SBI · Hlinka Gretzky Cup 2026 · Edmonton · Aug 3–8
USA named 23. Our algorithm agreed on 6.
On July 27, USA Hockey announced its 23-man roster for the 2026 Hlinka Gretzky Cup — 13 forwards, 7 defensemen, 3 goalies. We ran the identical 13F/7D/3G build through ProdigyPoints, our public composite score, over the same player pool (US players, 2009–2010 birth years). The two lists agree on 6 names.
Six shared forwards. Zero shared defensemen. Zero shared goalies. We are not asserting the roster is wrong — selection camps see things a stat composite can't. We are showing you the table.
Published 2026-07-29~5 min readHlinka series · Day 1Puck drop in 5 days
The average gap is not small. The 23 players USA Hockey selected average 3,056 ProdigyPoints. The algorithm's 23 average 4,059 — roughly a thousand points per player higher. Whatever the camp evaluators saw, it isn't what a production-based composite sees.
This has history
USA Hockey's Hlinka selections have drawn public criticism before — notable omissions from this roster are a recurring August storyline in American hockey media. We're not relitigating past rosters. But when a federation's picks diverge this consistently from production-based lists, the divergence itself becomes a measurable pattern — and that's precisely what the SBI series exists to measure.
Canada, meanwhile
Canada hasn't finalized its roster yet — 27 players are in camp in Edmonton through August 1, and the list gets trimmed before puck drop. One early observation from the camp invites: Landon DuPont, the consensus projected #1 pick for the 2027 NHL Draft, ranks 7th among Canada's own camp invites by ProdigyPoints. Charlie Murata leads. That's not a knock on DuPont — it's a clean illustration of what our composite measures (production, level, visibility) versus what scouts project (ceiling). When the tournament starts, we get six games of evidence about which lens sees sharper.
USA vs. Canada by the numbers
🇺🇸 USA (final 23)
🇨🇦 Canada (camp 27)
Average ProdigyPoints
3,056
2,412
Median ProdigyPoints
3,163
2,570
Players over 3,000 pts
12
2
Top player
Ryland Rooney · 4,637
Charlie Murata · 3,147
Read this table carefully before quoting it. USA's list is a final roster; Canada's is a camp roster that still includes players who will be cut — comparing them straight-up flatters neither. And ProdigyPoints weights league production heavily: USNTDP and USHL contexts score differently than CHL contexts. The number says USA's selected group out-scores Canada's camp group on our composite. It does not say USA wins.
🔒 30 seconds to unlock
The full 23-vs-23 table. Every name.
One email. We'll show you both complete lists side by side — every USA Hockey selection with their ProdigyPoints, and every player our algorithm would have taken instead.
The full side-by-side — all 23 roster picks vs. all 23 algorithm picks, with points
The blue-line divergence — the 7 defensemen the algorithm wanted, none of whom made the roster
The 200-point head start — how making this roster immediately moves every player's ranking via F17
What we're tracking as the tournament unfolds — the specific predictions this table implies
No spam. No password. Future SBI reports go to this inbox first.
The full table
#
USA Hockey's 23
Pts
Algorithm's 23
Pts
1
Ryland Rooney · F ✓
4,637
Aiden Kelly · F
5,169
2
Gunnar Conboy · F ✓
4,544
Caiden Pellegrino · F
5,068
3
Brooks DeMars · F ✓
4,441
Ryland Rooney · F ✓
4,637
4
Colin Grubb · F ✓
4,436
Gunnar Conboy · F ✓
4,544
5
Nolan Snyder · F ✓
4,318
Brooks DeMars · F ✓
4,441
6
Michael Tang · F ✓
4,226
Caine Bickel · F
4,441
7
Ethan Sung · F
4,162
Colin Grubb · F ✓
4,436
8
Gavin Burcar · F
3,782
Nolan Snyder · F ✓
4,318
9
Ty Bergeron · F
3,617
Oliver McKinney · F
4,256
10
Becker Wenkus · D
3,275
Landon Jackman · F
4,249
11
Luke Pietila · F
3,268
Sawyer Schmidt · F
4,241
12
Trevor Shorter · D
3,163
Michael Tang · F ✓
4,226
13
Cameron Minella · D
2,996
Enzo DiDomenicantonio · F
4,219
14
Nash Roed · F
2,900
Jason Segal · D
3,939
15
Gerry DiCunzolo · F
2,875
Noah Mazzola · D
3,854
16
Joey Cullen · F
2,872
Drew Roscoe · D
3,806
17
Rylan Hanson · D
2,429
Luke Warrener · D
3,572
18
Jake Prunty · D
2,264
Solomon Wolf · D
3,568
19
Mason Courville · G
1,523
Ivan Belikov · D
3,492
20
Kaeden Tate · G
1,270
Kage Flory · D
3,472
21
Beck Thoreson · D
1,233
Chase Pastor · G
3,245
22
Charlie Mineman · D
1,223
Sammy Inendino · G
3,108
23
Brandon Wright · G
826
Jesse Bernardinelli · G
3,059
✓ = on both lists. Points are ProdigyPoints (our public composite) as of July 29, 2026.
The blue-line divergence
The starkest split is on defense. The algorithm's seven — Segal, Mazzola, Roscoe, Warrener, Wolf, Belikov, Flory — average 3,672 points. The seven who made the roster average 2,369. Some of that gap is structural: several roster defensemen play NTDP or prep contexts where our composite has thinner inputs, and camp evaluators weight skating, defending, and system fit that box-score composites can't see. But a 1,300-point-per-player divergence on one position group is exactly the kind of number worth writing down before the tournament, so it can be checked against what actually happens on the ice in Edmonton.
The 200-point head start
Our ranking model has a factor called F17 — International Points. Any player who makes an IIHF-track international roster banks a 200-point base immediately, before playing a single shift. Team finish adds up to 500 more; individual performance adds up to 400, capped at 1,000 per season. Which means the selection decisions analyzed above don't just decide who flies to Edmonton — they directly move 23 players' rankings the moment the roster PDF goes out. Selection is a ranking event. That's why we watch it this closely.
What we're tracking as the tournament unfolds
The forwards the lists agree on (Rooney, Conboy, DeMars, Grubb, Snyder, Tang) — consensus picks should produce. If they don't, both lenses were wrong together.
The USA blue line — the position group where the roster and the algorithm disagree completely. Six games will generate real evidence.
DuPont vs. Murata — the consensus #1 2027 pick against the ProdigyPoints leader on Canada's camp list.
USA vs. Canada — our composite says USA's selected group is stronger on paper. The group stage has them in opposite pools; a medal-round meeting is the test.
When it's over, we'll publish the follow-up: what the table got right, what it got wrong, and what that says about selection — the same retrospective standard we hold every SBI piece to.
What this analysis doesn't see
Camp performance. USA's roster came out of a live selection camp in St. Paul. Our composite doesn't watch camp scrimmages.
Role construction. Teams need checkers, shutdown pairs, and penalty killers — not 23 leading scorers. A roster is not a leaderboard.
League context weighting. ProdigyPoints weights production by league level, but NTDP/prep/AAA contexts have thinner statistical inputs than the CHL or USHL. Some of the divergence is measurement, not evaluation.
Projection. Scouts pick for the player a 16-year-old becomes. The composite scores the player he has been. Both are legitimate; they are different questions.
What's next
Canada trims its camp roster by August 1 — when the final list lands, it gets the same treatment. The other six nations (Czechia, Slovakia, Sweden, Finland, Germany, Switzerland) publish rosters in the days before puck drop, and each one joins the series as it does. Follow along below.
FOLLOW PRODIGYCHAIN
Daily Top-10 player drops, AI scouting reports, more SBI breakdowns. We post where you scroll.