Earlier today I published an article on Hockey-Graphs on prospect statistics such as NHL translation numbers and how we can improve upon them (read here for more). The write up is centred on combining both quantitative and qualitative information in improving the overall system when projecting future prospects. The article would serve as an excellent primer to those who have not read about NHL translations previously.
One way to look at a player's development while decreasing factors like on-ice and personal shooting percentage variance is combining seasons. Instead of looking each draft year in junior as separate time periods, we use it to increase the sample with the year(s) prior.
We will turn to the Winnipeg Jets draft prospects and look at what these numbers suggest of their performance.
Jets Forward Prospects
* Mark Scheifele, Adam Lowry, Austen Brassard, Lucas Sutter, Scott Kosmachuk, Ryan Olsen, JC Lipon, Andrew Copp are currently in or have passed their Draft+3 season, and are shown for contextual reference.
* This method helps show potential outliers, such as Lucas Sutter's and Jimmy Lodge's draft seasons.
* Nic Petan has not increased much in scoring over his three most recent seasons in junior. His draft production was elite, and it could be that he merely reached the limits in offence that one could produce in his circumstances.
* Jimmy Lodge is interesting case. His numbers were awesome in his draft season, causing high expectations. When he failed to improve upon his numbers there were some concerns, but this shows that his overall performance has still be promising regardless.
* Axel Blomqvist has performed excellent as a free prospect, producing similarly or more to many of the Jets later round draft picks.
* Nikolaj Ehlers had elite production for his draft season, especially when you add in contextual nuances like age, league and era. He started off slow this season, but is slowly creeping up to last season's numbers despite playing on a weaker team.
* Chase De Leo has some promising top six upside as a high scoring late round pick, much like Jimmy Lodge. Due to his birthdate, De Leo is in his final season of junior eligibility (unless he comes back as an OA player) despite being in his Draft+1 season currently. This brings up the debate on whether it is better to judge players by their birth year or by their draft year. The answer likely is neither; the best model will likely use players age in detail (ex: 18.45 years old).
* C.J. Franklin was a solid scorer in the USHL, and has since adapted well to the NCAA in his Draft+2 season.
* The MHL has gone through a lot of changes over the past few seasons, so it is entirely likely that Pavel Kraskovsy's numbers are less reliable than the others. He has shown flashes of high end skill and has experience playing in pro with the KHL.
* Matt Ustaski impressed some at the Jets summer development camp, but these impressions have yet to translate into high end scoring.
Jets Defensive Prospects
* Brennen Serville, Zach Yuen, Aaron Harstad, Jacob Trouba, Brenden Kichton, and Tucker Poolman are currently in or have past their Draft+3 season, and are shown more for contextual reference.
* Josh Morrissey is in a bit of a scoring slump this season -while playing on a decimated Prince Albert squad- but his numbers still indicate him as one of the most offensively gifted Jets prospects yet.
* Here we can see the difference between players who are known more for their defensive acumen -like Brennan Serville and Nelson Nogier- and players listed as more "two-way" style -like Jan Kostalek and Jack Glover"
* For those wondering about Tucker Poolman, he did not play in a league with NHL translations until his Draft+2 eligible season. The Jets drafted Poolman that very year, so all his recent development has been his Draft+3 and Draft+4 seasons.
The current standard for prospect evaluations is NHL Equivalencies, also known as NHLe. These are created using a standard regression curve looking at players who play in the league of interest and then move to the NHL the following season.
The largest issue with these numbers is they lack context. Does a player score well because he's talented or because he has a fortunate shooting percentage year? Are other factors inflating his numbers, like line mates, line matching, zone deployment, or ice time? Does he have factors that project him to improve more than the average individual does as they grow?
These are all areas we are still trying to understand better.