2025 NHL Draft Predictions: AI Revolution, Top Prospects & Key Insights

The 2025 NHL Draft is shaping up to be one of the most intriguing events in recent hockey memory, not just because of the dramatic lottery shakeup that propelled the New York Islanders from the 10th to the 1st overall pick, but due to a remarkable experiment: artificial intelligence predicting how the first round will unfold. As franchises gear up to make selections that could chart the course of their future, major opportunities—and risks—lie ahead for both players and teams. The growing influence of machine learning and advanced analytics is now steering the conversation, challenging scouts, executives, and fans to reconsider what “smart” drafting really means.

The Drama of the Draft Lottery​

Few moments ignite more anticipation than the draft lottery. For the Islanders, a franchise searching for renewed identity and a push back into the playoff picture, landing the top pick presents a golden ticket. The 2024-25 campaign saw them struggling for consistency, matching their number of wins with regulation losses—a stat that painfully highlights their difficulties in closing out games. While skeptics question whether any single player can truly transform a middling squad, the Islanders’ jump to the top slot offers hope for a franchise-altering selection.
But does this year’s class offer a true can’t-miss star? Unlike the 2024 draft, which was headlined by the consensus phenom Macklin Celebrini, the 2025 pool is seen as a little more wide open, with various analysts and AI models pinpointing different players who could make a franchise-defining impact.

Artificial Intelligence Steps Into the Spotlight​

USA TODAY Sports set out to answer one of the draft’s most tantalizing questions: If you assembled the brightest AIs—from Microsoft Copilot to Meta AI, Google Gemini, and OpenAI’s ChatGPT—and fed them the wealth of available data, would their first-round predictions converge on consensus? Or would they unearth rising stars and sleeper picks far outside the traditional scouting wisdom?
Leveraging advanced natural language processing, multivariate statistics, and predictive modeling, the four AI programs churned out a full simulation of the first round. Their results not only highlighted the known commodities but also spotlighted some surprising names projected to climb—and fall—in the draft order.

Breaking Down the AI Simulations​

The first question on every Islanders fan’s mind: who is the AI’s projected top pick? Both historically and in AI-driven models, the value of a high-impact defenseman or center at the top of the draft remains a prevailing theme. This year, Matthew Schaefer, a towering blue-liner with dazzling mobility and a projectable two-way game, emerges as a consistent favorite across several models. While Schaefer may not yet carry the household name cachet of recent top picks, the consensus among AI platforms is that his blend of skill, hockey IQ, and raw physical tools make him the most likely “safe” bet to become a cornerstone player.
The Sharks, who now pick second, reportedly favor high-skilled Swedish center Anton Frondell. If true, this represents both a positional need and a swing for upside—traits AI models tend to value highly, especially after analyzing Frondell’s international and professional play against older competition.
Notably, AI models also flagged some significant trends: Brady Martin, a player surging up the consensus rankings, vaulted into several simulations’ top ten, raising the possibility that a QMJHL or WHL standout could disrupt the proceeding’s North American bias. Meanwhile, teams like the Nashville Predators and Philadelphia Flyers, each with multiple first-round selections, faced challenging simulations around risk-reward—balancing high ceiling prospects against more established, “safer” picks.

First Round AI Predicted Order (per USA TODAY/AI Consensuses)​

1. New York Islanders
2. San Jose Sharks
3. Chicago Blackhawks
4. Utah Mammoth
5. Nashville Predators
6. Philadelphia Flyers
7. Boston Bruins
8. Seattle Kraken
9. Buffalo Sabres
10. Anaheim Ducks
11. Pittsburgh Penguins
12. New York Rangers
13. Detroit Red Wings
14. Columbus Blue Jackets
15. Vancouver Canucks
16. Montreal Canadiens (via Calgary)
17. Montreal Canadiens
18. Calgary Flames (via New Jersey)
19. St. Louis Blues
20. Columbus Blue Jackets (via Minnesota Wild)
21. Ottawa Senators
22. Philadelphia Flyers (via Colorado)
23. Nashville Predators (via Tampa Bay)
24. Los Angeles Kings
25. Chicago Blackhawks (via Toronto)
26. Nashville Predators (via San Jose from Vegas)
27. Washington Capitals
28. Winnipeg Jets
29. Carolina Hurricanes
30. San Jose Sharks (via Dallas)
31. Philadelphia Flyers (via Edmonton)
32. Calgary Flames (via Florida)

Context: Draft Day Logistics​

Hockey’s next generation will hear their names called at L.A. Live’s Peacock Theater in Los Angeles, with the first round scheduled for Friday, June 27 at 7 p.m. ET live on ESPN. Rounds 2 through 7 follow on Saturday, aired on NHL Network, with a rapid-fire pace that will see teams make or break years’ worth of scouting decisions in a matter of minutes. The convergence of glitz, media attention, and the analytic rigor of modern front offices ensures the 2025 draft will be scrutinized for years.

Comparing the 2025 Class: Strengths and Weaknesses​

Notable Strengths​

• Depth and Variety at the Top: While there may not be a consensus “generational” player, the class boasts a strong collection of forwards and defensemen projected as top-pairing or top-line talents. AI simulations, which aggregate vast scouting reports and player tracking data, consistently rate the first 6–8 picks as potential long-term NHL stalwarts.
• Emergence of Non-North American Talent: Several draft projections, both human and AI-driven, point toward an increasing presence of European players among the top 10, reflecting broader league trends. Swedish center Anton Frondell’s name surfaces often as a second overall pick—a testament to the strength of Sweden’s developmental system and the willingness of teams to trust European prospects in prime slots.
• Balanced Skill Sets: Many of the top names combine elite skating, possession metrics, and decision-making, suggesting they can contribute on both sides of the puck at the NHL level.

Areas of Caution​

• Lack of “Clear-Cut” Franchise Changers: Unlike recent drafts headlined by players with universal acclaim, this year’s crop appears short on certain-fire, franchise-altering talent. This “flattened” talent curve, while deep, places a greater premium on scouting nuance and risk tolerance. Even AI-predicted consensus at the top is cautious, frequently describing leading prospects as “projectable” rather than “transformative.”
• Skepticism Around Goaltending Prospects: 2025’s goaltending pool is relatively weak, with few netminders appearing in any simulation’s first round. The analytics bear this out, as most AI models prefer skaters unless a goaltending prospect demonstrably outperforms their peers in high-impact tournaments or pro-league stints.
• Horizontal Volatility: AI projections emphasize not only the ceiling but also the potential for “bust” outcomes in the middle-to-late first round—a theme echoed by scouts who warn that, outside the top dozen picks, teams must scrutinize development environments and medical histories closely.

The Method Behind the AI Madness​

How, exactly, do AI models arrive at their simulated draft boards? Each platform employs its proprietary blend of:

• Historical Draft Value Models: These algorithms correlate draft position and subsequent NHL performance to gauge the likely return on investment for particular player archetypes.
• Advanced Scouting Data: Using player tracking, zone entry/exit data, and even biometric information, these models can assess subtle aspects of on-ice performance that humans might overlook.
• Contextual Adjustments: Factoring in organizational needs, recent trade activity, and pipeline strengths, AI models simulate likely decision points—not just “best available” player lists, but “most likely pick” scenarios for each team.

While specifics differ, the result is a layered, probabilistic approach to predicting the draft—a method that can surface underappreciated talents (sometimes dubbed “AI sleepers”) and bring transparency to what has historically been an opaque process.

Validation, Consensus, and Contradiction​

Perhaps the greatest value in this AI exercise is not simply in more accurate picks, but in forcing a reckoning with institutional biases and blind spots. With AI, patterns of over- or under-valuing certain positions, sizes, or backgrounds are laid bare. Moreover, by running consensus simulations across multiple platforms, trends emerge:

• Agreement at the Very Top: Most AI models align on the same top 3–5 prospects, matching the wisdom of most human scouts.
• Divergence in the Mid-Round: Here, “AI vs. Human” debates rage, with some models identifying skill sets—such as elite lateral mobility in defensemen or puck recovery metrics in wingers—not always prioritized by traditional eyes.
• Outliers and Surprise Inclusions: When AI foresees a late first-round climb for a previously under-the-radar player, it often correlates with a spike in underlying performance indicators, hinting that teams with robust analytics departments could steal value late.

The Broader Impact of AI on NHL Draft Strategy​

Whether these AI-driven projections come to fruition or not, their influence on NHL front offices is undeniable. More teams are building their own proprietary data sets, hiring analytics specialists, and embedding AI into their scouting and draft decision architecture. The 2025 draft may well be a test case for how much “machine” influence is too much—or whether, in a league known for conservatism, innovation will finally be rewarded.
Risks remain: AI models can amplify the impact of poor historical data, underestimate the value of intangible leadership, or fail to account for personal adversity. However, their ability to cut through noise—synthesizing thousands of games and millions of data points—presents a compelling counterbalance to gut instinct.

Key Players to Watch: AI’s Top-Rated Prospects​

Matthew Schaefer (D, projected for New York Islanders)​

Praised for his elite skating, physical play, and “coachability,” Schaefer doesn’t just fill seats on the depth chart—he promises to steady a defensive corps that’s struggled to play a modern puck-movement game. AI models love his plus-minus relative to team, breakout zone control, and penalty-kill impact. Still, some caution that his point generation may lag top offensive blue-liners, making him less dynamic but highly reliable.

Anton Frondell (C, projected for San Jose Sharks)​

A deft Swedish pivot with a creative streak and strong two-way presence, Frondell has earned consistent praise for his ability to shoulder heavy minutes and make an impact against men in the SHL. Analytics highlight his transition game and high-danger shot setup—traits valued in today’s positionless play.

Brady Martin (F, climbing into top 10)​

A polarizing prospect, Martin’s stock has soared on the back of a tear through junior competition. AI’s admiration stems from exceptional microstats: offensive zone retrieval, rush creation, and a knack for drawing penalties. However, critics note his uncertainty in adapting to NHL physicality, making him a calculated risk-reward pick.

Tactical Implications for Teams With Multiple Picks​

Clubs like Philadelphia, Montreal, and Nashville, each with multiple first-rounders, face amplified pressure to maximize draft value. AI simulations often suggest a “barbell” approach—pairing safe, NHL-ready selections with high-upside, developmental talents in the same round. The Flyers, for instance, could emerge as a bellwether, depending on whether they trust AI’s suggestion to challenge conventional wisdom or stick with established playbooks.

The Entertainment and Unpredictability Factor​

While AI analytics make for fascinating content and endlessly debatable predictions, they cannot simulate the unique psychology of a draft floor. A single trade, an unexpected medical revelation, or an old-school “feel” pick could upend the board. Yet, as teams become more technologically sophisticated, the margin for “off-the-board” picks may shrink.

Future Trends: Will AI Dominate the NHL Draft?​

The 2025 NHL Draft signifies a tipping point: for the first time, predictive modeling has reached the mainstream, offering transparent, data-guided alternatives to “hockey man” tradition. Still, human elements—interviews, personal rapport, and small sample luck—retain outsize effects on draft day outcomes. The future might lie in hybrid approaches: letting AI set parameters but allowing seasoned scouts to provide context and “veto power” where necessary.

What Needs Watching in 2025…and Beyond​

• Injury and Development Tracking: Expect AI to improve at quantifying injury risk, recovery timelines, and development environments—a potential game-changer for prospects with uneven medical histories.
• Geographic Expansion: As more elite talent emerges from non-traditional markets (including Western Europe and the U.S.A.), AI’s ability to parse comparative league strength will be vital.
• Psychometrics and “Intangibles”: The next generation of models may yet find ways to predict leadership and “clutch” factor using biometric and psychological data—a controversial, but potentially revolutionary, frontier.

Conclusion: The Draft as a Mirror to Hockey’s Future​

The 2025 NHL Draft, fueled by an unprecedented infusion of AI insights, encapsulates hockey’s ongoing battle between tradition and innovation. While machines may never fully replace the drama and unpredictability of the human element, they are dramatically raising the standard for preparation and decision-making. As New York, San Jose, and the rest of the hockey world await the announcements from L.A.’s glitzy stage, one thing is clear: the future of the NHL is being written not just by scouts and GMs, but by the silent, inexorable logic of the algorithm. Whether teams lean into this brave new world—or resist it—will shape the league for years to come.

---

Source: USA Today NHL AI mock draft: AI predicts the first round of the 2025 NHL Draft