2026 NHL Draft Rankings

AFP Analytics is pleased to release prospect rankings/tiers for the second consecutive season and fifth overall. We continue to make tweaks to how we come to these results.

The Model

Our model is ultimately a regression model that seeks to predict how likely it is a player plays 200 games. This year there were some behind the scenes tweaks that we’re going to see how they play out before discussing further. The goal is to better credit players, mainly those playing in Europe, who end up playing at different levels during a season. We haven’t fully integrated these tweaks across all leagues so we will hold-off discussing much further. Ultimately, the model looks at players on a league-by-league basis and uses past player results to predict the likelihood of a player reaching 200 NHL games. Players who generally have strong box score numbers at a younger age will grade out better. Some leagues have shown that additional factors like size and penalty minutes may also play a role.

Getting to Results

One of the big tweaks we are making this year is how we are ultimately coming to our tier rankings this year. In an NHL front office, well-run organizations would be blending their scouting department assessments and analytical models. This year we have taken our model output and blended with multiple public draft analysts’ rankings to come to a consensus. On the Max Term Podcast, we discussed players that we believe could outperform their expected draft slot and those we might be cautious of.

One of the points to emphasize is there are players that might not grade out particularly well in our model because the playstyle/skillset they bring to an organization is not focused on offensive production. Take Albert Smits for example. He is well-regarded by many public draft analysts because he is a steady, modern “stay at home” defenseman. His game isn’t necessarily focused on offensive output, which means he will not grade out overly well. Simon Edvinsson was a similar type of defenseman who did not grade out well in the model in his draft year. He has certainly developed into a high-level NHL defenseman. Analytics and “the eye test” will not always agree. The important thing in this situation is an understanding of why they differ.

The Results

Below, you will find the top four tiers of our final rankings, encompassing 42 players. For presentation purposes, we are grouping by tiers and not presenting individual probabilities.

Observations

Although we have not included the player’s probability of success, this year’s draft class is incredibly strong. Not since the 2020 draft class has the top end graded out so strongly. Gavin McKenna and Ivar Stenberg grade out very well. Even though we have split our top nine into two tiers, those players in tier two this year would be tier one in every other draft class we have modeled except 2020. Further, this is by far the strongest defensive draft class we have modeled in our five years. Rudolph, Bleyl, Reid, Carels, and Verhoeff all have the profiles to be top four defensemen in the league for many years to come.

Conclusions

There are players that our model is much higher and much lower on than the “scouting community” but many of the same sentiments around this draft align with the output we are generating. This draft is setup to be a very interesting top ten next week in Buffalo. Outside of McKenna going first, the combination of teams and an abundance of talent make the rest of the top ten very intriguing. We look forward to hearing everyone in Buffalo boo Gary Bettman on June 26^th.