Using cutting-edge AI analysis, we dive deep into the upcoming NHL season to uncover the top 5 players projected to face regression. Discover the statistical insights and trends that suggest these players may be poised for a challenging year ahead. Get ahead of the game with our exclusive player projections and stay tuned to see if the AI's predictions hold true on the ice.

The 2022-2023 season saw quite a few offensive explosions from players who were not expected to make the jump. Many of those players may find themselves primed for regression this season for a variety of reasons. Some are related to a change in playing situation, while others are related to an inflated shooting percentage or PDO factor. Tim Stützle, despite offensive explosion, is not a regression candidate because of his age, playing situation and development curve projections. Ditto for Gabriel Vilardi whose situation has changed drastically - he will see far more offensive situations to generate scoring opportunities. 

When predicting regression candidates, Kevin looks at the performance of the previous season, the factors that impact scoring and accounts for manual inputs such as a coaching change or situational change (positive/negative). Those will be automatically accounted for once the first game is played, but there needed to be something to predict for that prior to the season. Once all factors were considered, the player's performance and volatility was measured to understand how consistently they perform to expectations. Over or under performing shows volatility meaning the players who over performed, but were highly volatile, are the biggest candidates from regression. 

Erik Karlsson

Karlsson's inclusion on this list should come as no surprise given his performance lat year came out of seemingly nowhere. The one major change Karlsson benefitted from was the trade that sent Brent Burns to Carolina. This freed up significant offensive opportunities and allowed Karlsson to play more free. In the process of accepting for change in situation, Kevin would've projected Karlsson for 79 points last season, not 101. It would have been a significant increase from his previous seasons simply accounting for a situational change, but not to the extent of a 66 point increase. 

Karlsson's situation changed again, however he will find himself with another elite defenceman in Kris Letang. It seems as if Pittsburgh intends to run a power play with two defencemen, meaning he will see significant time. However, he will not have the same ice time as he did in San Jose, but the talent around him increases significantly. 

Karlsson is projected for a range of 72-89 points this season, right around a point per game. He has the most volatile projection in the data set. Generally speaking, the point projection ranges are 6-10 points, not 17. Due to the volatile nature of Karlsson's play and the change in situation, his projection has a wide range. HIs most likely outcome is 83 points, right around a point per game player. While that is still in the NHL's elite, it is a progression of about 20 points. 

Andrei Kuzmenko

Kuzmenko's appearance on this list is two-fold as well. He exploded onto the scene in his rookie season and put up 74 points and 39 goals. However, Kuzmenko shot 25% or better from November to April. The best shooters in the history of hockey shoot 18%. Kuzmenko will certainly see significant regression of his shooting percentage. A player of Kuzmenko's talent has a shooting percentage 12%. If Kuzmenko regresses to anywhere near that number, his goal totals will fall significantly. It is more reasonable to expect Kuzmenko to score 25 goals, as opposed to 40. 

Kuzmenko found serious chemistry with Elias Pettersson and there is no reason to believe those two won't continue to see success together. He's going to continue to see heavy offensive zone starts and premier power play minutes. There is a new coach running the power play this season, the impact on point totals is currently unknown. Kuzmenko's true talent as a finisher is not 28%, it is likely less than half of that. While that is still elite, his shooting percentage would be enough to land him on this list without consideration of harder matchups, different special teams setups and volatile performance. A more reasonable expectation for Kuzmenko this season is 22-27 goals and 60-70 points. 

Brandon Montour

Montour finds himself on this list for a few reasons, the most significant being his injury. Had he remained healthy, he would've been a candidate for regression this season from his 73-point campaign last year. Accounting for his increased role at the beginning of the season where he benefitted from Ekblad's power play minutes, Montour was projected to score 54 points, a 17-point increase from 2021-22. Montour scored 19 points in the first 20 games of 22-23 and 25 points in his final 20 games. He was averaging over a point per game in those stretches. In the middle portion of the season, his production dropped 34%. That level of volatility causes the model to have concerns about his play in a full season.

An important note is that Montour saw increased special teams opportunities with Aaron Ekblad missing portions of the season. Montour will likely not have those opportunities when he returns to the lineup in December. He is unlikely to be a three shots per game defenceman and is projected around 2.1 upon his return - a 30% drop. Missing training camp and the beginning of the season has a significant impact on player production in the dataset going back to 2005. Players who miss training camp through injury or contract dispute see between 20-30% drop in production for the season. Combine the factors together and it is easy to see Montour's path to regression. If Montour plays 50 games this season which is a conservative estimate, Kevin projects him to produce 30-35 points, a projection that accounts for games missed, missed training camp, volatility and regression. 

Josh Morrissey 

Any defenceman that doubles their point totals season over season to record more than 70 points is an easy candidate for this list. Morrissey had an outstanding year for the Winnipeg Jets last season and found himself in the Norris conversation for a large portion of the season. Morrissey started his regression towards the end of last season and it coincided with the Jets tumbling down the standings and barely squeaking into the playoffs. Buoyed by a 50% jump in shooting percentage, Morrissey led Jets defencemen with 16 goals and 76 points last year - 43 points ahead of Neal Pionk, who finished second. 36 of Morrissey's 76 points were secondary assists, nearly matching his career high for points in a season (37). That number trailed only Karlsson, Quinn Hughes and Rasmus Dahlin last season. Karlsson is a regression candidate while the other two are younger defencemen improving their offensive game at the NHL level.  

Morrissey's production on the power play (28 points) nearly doubled his previous career high total of 15 points. Winnipeg's top power play unit was not buoyed by a better on-ice shooting percentage, in fact their success rate dipped 2% from 2021-22. The Jets were the benefactor of more power play opportunities, and scored more goals with the same shooting percentage. It is reasonable to believe the secondary power play assists last season are not repeatable to the same extent as goals or primary assists. While Morrissey is due for regression, it may not be as stark as some other players because the power play produced at a sustainable rate last season. Kevin projects Morrissey's goal total between 9-13 in 2023-24 and a point total between 55 and 63 points. 

Ryan Nugent-Hopkins

107 points. 107. No one expected Ryan Nugent-Hopkins to crack the 100-point plateau 12 years into his career. Buoyed by an incredible shooting percentage jump and a historically good power play, Nugent-Hopkins wasn't the only Oiler to see a significant rise in production last season. Playing with two of the generations best talents can have that effect sometimes. Nugent-Hopkins is an elite player, he's more likely a mid-80s point player with true talent on this Oilers team. 

Nugent-Hopkins was the benefactor of an 18.4% shooting percentage last season, more than six percentage points better than his career shooting percentage of 12.2%. His shooting percentage is highly likely to regress this season. It is reasonable to believe his shot attempts will continue at a similar clip given how the Oilers like to play. With this is mind, a 12.2% success rate on 200 shots is approximately 25 goals. This is slightly below Kevin's projection of 28 goals for Nugent-Hopkins. 

The main reason his production is likely to dip this season is the power play. There is no denying the talent on the Oilers power play is terrifying. However, they are unlikely to repeat the same level of success this season at 32.4%. If the Oilers power play succeeds at 28% this season, they will score 77 goals on 275 attempts, 12 goals less than last season. Nugent-Hopkins scored 21.6% of his total career power play points last season (53 of 245). 25 of his 53 power play points were secondary assists which have proven to not be as repeatable in consecutive years. The combination of shooting percentage, power play production and secondary assists make Nugent-Hopkins a prime candidate for regression in 2023-24 to his projected totals in the mid to high-80s.