Built from tracking data and film study
Rookie camp opened last week, which means everyone gets a prospect list. Here is ours, and it is built differently.
Most rankings start with viewings and use numbers to confirm what the eye already decided. I went the other way. I pulled InStat tracking data for every Calder-eligible player in the system, then pulled full league exports so each guy could be measured against players his own age, at his own position, in the same league he actually played in. That is roughly 9,000 skaters across thirteen leagues as the comparison set. Then I went back to the film to see whether what I was watching matched what the numbers said.
Sometimes it did. Sometimes it did not, and those are the interesting ones.
Every player mentioned in this piece has a card on the site. You can pull any of them up here, see every league he played in, and check my work against the raw percentiles.
After I had my list, I read the two that matter. Charlie O’Connor and Alexander Appleyard at PHLY published their sixth annual rankings last week, and if you are not a subscriber I highly recommend it. Charlie is one of the best Flyers writers out there. Steven Ellis at Daily Faceoff published his top ten in August.
I agree with those guys more often than I disagree. That is not a coincidence, they are very good at this. But where my list moves a player, the number that moved him is right there in the write-up. Argue with it. That is the idea.
Three ground rules before the list.
Everything here is 2025-26 only. Not career work. A player coming off an injury year or a fourth-line role reads that way.
Calder eligibility, same cutoff PHLY uses. That graduates Denver Barkey, who Ellis had fourth on his list under a different rule.
Cross-league comparison is the whole problem. A point in the J20 Nationell and a point in the SHL are not the same object. So I never compare raw rates across leagues. Everything converts to a percentile against age-matched peers inside each league, and those are what get compared. When a player has multiple leagues, he gets multiple rows, and the gap between them is usually the most interesting thing about him.
1. Porter Martone, RW, 19
No argument here and I am not going to invent one. Martone graded 93rd percentile in offensive volume among NCAA forwards born within a year of him, and 96th in play driving. Then he arrived in Philadelphia and put up 10 points in nine games.
What the data adds is that the NHL cameo was not a fluke of shooting luck. His territorial results held at 60th percentile against NHL forwards his age, on 16.9 minutes a night, in a playoff race. The offense traveled. The play driving traveled.
The soft spot is real and worth naming: 18th percentile in defensive work in the NHL sample, 20th in the NCAA. He does not retrieve pucks, he does not take them away in his own end, and he does not finish checks as often as I assumed he did before I went looking. That is a fine trade at his production level.
Knowing Tocchet, the inevitable scoring slump this year is going to be telling. Just because you do not have your A game every night does not mean the night is an automatic F. You work on your B and C game so you can still contribute off the scoresheet. That is the part of Martone’s development I will be watching.
Below I’ve included both Martone’s NCAA comparison (if his game PERFECTLY translated to the NHL from the NCAA) and his NHL metrics (playoffs & Regular Season) vs other guys.
2. Oliver Bonk, RHD, 21
My first real departure. PHLY has him third, Ellis third, and both frame him as a defense-first blueliner. Ellis specifically describes him as known more for his defensive game than his offense.
The data says the opposite, and once I went back to the film with that in mind I could see it. Bonk graded 87th percentile in offensive volume among AHL defensemen his age and 38th in defensive work. He played 21.8 minutes a night for the Phantoms, which is a workhorse number, but the shape of what he did in those minutes tilted toward offense. He is a puck-mover still learning the defensive side, not a defensive defenseman with some offense.
That is a better prospect, not a worse one, and it is why I moved him up.
The season arc everyone describes checks out cleanly. The first ten games were bad, the last 36 were good.
3. Yegor Zavragin, G, 20
Everyone has Zavragin second. I have him third and I want to be honest about why.
The VHL numbers are outstanding. A .948 save percentage, 99th percentile in that league, with positive goals saved above expected. Ellis is right that second-tier Russian save percentages run inflated, and my data supports that caution, but Zavragin was well clear of the inflation.
The KHL is the problem. In 12 games with SKA he was 59th percentile in save percentage and, more tellingly, negative in goals saved above expected. He conceded slightly more than the quality of shots he faced would predict. The Larionov benching after the October CSKA loss does look like a coaching decision rather than a performance one, but the underlying number does not scream that a starter was wrongly buried.
The tools are not in question. He is a superb athlete with genuine starter potential. This is a ranking of what happened last season, and last season was not the jump forward his D+1 year was.
4. Alex Bump, LW, 22
Bump’s AHL profile is one of the best in the system: 91st percentile offense, 98th percentile play driving among AHL forwards his age. When he moved up, 71st percentile offense in 24 NHL games at 12.8 minutes. His closest overall profile match is Artturi Lehkonen, which tells you exactly what kind of winger this is.
The gap in his game is distribution. 25th percentile in playmaking in the NHL, 28th in the AHL. He shoots, he drives the net, he does not create for others. That is less a flaw than a job description, and the chemistry with Barkey shows up in the tracking as a real on-ice results spike rather than a vibes thing.
He earned the roster spot. The open question is whether a shoot-first winger with limited passing holds a middle-six job when the shooting percentage regresses.
5. Jett Luchanko, C, 20
Moving him up, and the rookie showcase against the Rangers this weekend is why.
Public sentiment on Luchanko fell further than on any prospect in this system over the past year, and the data was already kinder to him than the discourse. In the OHL he was 86th percentile in distribution and 90th percentile in play driving among forwards his age, on 19.5 minutes a night. Those are not the numbers of a bust. They are the numbers of a good passer who does not finish, which is the actual concern with him.
His seven NHL games are a horror show offensively, 1st percentile, but 83rd percentile in defensive work. At 19, in a tiny sample, that is a kid clinging to the parts of his game he could execute at that level.
Here is what moved him up. Watching him this weekend, the core muscle injury that hung over him for a year and a half is clearly behind him. The first three steps are back. He looked like the fast-moving playmaker we were all so high on in 2024, not the version that spent last season a half-step late. A full year in Lehigh Valley at 20 minutes a night should do the rest.
6. Max Westergård, LW, 20
My biggest riser. PHLY has him 17th.
The J20 Nationell numbers are absurd: 98th percentile offense and 100th percentile play driving among Swedish juniors his age, on 20.3 minutes a night. That is not a good junior season, that is the best territorial season in his entire cohort.
The counterweight everyone cites is two points in 35 games against men. But look at how those games happened. His SHL ice time averaged 6.4 minutes. A dozen appearances were under four minutes and one was 41 seconds. You cannot score in 41 seconds. That is not a player failing against men, that is a player not being played.
When he did get real minutes against men, in the Champions Hockey League, he posted 97th percentile play driving and 76th percentile transition. Eight games, so hold it loosely, but it points the same direction as the junior work rather than the opposite one.
I am betting big on Max’s upside. He just needs to earn his coach’s trust.
7. Brek Liske, RHD, 18
Liske was 84th percentile in distribution and 92nd percentile in play driving among WHL defensemen his age, playing 21.8 minutes a night as a draft-eligible. His closest profile matches on playmaking are Thomas Chabot and Mikhail Sergachyov.
The scanning habits and crisp passing you see on film are not a projection here. They are measurable and they sit at the top of his league. His year-over-year is the part that sold me: his distribution jumped from 50th percentile to 84th and his offense from 38th to 65th, across two full seasons of 79 and 68 games. That is a step change, not a draft-year blip.
He needs to work on his skating to produce at the next level and I am not going to pretend otherwise. But right now, in the WHL, it is not stopping him from moving pucks at an elite rate for his age. Appleyard makes the case he should have gone earlier on day two and I think that is right.
8. Jack Nesbitt, C, 19
Moving him down, and the reason is a number that got worse rather than better.
The concern with Nesbitt has always been skating. My transition score, which measures carrying pucks in and out of zones against giveaways, has him at 35th percentile among OHL forwards his age. Meanwhile his offensive volume is 93rd percentile and his play driving is 90th.
Now add the year over year. In 2024-25 his transition graded 48th. This season it fell to 35th while his offense climbed from 87th to 93rd. Both samples are large, 77 games and 68 games, so this is not noise. He is producing more while transporting the puck worse relative to his peers.
That combination only exists when a player wins through size, reach and hockey sense rather than feet. It works in junior. The gap between 93rd percentile production and 35th percentile transition is exactly the gap that closes on you when everyone around you gets faster.
Michigan is the right landing spot and a college schedule with heavy weekday strength work is precisely the intervention this profile needs. But I am not ranking him ahead of players whose underlying game is trending the right way.
9. Spencer Gill, RHD, 20
The most uncomfortable ranking on my list, and I want to flag exactly why before anyone gets excited.
The numbers are terrific. 86th percentile offense, 88th percentile play driving, 78th percentile transition among QMJHL defensemen his age, on 23.6 minutes a night, the heaviest workload of any prospect in this system. His overall profile matches are Brock Faber and Owen Power.
Now the caveat. He played 32 games. Thirteen were playoffs. Two games in October, then nothing until February. So this is a small, late, playoff-weighted sample from a player who was fresh while everyone around him was worn down.
What pushed me to rank him here anyway is the prior season. In 2024-25 he played 51 QMJHL games and graded 94th percentile in play driving and 92nd in distribution. Two seasons of strong numbers, one of them full length, is a different animal than one hot spring.
He is still the guy on this list most likely to look silly in six months, in either direction. Lehigh Valley this year will tell us.
10. Maksim Sokolovskii, LHD, 18
PHLY has him seventh, Ellis tenth. I land at ten, and the data reads him as a narrower player than the hype suggests.
Sokolovskii was 77th percentile in defensive work and 75th in physical play among OHL defensemen his age. He was 42nd in offense and 30th in distribution. His closest overall matches are Joel Edmundson, Zach Whitecloud and Carson Soucy. That is a specific and useful type of player, and it lines up with Ellis calling him a shutdown option.
The number that gives me most pause is deployment. He averaged 14.0 minutes a game, which ranks 18th percentile for ice time among OHL defensemen. For a first-round pick on a strong London team, that is sheltered usage. The game log shows it plainly. He opened the season at 25:41, fell under five minutes by November, and only climbed back over 20 for the playoff series.
This was a real swing by Flahr and Danny B. He is either the best pick at the end of the first round in years or a bust who keeps getting looks because of his size. Year two in London, with a bigger role, will start to tell us which.
11. Jack Berglund, C, 20
On résumé last season was massive: World Junior gold as captain and leading scorer, an expanded SHL role under Cam Abbott, then a senior World Championship call-up.
The data is the dissenting voice. Against SHL forwards his age he was 62nd percentile in offense, 33rd in distribution, 48th in transition, 48th in play driving. That is an average SHL season for a 20-year-old, which is fine, but it is not the profile of a player who moved up a tier.
The role change is visible and real. His ice time went from 38 seconds in October to 16 and 19 minutes by late January. That trajectory is the best argument for him and it is what I would weight most.
He is also the clearest example of why the player cards split leagues the way they do. Berglund’s U20 line is 17 points in 14 games. His SHL line is 62nd percentile. Both are true and only one of them tells you anything about his NHL future. He should dominate juniors. He is 20.
12. Carter Amico, RHD, 19
In the USHL, Amico was 82nd percentile in defensive work and 84th in play driving among defensemen his age, on 21.3 minutes a night. His closest overall matches are Simon Edvinsson and MacKenzie Weegar, which for a 6’5″ right shot is exactly the shape you hope for.
The Boston University stretch was as bad as advertised: 28th percentile offense, 8.6 minutes a game across 16 games, several appearances of two minutes or less. That is a player being hidden.
O’Connor weights the development camp and summer showcase work heavily and is explicit about it. I cannot measure that, so I am not going to pretend to. Worth noting that Ellis published in early August, before the showcase, and left Amico out of his top ten entirely. That disagreement is mostly about calendar, not evaluation.
13. Cole Knuble, RW, 22
The consensus ceiling on Knuble is a likeable bottom-sixer. I think that sells him slightly short.
At Notre Dame he was 90th percentile in offense and 90th percentile in distribution among NCAA forwards his age, on 20.2 minutes. Then in seven AHL games he posted 97th percentile playmaking. Tiny sample, but it points the same way.
The reason he is not higher is that his physical and defensive scores are mediocre and his size is unremarkable, which is the real ceiling argument. But the passing is a genuine NHL skill, and the idea that he could force his way up the ladder the way Barkey did looks well founded to me.
While this weekends game did not HURT his stock it does have me curious to follow along with his game in Wilkesbare this year.
14. Shane Vansaghi, RW, 21
The book on Vansaghi is that the offense has not emerged and he projects bottom six. The data pushes back a little.
He was 73rd percentile in offense and 84th percentile in play driving among NCAA forwards his age. The team tilted the ice when he was out there. That is more than a fourth-line profile.
Where the scouting take is clearly right is everything requiring touch. 44th percentile transition, 40th defensive work. The fork in front of him is real: either the handling improves or he simplifies until only his strengths matter.
His four World Junior appearances are worth ignoring entirely. He averaged 6.4 minutes and his usage declined every single game.
15. Martin Psohlavec, G, 18
My sharpest disagreement with PHLY, who have him ninth.
On the surface his Czech U20 season is excellent: .931 save percentage, 94th percentile in the league, 97th percentile in goals against per 60.
Then you look at what he faced. Psohlavec’s shot difficulty faced ranks in the 0th percentile of his league. He faced the easiest shot diet of any goalie in the Czech U20 Extraliga, and his goals saved above expected came in slightly negative. Once you account for quality, he performed marginally below what an average goalie would have managed behind the same team.
Appleyard is upfront that Psohlavec was seen by many as a mid-to-late round pick and that the Flyers bet on the athlete and the person rather than the current product. That is a perfectly reasonable way to draft a goalie, and the point about 6’5″ athletes making up a large share of modern NHL netminders is well taken.
I just cannot rank a save percentage that the underlying numbers say was manufactured by his defense. Saskatoon and the WHL will tell us far more.
16. Heikki Ruohonen, C, 21
An even profile with no holes and no peaks: 62nd percentile offense, 71st distribution, 61st transition against NCAA forwards his age.
The environmental defense of his production is legitimate. Harvard did not have a single ten-goal scorer last season, so everyone there was suppressed. His World Junior work, nine points in seven games leading Finland, shows what happens when the surrounding talent improves.
His closest overall comparison is Morgan Geekie, which is a genuinely useful middle-six outcome. He is a few years away and a point per game at Harvard this season is the marker to watch.
17. Matthew Gard, C, 20
Gard projects as a matchup center leaned on for tough minutes rather than a scorer, and the numbers land in the same place: 65th percentile physical play, 61st defensive work, 34th distribution, 28th play driving.
That last number is the concern, and the year over year makes it worse. His play driving fell from 59th percentile in 2024-25 to 28th this season, across 66 games and 60 games. Both samples are full. The production ticked up after the trade to Seattle without the underlying play moving at all.
He is what he was drafted to be. Whether that reaches the NHL depends on his skating, which still needs significant work.
18. Carson Bjarnason, G, 21
There is no way to soften this. In the AHL, Bjarnason ranked 8th percentile in save percentage, 13th in goals against per 60, and 7th percentile in goals saved above expected. He faced a roughly average shot diet and conceded far more than it warranted.
The fair counterpoint is that goalies take time and that three and a half strong months preceded the wheels coming off.
Here is the thing that reframes the Kolosov comparison below. Bjarnason faced lower expected goals against per 60 than Kolosov did and still posted worse results. The read that Kolosov was the better goalie of the two last season is not just an impression. The tracking data says it outright.
His card also shows the level change, which matters here. He went from 76th percentile in save percentage in the WHL to 8th in the AHL. That is not a slump inside a level. That is a collapse on arrival at a new one.
19. Jack Murtagh, LW, 20
Murtagh is the only forward on this list with no category above the 55th percentile. Offense 29th, distribution 21st, play driving 30th, on 12 minutes a game.
The hockey IQ questions that followed him into the draft show up here as clearly as anything in this data set. He was buried in the bottom six for over half the season, which is both an explanation and a symptom.
The tools are real and I am not disputing them. Great skater, good frame, fantastic shot. But a full NCAA season produced nothing measurable, and this is a ranking of last season. Worth saying that he is one of the youngest players in this group, having turned 18 barely two months before the draft, so there is more runway here than the ranking suggests.
20. Aleksei Kolosov, G, 23
Kolosov’s AHL season was mediocre rather than bad: 38th percentile save percentage, 30th in goals saved above expected. As covered above, still better than Bjarnason’s.
His four NHL appearances should be excluded from any evaluation. Two were relief outings, including one where he faced three shots and stopped none in eight minutes. That kind of appearance wrecks an aggregate and tells you nothing about the goalie.
2026-27 is likely his last chance with this organization, and with Vladar and Woll ahead of him that seems right.
Honorable mentions
Nathan Quinn, Hunter McDonald, Ty Murchison, Alex Ciernik, Marek Sklenicka, Max Laatikainen.
On McDonald and Murchison, O’Connor’s response to a reader in his own comments is the right framing: they are ahead on today’s depth chart and behind on ceiling, and a prospect list should reward the latter.
What the numbers could not see
Three honest limitations, because a list like this is only worth anything if you know where it is thin.
Goalie exports carry a fraction of the detail skater exports do, with no age data at all. Zavragin, Psohlavec, Bjarnason and Kolosov are ranked on a thinner evidence base than everyone else, and I would rather say that than pretend otherwise.
Tracking data cannot see skating mechanics. Heavy feet in tight spots, a stride that is strength-limited rather than habit-limited, the things you pick up in twenty seconds of film. There is no instrument for that. Where I talk about skating in this piece, it is coming from watching, not from a spreadsheet.
And InStat game counts do not always match official league counts. Gill reads as 19 regular season games in my data and 16 in PHLY’s. Bump reads as 37 AHL games where Ellis has 35. Small gaps, but they are real, and anyone rebuilding this work should expect them.
Every player in this piece has a full card on the site. Each one breaks his season out by league, scores him against players his own age in that league, tracks him year over year, and shows the NHL profiles he most resembles. Browse all of them here.
Leave a Reply