) have become more common, with 15 players 5’10” or shorter selected in the top 16 picks from 2020–2024 (e.g., Connor Bedard, 5’10”, 2023 #1 overall; Cole Caufield, 5’7”, 2019 #15 overall), compared to only 4 from 2010–2014. This reflects a shift toward valuing speed and skill.

  • Defensemen: Average height is 6’1.54” (186.8 cm). Taller defensemen (6’3”+) like Victor Hedman (6’6”, 2010) remain preferred for shutdown roles, but smaller, mobile defensemen like Lane Hutson (5’10”, 2022 #62 overall) are increasingly drafted. From 2015–2019, 32 defensemen 6’3”+ were drafted in the 1st round, compared to 8 under 6’0”.

  • Goaltenders: Average height is 6’2.69” (189.7 cm). Shorter goaltenders like Juuse Saros (5’11”, 2013 #99 overall) have succeeded, with 7 of the top 15 goaltenders in SV% in 2023–24 being 6’2” or shorter (e.g., Saros, .918 SV%). Taller goaltenders (e.g., Connor Hellebuyck, 6’4”, 2012 #130 overall) remain valued for net coverage.

  • Trend: Height is less restrictive than a decade ago, with teams prioritizing skill for forwards and mobility for defensemen. Goaltender success depends more on technique than size.

• Weight:

  • Forwards: Average weight is ~185 lbs. Lightweight forwards (<175 lbs) like Matthew Phillips (140 lbs, 2016 #166 overall) are rare but can succeed (Phillips: 93 NHL games). Heavier forwards (200+ lbs) like Tage Thompson (6’6”, 215 lbs, 2016 #26 overall) are valued for physicality.

  • Defensemen: Average weight is 190–200 lbs. Heavier defensemen (210+ lbs) like Jamie Oleksiak (6’7”, 257 lbs, 2011 #14 overall) excel in physical roles, but lighter defensemen (180–190 lbs) like Cale Makar (5’11”, 187 lbs, 2017 #4 overall) dominate with skill.

  • Goaltenders: Weight is less critical, ranging from 180 lbs (Saros) to 220 lbs (Hellebuyck). Agility and positioning outweigh mass.

  • Trend: Recent drafts (2022–2024) show a slight increase in average weight for 1st-round forwards (190.6 lbs in 2022, 192.6 lbs in 2024), suggesting a preference for physicality among top picks, though skill remains paramount.

2. Position Analysis

• Forwards:

  • Centers: Slightly taller (6’1”) and more likely to be high picks due to playmaking demands (e.g., Jack Hughes, 5’11”, 2019 #1 overall, 0.94 PPG). Centers often require longer development for two-way play.

  • Wingers: Average 6’0”, valued for scoring (e.g., Matvei Michkov, 5’10”, 2023 #7 overall, projected 0.80 PPG). Wingers may transition faster due to fewer defensive responsibilities.

• Defensemen:

  • Smaller defensemen (<6’0”) like Matthew Schaefer (6’2”, 2025 prospect) focus on puck-moving and skating, often needing extra development time. Larger defensemen (6’3”+) like Noah Hanifin (6’3”, 2015 #5 overall) are drafted for physicality and defensive reliability.

  • From 2015–2024, 45% of 1st-round defensemen were 6’3”+, compared to 20% of 4th–7th-round defensemen.

• Goaltenders:

  • Goaltenders are often drafted in later rounds (only 3 goaltenders in 1st round from 2015–2024). Success depends on technique and mental resilience, with players like Saros excelling despite later selection.

3. Draft Year Leagues

• CHL (OHL, WHL, QMJHL):

  • Produces ~40% of drafted players annually. OHL leads with 27% of Central Scouting’s North American skaters (2016–2017), followed by WHL (25%) and QMJHL (15%). Conversion rates: OHL (56–64% of listed players drafted), WHL (50–60%), QMJHL (38–41%).

  • Examples: Connor McDavid (OHL, 2015 #1 overall), Alexis Lafrenière (QMJHL, 2020 #1 overall).

• USHL:

  • Represents 10–15% of Central Scouting’s pool but has a high conversion rate (53–58% drafted). Players like Cole Caufield (USHL, 2019) benefit from competitive play and NCAA transitions.

• NCAA:

  • Older players (19–20) are drafted at a lower rate (14–31% of non-listed players). NCAA players like Adam Fantilli (2023 #3 overall) often transition quickly due to physical maturity.

• European Leagues:

  • SHL (Sweden), Liiga (Finland), and KHL (Russia) produce NHL-ready players due to professional competition. Examples: Juraj Slafkovsky (Liiga, 2022 #1 overall), Victor Frondell (Allsvenskan, 2025 prospect).

  • Russian players faced restrictions post-2022 due to geopolitical factors, reducing KHL draft share.

• Other Leagues:

  • BCHL and AJHL produce fewer draftees but yield standouts like Bradly Nadeau (BCHL, 2023 #30 overall).

4. Draft Position Impact

• 1st Round (Picks 1–32):

  • High expectations, with 82% playing 50+ NHL games and 65% reaching 200+ games. Top-10 picks often debut in draft+1 or +2 (e.g., McDavid, 2015, NHL debut in 2015–16).

  • Development: 30% go directly to NHL, 50% spend 1–2 years in CHL/Europe, 20% in NCAA/AHL.

• 2nd–3rd Rounds (Picks 33–96):

  • 40% reach 100+ NHL games, 25% reach 200+ games. Players like Drake Batherson (2017 #121 overall, QMJHL) often spend 2–3 years in CHL/AHL.

  • Development: 60% return to CHL/NCAA, 30% to AHL, 10% to Europe.

• 4th–7th Rounds (Picks 97–224):

  • Only 10–15% reach 100+ NHL games, 5–8% reach 200+ games. Examples: Yanni Gourde (2012 undrafted, QMJHL, 500+ games), Juuse Saros (2013 #99 overall, 400+ games).

  • Development: 70% spend 2–4 years in CHL/NCAA/Europe, 25% in AHL, 5% direct to NHL.

5. Pathways to the NHL

• Immediate NHL Transition:

  • Primarily 1st-round picks, especially top-5 (e.g., Bedard, 2023 #1 overall, NHL in 2023–24). Only 2–3 players per draft play 50+ NHL games in draft+1.

• AHL Development:

  • Common for 1st–3rd-round picks, with 60% of 2015–2019 draftees spending 1–2 seasons in AHL (average 80 AHL games before NHL debut). Example: Cale Makar (2017 #4 overall, NCAA to AHL to NHL).

• CHL (Junior Leagues):

  • Standard for CHL-drafted players (18–19 years old), with 80% of 1st–3rd-round picks returning for 1–2 years. Late-round picks may stay 2–3 years to dominate (e.g., Brayden Point, 2014 #79 overall, WHL, 0.92 PPG in NHL).

• NCAA:

  • Older draftees (19–20) play 1–3 college seasons, transitioning to AHL or NHL. Example: Matthew Wood (2023 #15 overall, BCHL to NCAA, projected NHL debut 2026–27).

• European Leagues:

  • Players like Slafkovsky (Liiga) or Frondell (Allsvenskan) stay 1–2 years in professional leagues, then move to AHL/NHL. European draftees average 1.5 years longer to debut than CHL players.

• Time to NHL:

  • 1st Round: 75% debut by draft+2, 90% by draft+3.

  • 2nd–3rd Rounds: 50% debut by draft+3, 70% by draft+4.

  • 4th–7th Rounds: 30% debut by draft+4, 50% by draft+5.

Optimal Pathways by Player Profile and Draft Position

Using data from 2015–2024, optimal pathways are tailored to height, weight, position, draft-year league, and draft position:

• Forwards:

  • Smaller Forwards (<5’11”, <175 lbs):

    • 1st Round: Drafted from CHL/USHL, return to junior for 1–2 years to dominate (e.g., 1.5+ PPG), then 1 season in AHL. Example: Cole Caufield (2019 #15, USHL/NCAA, 1 year NCAA, 0.5 AHL season, NHL debut 2021).

    • 2nd–3rd Rounds: 2–3 years in CHL/NCAA, 1–2 years AHL. Example: Alex DeBrincat (2016 #39, OHL, 2 years OHL, 1 AHL, NHL debut 2017, 0.82 PPG).

    • 4th–7th Rounds: 3–4 years in CHL/NCAA to maximize scoring, 1–2 years AHL. Example: Yanni Gourde (undrafted, QMJHL, 3 years QMJHL, 2 years AHL, NHL debut 2015).

    • Rationale: Smaller forwards need extended skill development to compete against NHL physicality.

  • Larger Forwards (6’1”+, 190+ lbs):

    • 1st Round: Drafted from CHL/Europe, 1 year in CHL/Europe or direct NHL. Example: Juraj Slafkovsky (2022 #1, Liiga, NHL debut 2022–23).

    • 2nd–3rd Rounds: 1–2 years CHL/Europe, 1 year AHL. Example: Tage Thompson (2016 #26, NCAA, 2 years NCAA, 1 AHL, NHL debut 2018).

    • 4th–7th Rounds: 2–3 years CHL/NCAA/Europe, 1–2 years AHL. Example: Ryan O’Reilly (2009 #33, OHL, 1 year OHL, NHL debut 2009).

    • Rationale: Physical maturity enables quicker transitions, especially for high picks.

• Defensemen:

  • Smaller Defensemen (<6’0”):

    • 1st Round: Drafted from USHL/NCAA/CHL, 1–2 years NCAA/CHL, 1–2 years AHL. Example: Lane Hutson (2022 #62, NCAA, projected NHL debut 2025).

    • 2nd–3rd Rounds: 2–3 years NCAA/CHL, 1–2 years AHL. Example: Samuel Girard (2016 #47, QMJHL, 2 years QMJHL, NHL debut 2017).

    • 4th–7th Rounds: 3–4 years NCAA/CHL, 2–3 years AHL. Example: Andrew Nielsen (2015 #65, WHL, 3 years AHL, limited NHL).

    • Rationale: Smaller defensemen need time to develop puck-moving and defensive skills.

  • Larger Defensemen (6’3”+, 200+ lbs):

    • 1st Round: Drafted from CHL/Europe, 1–2 years CHL/Europe, 0–1 year AHL. Example: Noah Hanifin (2015 #5, NCAA, 1 year NCAA, NHL debut 2015).

    • 2nd–3rd Rounds: 1–2 years CHL/Europe, 1–2 years AHL. Example: Brandon Carlo (2015 #37, WHL, 1 year AHL, NHL debut 2016).

    • 4th–7th Rounds: 2–3 years CHL/NCAA, 2–3 years AHL. Example: Connor Carrick (2012 #137, USHL, 2 years NCAA, 2 years AHL, NHL debut 2013).

    • Rationale: Physicality accelerates transitions, but defensive awareness requires seasoning.

• Goaltenders:

  • All Sizes (height less predictive):

    • 1st Round (Rare): Drafted from CHL/Europe, 1–2 years CHL/Europe, 1–2 years AHL. Example: Jake Oettinger (2017 #26, NCAA, 2 years NCAA, 1 AHL, NHL debut 2020, .914 SV%).

    • 2nd–3rd Rounds: 2–3 years CHL/Europe/NCAA, 2–3 years AHL. Example: Juuse Saros (2013 #99, Liiga, 2 years Liiga, 2 years AHL, NHL debut 2015).

    • 4th–7th Rounds: 3–4 years CHL/NCAA/Europe, 2–4 years AHL. Example: Adin Hill (2015 #76, WHL, 2 years WHL, 3 years AHL, NHL debut 2017).

    • Rationale: Goaltenders require the longest development (4–6 years), with late-round picks needing extended AHL time to refine technique.

Key Findings

1. Height and Weight:

   • Smaller forwards (<5’11”) are increasingly successful (15 top-16 picks since 2020), but larger forwards (6’1”+, 190+ lbs) dominate 1st-round selections (60% of 2022–2024 1st-round forwards).

   • Defensemen favor height (6’3”+ for 45% of 1st-round picks), but smaller, mobile defensemen are rising.

   • Goaltender height is less predictive; technique drives success (e.g., Saros, 5’11”, .918 SV%).

2. Position:

   • Centers require longer development for two-way play, while wingers transition faster due to scoring focus.

   • Defensemen, especially smaller ones, need 2–4 years to develop defensively.

   • Goaltenders face the longest timelines, often debuting at 22–24.

3. Draft Year Leagues:

   • CHL produces the most draftees (40%), with OHL leading (27% of skaters). USHL (10–15%) and European leagues (15–20%) are growing in influence.

   • NCAA players, being older, have shorter post-draft timelines but lower draft rates.

4. Draft Position:

   • 1st-round picks: 75% debut by draft+2, with 30% going directly to NHL.

   • 2nd–3rd rounds: 50% debut by draft+3, requiring 1–2 years AHL.

   • 4th–7th rounds: 30% debut by draft+4, needing 2–4 years in junior/NCAA and 1–3 years AHL.

5. Optimal Pathways:

   • High Picks (1st Round): CHL/Europe to NHL (1–2 years) or AHL (1 year).

   • Mid-Rounds (2nd–3rd): CHL/NCAA/Europe (2–3 years), AHL (1–2 years).

   • Late Rounds (4th–7th): Extended CHL/NCAA/Europe (3–4 years), AHL (2–3 years).

   • Smaller players benefit from junior/NCAA dominance; larger players leverage physicality for quicker transitions.

Limitations

• Incomplete Data: Players from 2022–2024 drafts are still developing, limiting long-term success metrics.

• Individual Variability: Injuries, team depth, and coaching decisions impact pathways.

• Geopolitical Factors: Restrictions on Russian players post-2022 reduced KHL draft share.

• Statistical Noise: Draft-year stats (e.g., PPG) have weak correlations with NHL success due to development changes.

Conclusion

Optimal NHL pathways depend on a player’s physical attributes, position, draft-year league, draft position, and current landscape of drafted team.

• Smaller Forwards: Thrive in CHL/USHL, needing 2–4 years of junior/NCAA dominance (1.5+ PPG), then 1–2 years AHL. Late-round picks require longer junior time.

• Larger Forwards: Transition faster from CHL/Europe (1–2 years) to AHL/NHL, especially 1st-round picks.

• Defensemen: Smaller players need 2–4 years in NCAA/CHL for skill development, while larger players move quicker (1–2 years CHL/Europe, 0–1 year AHL).

• Goaltenders: Require 4–6 years across CHL/NCAA/Europe and AHL, with late-round picks needing extended AHL seasoning.

• Draft Position: 1st-round picks debut fastest (draft+1 or +2), mid-rounds need 2–3 years, and late-rounds need 4–6 years, often via AHL.
  The CHL remains the dominant talent pipeline, but USHL and European leagues are rising. Teams increasingly value skill over size for forwards, though physicality matters for defensemen and recent draft trends. Tracking 2022–2024 draftees will refine these insights as they reach the NHL.

  Leave a comment

Out of curiosity, I decided to run a regression model to identify which microstats correlate most strongly with points in the NHL. Here’s what I found, how I built the model, and what it means for teams looking to boost their scoring.

My Process

Using R and AI, I built a regression model to predict points based on a range of microstats, leveraging play-by-play data (similar to what’s available from providers like Instat). I narrowed the model to 75 predictors, down from 91 in a previous version, to simplify the analysis and reduce overfitting. Packages like dplyr handled data wrangling, scales helped with formatting, and car checked for multicollinearity using Variance Inflation Factors (VIF). The goal was to create a robust, interpretable model that still captured the complexity of NHL scoring.

The model performed strongly, with an R-squared of 0.9918, meaning it explained 99.18% of the variance in points. However, the residual standard error (0.2179) was slightly higher than before, suggesting predictions were a bit less precise due to the trimmed set of predictors. I also found that four variables, controlled entries, scoring chances, shots, and zone exits per 60 minutes caused singularities, likely because they were too closely related to other variables.

Key Findings

So which microstats most strongly correlate with points?

• Passes (Estimate: 1.656, p < 0.001): Every additional pass strongly boosts points. This isn’t surprising since puck movement usually creates scoring opportunities.

• Shots off Rush (Estimate: 0.696, p < 0.001): Shots from rush plays have a big positive impact, reflecting the high-danger nature of fast breaks.

• Exits (Estimate: 4.295, p < 0.001): Successful zone exits are a massive driver of points, likely because they kickstart offensive transitions.

• Shots off Forecheck or Cycle per 60 (Estimate: 0.161, p < 0.001): Sustained pressure through forechecking or cycling leads to more points.

• Rush Assists per 60 (Estimate: 0.180, p < 0.001): Assists on rush plays are a strong predictor, aligning with the importance of speed in today’s NHL.

Some results were also very interesting to see and interpret. For instance, Botched Retrievals (Estimate: 0.232, p < 0.001) and Failed Exits (Estimate: 0.238, p < 0.001) positively correlate with points, which seems accurate since having excellent on-puck decision-making will result in increased offence. On the flip side, Rushed Exits (Estimate: -0.172, p < 0.001) and Build-up Passes per 60 (Estimate: -0.138, p < 0.001) hurt points.

What Coaches Can Do

These findings are actionable. To maximize points, teams should:

• Prioritize puck movement: Encourage more passes to create scoring opportunities.

• Focus on rush plays: Practice quick transitions and shots off the rush, as they’re high-value actions.

• Improve zone exits: Drill clean exits to spark offensive plays without losing possession.

• Sustain pressure: Emphasize forechecking and cycle plays to generate shots in high-danger areas.

Final Thoughts

Microstats are a goldmine for understanding what drives scoring in the NHL. By focusing on high-impact actions like passes, rush shots, and zone exits, teams can fine-tune their strategies to generate offense. Of course, no single theory can explain everything. Hockey is simply too dynamic. But organizations with strong development teams capable of teaching and reinforcing these micro-details will set themselves apart in a league where every edge matters.

Subscribe now

Leave a comment

Development

Undoubtedly the most important tool one of these organizations can have is a capable player development department. For non-destination or small market teams, you must create through the draft. For this type of market, you have a higher chance of retaining a player you draft and develop rather than someone you acquire on a short-term deal in free agency or at the trade deadline. Just look at the Winnipeg Jets. They’ve drafted, developed, and retained players such as Mark Scheifele, Nikolaj Ehlers, and one of the highest-valued goaltenders in Connor Hellebuyck.

Analytics

For organizations operating with smaller budgets and fewer options, analytics is another critical tool. These teams often have a lower risk tolerance compared to higher-valued teams with broader fan bases to sustain them. The concept and utilization of analytics is where staff versatility can come into play. We want to be able to combine the development and analytics approach into one. With the limited opportunities in a smaller market, we can relatively create these positions cost-consciously and build a team of say 2-4 people.

1. Player Acquisition: Being required to spend wisely necessitates having trade value cost acquisition models in addition to a short- and long-term view. (For example, if we know a player may not be here long-term, should we still sign him with the intent of flipping him for picks?)

2. Scouting: A lower-valued team can’t be wasting picks so using a hybrid model, mixing the scout’s eye test and analytics is the optimal strategy.

3. On Ice: Game preparation focuses on how your team can expose the opponent as well as possible areas the opposition may attack

4. Player Focus: Building programs that are essential for maximizing individual performance

   ---

If we can push the focus on these areas, we can not only identify players and prospects wisely but we can compete in the present while compiling draft picks as our non-committed players choose to depart. If we know a player is leaving, how can we use these tools to maximize his play so that A) we can get all value out of him before he leaves and B) we can then maximize the return whether that be at the deadline or for signing rights?

Conclusion

It’s certainly an easier concept to say than to do. All successful companies maximize value out of the resources available to them, so this isn’t a foreign concept. The key lies in finding the right people for the job—individuals with knowledge, experience, and a growth mindset.

For small-market teams, even building specific models can be an area where consulting or hiring interns is implemented rather than having a full-time position. If resources allow, a dedicated team member is ideal, but if not, bringing in consultants to work on projects such as draft models as needed can be a cost-effective solution. By investing in the right people, fostering collaboration, and strategically combining analytics and development departments, small-market teams can level the playing field. Success doesn’t always come from having the biggest budget but from maximizing the resources available to you.

Subscribe now

Leave a comment

Research

Hockey is extremely unique compared to other popular sports like baseball, football, and basketball. The biggest difference is that players are on skates rather than two feet touching the ground. Michael Bracko published a great paper diving into the skating details of NHLers. Having these benchmark values allows us to collect video to identify the player’s current metrics and then come up with a plan to help them reach their target of being a fast skater.

Software

Arguably one of the best skaters in the NHL, Avalanche forward Nathan MacKinnon is a prime example of a wide ice-carving stride. This is backed up by NHL Edge, he sits in the 99th percentile of speed bursts over 20 mph, 95th of top skating speed, and 99th in skating distance. Below you can see multiple points on his body. These were created using:

import cv2
import mediapipe as mp
import pandas as pd
import numpy as np

Along with the annotated video, the code produces a file that shows the x and y values of each of the points on the body that you see below. In the file, we can also see custom stats so in mine I currently see speed, trunk angle, knee flexion angle, time to take off, and abduction angle.

We can use this tool for not only skating but shooting as well. Who better to analyze than Auston Matthews? Shooting the puck is a lot more like hitting in baseball. In Daryl Belfry’s book Belfry Hockey, he spoke about how they worked on improving Austons shot by taking components of Daniel Sprongs shot and working that into Auston’s game. At the moment I’m working on adding in the stick and puck detection but you can imagine how having the exact angles and details of Sprong's shot could be used to put that into Matthews’ game.

Moving Forward

I can imagine that NHL clubs have this type of technology available to them with the brains in teams today. If you read my work you know that I’m a huge proponent of player development in teams. Having a forward-thinking and up-to-date department not only helps from developing your own players but even flipping players for assets depending on the contention level of a team. If you’re intrigued by this type of technology feel free to reach out with some of your own video. I plan to spend this winter testing this out on my own game on the outdoor rinks to see any changes critiques I can make.

Leave a comment

Subscribe now

Process

Luckily for my own sanity, I did not have to collect all of this data manually. I was able to utilize the Top Down Hockey scraper courtesy of Patrick Bacon. With this tool, I downloaded data starting from the 2020-2024 USports season as well as specific leagues. In the coming weeks I will be adding more leagues but to build a starting point I used the OHL, OJHL, GOJHL, BCHL, and MJHL. Once I had this data I identified the names that occurred in both one of these leagues and in USports. Once I had the targeted players I was able to put that into my code and identify the equivalencies for each league. Keep in mind that I did equivalencies separately for defensemen and forwards just to ensure accuracy.

Projected Points = PPG (Junior) × Equivalency × Games Played (USports)

Results

So what leagues translated most to USports men’s hockey? It’s quite easy to guess that yes it is the OHL. There’s a reason it is the most NHL-drafted league. Coming in second is the BCHL following with the MJHL, OJHL, and GOJHL.

As with all research, this model has its limitations. For example, something I did in the past was conduct a correlation matrix to identify which statistics led to points in USports. I found that shots on goal, passes to the slot, scoring chances, and entries had a positive correlation with points, suggesting that a more advanced model incorporating these metrics could provide an even more accurate projection.

Conclusion

In a league where more often than not teams don’t have people working with data, the edge this work provides is significant. Beyond recruitment, this model has practical applications for coaching strategies, player development, and talent evaluation. For example, if we know that successful zone entries lead to a higher probability of points, we can tailor practices to improve this skill, giving players specific areas to work on that directly impact their performance. By refining and expanding this research to include additional leagues and performance metrics, we can help teams make more informed decisions, while also helping players reach their full potential and providing them with the necessary tools to transition beyond USports.

Leave a comment

Subscribe now

What are Microstats?

So it’s likely that if you’re reading this you’re aware of what expected goals are but what leads up to something being a so-called expected goal? How does the puck get from the defensive zone to the offensive zone, does the team enter with control or is it via a dump-in, is the pass leading to the goal a low-to-high pass or a behind-the-net pass? Microstats dive even deeper than traditional analytics and in my opinion, should be the most targeted category of information among front offices. In my explanation, you can see how translatable those stats such as botched retrievals are to a team’s development and what they work on in practice.

What I did

Instat provides play-by-play data for all games. Using R, I created algorithms to track key sequences of events. I developed three key microstats: first, botched or failed retrievals; second, retrievals that successfully lead to exits; and third, zone entries that result in shots. I used simple logic to identify in what sequence events happened. For example, when doing botched retrievals the program identified things such as puck losses after a dump-in by the opposition and inaccurate passes by the team I’m conducting this for.

Final Thoughts

I understand how this might raise eyebrows among more traditional coaches or management. However, I’m just working with what I got. Maybe it’s not 110% accurate but if we can have something that is 90% accurate and use these in-depth details of the game to make our team better by having more successful retrievals or being able to identify what entries led to the most shots on goal, then I view that as a valuable tool.

Leave a comment

Subscribe now

Process

So how does this work? Well, what I did was I took all data from the 2023/2024 regular season and placed all of the stats into percentiles. For this, we had 100 being best in the league, 50 being average, and 0 being worst in the league. This ranking system allows us to compare players across various metrics, making it easier to identify standout performances and areas for improvement. By visualizing stats as percentiles, we can quickly assess how a player measures up against the rest of the league, providing insights that might be overlooked in raw data. Ultimately, this approach gives us a clear snapshot of a player's performance and helps guide further analysis.

Options

The fun thing about having publicly available data is we can use just about anything we want and we can make it just about as complex as we want. In this case, if you wanted to look at things strictly from an offensive perspective you could change out the legend to be playmaking, scoring, and overall offense. You can take it in any direction you desire.

Usage

So how can we use this visual? As I mentioned, the sky’s the limit. In the case of the polar chart here today, this gives analysts and coaches a small glimpse into what a player is about from a simple but wider perspective. Is he good at generating offense? Can he hold his own in the defensive zone? How is his usage and is he winning those minutes played from not only a shot but a quality point of view?

Wrap Up

By utilizing tools like the polar chart, we make it easier to enhance our understanding of player performance in a visually engaging way. This not only aids in the evaluation of current players but also could inform scouting and recruitment strategies, ultimately contributing to more data-driven decision-making in the sport. While these stats shown in the visual are important and can be changed based on preference, it is even more important to get an even deeper understanding. For example, I heavily use these visual types in my player development reports but I dive incredibly deeper into each stat such as looking at the entry locations, seeing how the player uses his body in battle situations, analyzing retrieval routes and quick passes. Ultimately, combining these detailed analyses with visual tools creates a holistic approach to player evaluation. This ensures that we not only identify trends but also understand the context behind them, leading to more informed and effective decision-making in the world of hockey analytics.

Leave a comment

Launch Angle

In baseball, a huge area that can be tweaked by development personnel and analysts is the angle at which the ball is launched, aka pitched. We can relate this to hockey in which the release point or angle is where, from load to shot, the puck is released. Small adjustments in release angle can significantly impact the trajectory and speed of the puck, creating more deceptive or powerful shots. For instance, changing the release point can throw goalies off, making it harder for them to predict where the puck is going. By optimizing the release angle, players can also find better shooting lanes in traffic and increase their chances of hitting high-danger scoring areas. Within the launch angle, you can dive into optimal stick flex, optimal stick height, hand positioning on the shot, and weight transfer.

Shifting

Another extremely analytically driven aspect of baseball is the concept of shifting based on the player at bat. This concept was heavily fought against by traditional players and management. From a hockey perspective, we can immediately think of team systems. So things such as a breakout or an offensive zone system like 1-2-2 or 1-2-2. Teams traditionally follow the exact same systems maybe with the exception of when the team is struggling. We all know that teams have pre-scout meetings where they analyze and identify where to expose opposing teams’ systems so why not switch things up to throw in a surprise element and to make themselves less predictable?

Pitcher/Player Biomechanics

Biomechanical data is where baseball sets itself apart from other sports. Hockey and baseball are similar in some ways when you think of the bodily rotation of a shot and swing of the bat. Some immediate thoughts of hockey biomechanics are stride length, stride frequency, knee bend, power transfer of a shot, and hand positioning. That is only the beginning.

Why Does This Matter?

Ultimately, this is a very niche aspect of analytics. As I said, baseball is extremely advanced when it comes to development and how to use analytics and biomechanics to analyze a player. Because of that, the opportunity for someone to revolutionize the use of this data in hockey is more open than ever. If someone was to do this there is no doubt in my mind that as the generational turnover of management occurs, they’d be swept up by a team very quickly.

Leave a comment

Subscribe now

Problems with Analytics

One of the main components that analytics do not account for is the human aspect. Players are going to have down games or even down weeks. You may never know what’s going on. Maybe that person had a challenging work day, they could be sick, or maybe they have family troubles. Not only that but everyone knows that more often than not a player will play through injury. That’s the culture. You can never know.

Insights

So how do we utilize this visual? Well, initially we can look at this from the development perspective. If a player or multiple players for that matter are having their stats decrease towards the end of the season, that may be an indication of conditioning levels or workload management. This could result in refined practice plans or possibly even doing more off-ice work in comparison to on-ice work. The second analysis we can identify is analyzing trends in a player’s game, for example, we can see this player plays generally bottom-six minutes in which he’s thriving from an offensive and defensive standpoint. For a different player maybe he is having trouble entering the zone with control so we can generate a presentation to show optimal entry locations and options he may have that he never noticed in-game.

Concluding Thoughts

Being able to provide context to a player’s game is always the goal of my analyses. Having the ability to say this is why you’re struggling to score goals or this is how you can position yourself to generate more high-danger chances is ultimately what my role is. By using different types of visuals we can get down to the nitty-gritty version of a player, both the strengths and weaknesses.

Leave a comment

Subscribe now

In today’s constantly innovating world of sports technology, being a conduit of information is an extremely valuable role. My goal in generating post-game reports was to provide coaches with a way to get the most important information to their inboxes immediately after the data became available. With that, I wanted to share how I built these reports. While it takes many lines of code, you don’t need to be an expert coder to create them. In the reference image below, you will see the team box score at the top and the player box score below.

Packages

One of my most utilized and favourite packages in R is GT. This package aims to create “a wide variety of useful tables.” You can find an intro to the package here. Along with GT, I also use the packages dplyr, scales, and readr. The dplyrpackage is essential for data manipulation, scales assists in infrastructure changes in ggplot2, and readr is used to work with data in CSV form.

My Process

Based on my experience playing the game and working with coaches, I wanted to ensure one key aspect when building out my report styles: creating something that isn’t fancy but maintains professionalism. Building something digestible and straightforward was my highest priority. Additionally, I made sure to have a master document that explained every visual within my reports. This was essential to avoid any standstill time if the person interpreting the visual had a question about a column name or my process behind the calculation.

Another critical aspect was ensuring that the data presented in the box scores was exactly what the coach wanted. Every coach will have different preferences for which stats matter most to them.

Closing Thoughts

I really enjoy these reports because they look sharp and add substance to the traditional box score. These reports can be easily created for any league and automated. If you want to build these reports, feel free to message me directly, and I will assist in any way I can! If you’re interested in learning R, I highly recommend this guide made by the current analytics coordinator of the New Orleans Pelicans.

Subscribe now

Leave a comment

Individualized development and individualized attention is one of the most powerful things you can have as an organization

Using Data

All professional sports teams not only have a player development staff but they have access to millions of data points collected during practice and game settings. Professional teams have access to one thing many leagues don’t have: accessibility. How these teams use this data is the sticking point. For this, you need a process of how you go about evaluating players and then how you want to present that information. Whether it be looking at positioning for how a player can increase their retrievals in the offensive or defensive zone or identifying the player’s optimal shot locations and types. Data is not only for game prep and in-game decision-making.

Roles

A very common misconception in analytics is when player A is signed by team B in free agency and is expected to excel due to a past season's analytically encouraging performance. This is where an even more in-depth look into data is essential. The next season playing on his new team, team B, player A suddenly struggles heavily. When looking into the numbers, we may see this player was tasked with playing not only more minutes a night but playing those minutes versus tougher competition. By putting the quality of competition into perspective we can see how the player was set up to fail. This also isn’t to say that a player can’t suddenly take on a larger role. This simply provides insight into how and why a player may perform a certain way.

Innovation and Research

If I was a general manager in the NHL my main focus would be on implementing data into my player development operation. It’s not as easy as saying this will suddenly equate to every prospect or free agent acquisition becoming a star or top-six player. The common saying is the analytics don’t say do this, they say if you do this you have the highest probability of achieving the most favourable outcome. In pro sports minor tweaks or changes in a player’s game can be turning replacement-level defensemen into a reliable defensive defenseman. While to you that may not be much, those small margins are what sets teams apart. After all, winning is the only thing that matters. Rather than identifying value, an organization can take it upon itself to create value. If you are curious about how data can assist in player development operations check out Kyle Boddy, the creator of Driveline Baseball.

Subscribe now

Leave a comment

Player Development

Research and Findings

A correlation matrix is an excellent use of numbers to identify what matters. For my research, I conducted the process by using a U Sports team with a focus on how skaters obtain points. In doing so I was able to find that the top three indicators of points were 1) shots on goal, 2) passes to the slot, 3) entries, and 4) scoring chances. On the other end of that scale, I found the stats that displayed the largest negative impact on points to be 1) puck battles won %, and 2) overall puck losses.

What does it tell us?

My immediate reaction to the results was I felt they made subjective and objective sense. Upon digesting the results, I wanted to brainstorm how coaches and analysts can use this on the player development side of operations. There are so many components to what makes a good hockey player. By having these results, we can use them as focuses during practice and video sessions. To create passes to the slot, we need to look at our team tactics in the offensive zone. How can we open up space, and what opposing players can we target to create a competitive advantage? When it comes to entries, what breakout systems lead to a successful entry conversion? What routes does the opposing team limit and leave open, and how can we operate in the neutral zone to create those routes for our teammates? The sky’s the limit.

Further Research

The MLB and baseball world in general have so many great online voices who oftentimes end up using that platform to propel themselves into a front-office position. The biggest flaw I see in hockey and the NHL is the accessibility of this data. I firmly believe that making data accessible to fans will create even more fan engagement and excitement for the league, as well as provide exposure to those who choose to create content using the data. Ultimately showcasing extremely talented minds for front offices.

It’s playoff season. Teams travel throughout the province playing on average three games a week. As a spectator, I came away with a few findings. Particularly, I found it abysmal how coaches lacked adaptation game by game. For team anonymity, I’ll be referring to the teams as Team 1 and Team 2. Consistently throughout the series, ultimately a sweep, the coaches on Team 2 made zero tactical adjustments against Team 1 who were picking them apart with their speed. It could be a testament to how prepared and educated Team 1’s coaching staff was (they went on to win the championship). The team observed throughout the regular season was not the juggernaut we saw in the playoffs.

PK 1-1-2 NZ

Team 1’s execution of their penalty kill system was something of an art. The leading forward pressured just enough while Team 2 team fell right into the trap of taking the wide route and then being cut off by the middle ice forward and single defensemen. While team 1 was conservative in the NZ, in the OZ they were hounding opposing defensemen as they attempted to backtrack in their zone to execute a retrieval.

Dump and Chase

While the data can’t be completely relied upon due to the difference in league and quality of competition, spectators aren’t blind to the concept that controlled entries outweigh uncontrolled entries. Owning possession upon entering the zone generates chances. It’s not rocket science. Although that may be true, a smart team forces their opposition to dump the puck forcing a possession change. Team 1 simply executed that concept to perfection. While I did not track any games, I can’t count on both hands how many times Team 1 forced Team 2 to dump the puck into their DZ and came away with the possession retrieval leading to a clear path not only through the neutral zone but into the offensive zone with ease.

Conclusion: Adapting

Team 1 controlled every aspect of this first-round series. When Team 2 did manage to get some momentum in games it was due to stretch passes up ice beating Team 1’s forwards. Team 2 also went 4/46 on the powerplay. I don’t know whether to say credit to Team 1 or shame on Team 2. Along with that, Team 1 scored 10 goals while shorthanded. Talk about powerkill. So why is this of value? As a coach, this provides immense insight into decision-making. You can’t wait until game 5 and then believe you’re going to suddenly execute to perfection. Along with that, changing systems/tactics come time for the postseason, as Jared Bednar did versus Winnipeg, allows your team to catch the opposition by surprise. As a staff, you can assume your opposition pre-scouted your regular season play so they feel prepared for what you’re going to throw at them come time for the playoff matchup. I am a firm believer in preparation, if we as researchers, analysts, and coaches can identify effectiveness and efficiency through systems, creating a response toolkit would be something I’d push every staff to develop.

Data/Images courtesy of Hockey Tactics 2024, Thibaud Chatel, Murat Ates (The Athletic)

Leave a comment

Subscribe now

The Math

P = GF/(GF + GA)

In the formula, P is equal to the percentage of games the team is expected to win, GF for total goals for, and GA for total goals against.

Putting the Concept into Action

If we wanted to look at this from an individual perspective, we could look at WAR by individual players. With this, we could play around with adding or subtracting players. Also known as marginal analysis. For example, according to the Evolving Hockey WAR model, Zach Hyman has contributed 4.6 wins to the Oiler’s season equal to 8.9 standings points. In taking the Oilers Pythagorean Win Expectation of 45 wins or 91 points, without Hyman, the Oilers would have 40.4 wins or 82 points. That represents a significant contrast. Here’s where we get into the limitations.

A way we can directly correlate the stat into the game is with regards to the Conn Smythe. The official definition of the award is “ the most valuable player for his team in the playoffs”. In the playoffs, winning is the most valuable contribution you can make to your team. The stat provides a clear winner based on the team’s playoff performance. In addition to that, traditional hockey fans know that the award is close to never being given to a player on the team who didn’t win the cup. The stat itself will automatically have this in mind because the higher number of goals scored generally means a team went far in the playoffs. For example, in last year’s cup final the Golden Knights and Panthers were 1-2 in goals scored so the stat assumes this.

Limitations & Thoughts

The flaw in this equation is that the number can be very inflated due to single-game scores. For example, the Canucks lost to Minnesota 10-7 in a game that was a crap shoot. Another example is the Penguins played the Sharks back in November and put up 10, inflating their goal total due to the quality of the competition. You can look at it from the opposite angle well every team played the Sharks twice so it cancels out.

Another way we can view this is from the concept of supply and demand. The Oilers signed Zach Hyman to a seven-year contract for 5.5 million dollars annually. In the case of Hyman's impact on the Oilers, attributing 4.6 wins to his presence doesn't fully account for the dynamic nature of team composition and roster management. Without Hyman, the Oilers would likely have utilized available cap space to fill his position with another player, potentially yielding a similar contribution to the team's performance. Therefore, while individual player metrics provide valuable insights, they may not accurately capture the true impact of player movements on team outcomes due to the principles of supply and demand in the player market.

Conclusion

Inspired by the work of Bill James, his concepts hold significant promise for advancing analytics in the NHL. All that needs to be done is the transferring of concepts from baseball to hockey. By utilizing metrics such as Pythagorean Expectation and WAR, fans can play around and gain insight into just how valuable their favourite players are and how a team would be performing based on their goal-scoring differential. However, it's crucial to recognize the limitations and complexities involved in the concept, such as the impact of single-game scores and the dynamic nature of roster management.

Data Courtesy of Evolving Hockey & Get In the Game by Tim Chartier

Subscribe now

Eye Test

In a limited role, Puustinen has thrived. In only 10 minutes per game, the Fin ranks fifth on the Penguins in points/60. Sure, the team isn’t good, but this guy is outperforming Kris Letang, Rickard Rakell, Reilly Smith, and oh yeah Erik Karlsson. Following the points, he is a great passer and overall playmaker. He ranks first on the Penguins in total and first in assists per 60. So he isn’t a secondary assist merchant. As a matter of fact, although he only has 2 goals on the season, he’s in the 71st percentile of shot speed in the league. Not bad. At 24, Puustinen can certainly be a presence in the top six for years to come with time spent working with the player development staff and in the gym.

Analytics

Let’s go back to the shot speed but from a generation perspective. His ability to create chances is at a high level as well. In fact, based on Evolving Hockey’s model he should have close to six goals already. Finishing is certainly a skill but I’m confident the scoring will come as he grows. With everything in mind, it’s important to look at things from a quality perspective. It’s fair to say take this with a grain of salt considering the weakness of the competition Puustinen is playing against. Ranking 20/32 on ice time per game, Puustinen has been excellent in limiting quality chances against. Every source of offence, Puustinen is limiting it. High quality or low quality.

Outlook

As an upcoming RFA, you can expect Penguins management to resign Valtteri Puustinen. However, there’s no doubt it’ll be on a two-way contract. I’m of the firm belief that Puustinen can be a threat, especially on the powerplay. With the changing dynamic of Penguins hockey, Puustinen has the legs to kill off the rush. Something the Penguins haven’t had for three years. Say he doesn’t progress, he is a great bottom six and PP2 option. I imagine him as a staple near the left dot at 5v4.

Data courtesy of Natural Stattrick, Evolving Hockey, NHL Edge

Subscribe now

The Pittsburgh Penguins suffered immensely all year with finding a reliable bottom-six however, the search is over. The Pittsburgh Penguins had horrible injury luck this year which led to some interesting lines. The team has finally found exactly what they needed. Frederick Gaudreau. With a full lineup, it gives Mike Sullivan the chance to have a solid bottom six. Might I say, it was at the perfect time, during the playoffs?

Frederick Gaudreau Complements Third Line

Position in the Lineup

Although he suffered heartbreak vs this team back in the 2017 Cup Final, Frederick Gaudreau has fit in perfectly in Pittsburgh. In March he was slotted into the fourth line due to the injury of Teddy Blueger and Evgeni Malkin. He then made the jump to third-line center over Mark Jankowski who immensely underachieved this season with horrible offensive numbers with six points in 45 games. Gaudreau has played extremely well in the system racking up 10 points in 19 games along with two points in three playoff games. The aged centermen/winger has brought size and skill to the Penguin’s third line along with Jeff Carter and Jared McCann.

Style of Play

You’d be surprised by how quick this guy can be with his six-foot frame. He brings a wide range of tools to the team. He has the ability to score goals, throw hits, shoot the puck, play defence, and drive. He’s almost like a bigger version of Brandon Tanev. Everybody loves a Tanev-type player. Gaudreau’s biggest attribute is his two-way game. He makes a big impact offensively and defensively which is what Pittsburgh’s third line is all about, all three players on that line have good offensive and defensive impact. You can see the amount of energy Gaudreau brings when on the ice. He makes quick decisions in transition and is always the first one on the forecheck. In the playoffs, his line has been lined up against the Komaro, Barzal, and Eberle line which is filled with skill so you can see how much Mike Sullivan trusts them to shut down top units.

Analytics

Currently playing at 3.18 points per 60 rate according to moneypuck.com. Gaudreau has shown up big for the Penguins in the first three games of the playoffs. To put that into perspective, Sidney Crosby is playing at 0.98 and Kris Letang at 2.16. During the regular season, he played in 19 games and carried a solid Goals Above Replacement (GAR) rate of 3.6 while maintaining a Wins Above Replacement (WAR) rate of 0.6. Not bad for an undrafted player. Although the current line he plays on has only played five games together, they carry goals for per 60 rate of 8.99 and a goals-against-per-60 rate of zero. You can see he’s impactful in both zones. It isn’t a large sample size but if he got to play the full 56 you’d expect around the same numbers.

Future

After this season the 28-year-old will be a free agent but I have a feeling he’ll be back. He carries a low cap hit of $700,000 and his value won’t rise too much so the Penguins could easily resign him for the same amount or a bit more. He’s shown he can be a depth player for this team who seems to contend year after year. He’s a Pittsburgh-type player and I highly doubt they let a two-way player like him walk away after the impression he has made on not only management but fans too. With the Penguins being tight with cap, he’s a solid depth player who doesn’t cost much at all, hard to pass up resigning him.

Data courtesy of Evolving Hockey & CapFriendly

Subscribe now

Towards the end of 2021, Evan Rodrigues was undoubtedly one of the top and most exciting players in the NHL. His release on a wrist shot was a thing of beauty when he had the puck on the right wing. Since then he’s cooled down. It doesn’t help he’s playing in a more defensive bottom-six role. With 37 points in 61 games in the 2021-22 season, there’s no doubt he’s had an amazing year compared to expectations. The Toronto native has been no more than a fourth-line player in his 8 year NHL career. With his career-high being 29 points while playing for the Buffalo Sabres, it was worth sending Dominik Kahun, who’s no longer an NHL’er, the other way.

Evan Rodrigues Shines Playing In Top Six For Penguins

Style Of Play

“Erod” as they call him is a Swiss army knife player. The most notable part of his game is the shot. His ability to load up and release is a thing of beauty. He can be a playmaker and sniper with his shot and awareness. The Toronto native is a zone-entry beast, averaging over 26 entries per game. It doesn’t stop there, he’s generating over five scoring chances off the rush, which is up there with Hart trophy candidates Johnny Gaudreau and Connor McDavid.

Besides the offence, his ability to play defensively is what makes him so valuable. Rodrigues has excelled in both a shutdown role and a top-line scoring role. As he’s cooled off since his miracle run, it’s easy to see he’s overthinking things. Who wouldn’t if they had 2 goals since February 1st? However, after collecting 40 points in 68 games this season, it’s fair to say he’s surprised all Penguins fans.

Contract Situation

If you had asked Penguins fans what Evan Rodrigues was worth in January, they would have told you at least 3 million. Since then, his play has very much diminished. With only 2 goals since February 1st, it’s fair to say he won’t be getting the big-ticket raise. This may be a blessing in disguise for the Penguins as they have some major negotiations this summer.

However, in the final year of his contract, the 28-year-old can hope for a raise, as he deserves with the way he raised his game in the absence of the stars. He’s one of the main reasons the Penguins are where they are. With no scoring while the stars are out, who knows where they’d be in this tight Metropolitan division race?

Outlook

After the addition of Rickard Rakell, it was hard to identify where exactly down the lineup he would be. He’s had time on a wing with Sidney Crosby, Jake Guentzel, Jason Zucker, and Kasperi Kapanen. He’s also slid in at center quite a few times in addition to being on both powerplay units. He can play anywhere if you maximize the opportunity for him to use his offensive game. Unlike his high point of the season, he’s in a goal-scoring slump while shooting only 8.72%. But Evan Rodrigues has a prime chance to solidify a spot come playoff time in the Penguins lineup. Although he’s cooled off, “Erod” deserves more love from Steel City for his magical performance this season. He’s not going to get the big payday but there may be some teams wanting his services. Only time will tell his future.

Data courtesy of Evolving Hockey

Subscribe now