There is no more contentious prospect in this years NHL draft than Kingston Frontenacs winger Lawson Crouse, who is very likely to be selected in the top 10.
Any discussion that involves Crouse becomes divisive almost immediately. Its not because of a forged birth certificate allegation, or because of his nationality, or that hes been injured the whole season or that his position tends to be risky to project; and hes not a KHL flight risk.
No, this debate is about a much simpler issue: Crouses production in junior -- or a lack thereof relative to most elite CHL prospects.
Even with that in mind, Crouse is a very good prospect, and deserves to be taken in the top 10 this summer, regardless of his lackluster production on the score sheet.
NHL draft prospect analytics
In the past five years, the public demand for and usage of advanced statistics have risen considerably in hockey; not surprisingly, that has spilled over into draft and prospect analysis. Its been no secret that there is a significant relationship between scoring in junior and NHL success. A bare-bones study of just points per game of first-year, draft-eligible CHL players shows a 40 percent correlation between junior-level scoring and NHL games played. In other words, this 0.4 correlation means that 16 percent of NHL games played are thanks to a players scoring rate in junior.
Where does the other 84 percent come from? Well, it could be attained in part with a bulkier statistical algorithm. And scouts also identify particular cases that deviate from the production.
But, theres also a lot that prevents stats from fully predicting the future of NHL prospects. For one, luck is a major factor in the NHL draft, in that teams rarely show the ability to reproduce drafting results season to season at any notable level. We see sample size issues, as well: When trying to use an AHL season to predict production the next season in the NHL, that one season of data can account for only about one-third of the translation to the NHL. This is important because most CHL players dont get a lengthy sample to be evaluated statistically before they are drafted.
As is the case with NHLers, bare-bones offensive stats can miss critical context, such as how a player is deployed by the coach, defensive value and luck.
Most of the more complex algorithms dont significantly outperform the NHL teams in terms of drafting skill, and all tend to hover around the same output level. The key is that the stats find different good players than the scouts do (while also whiffing on different players). The key difference tends to be assessing small forwards better.
Given the large body of work showing the relationship between junior scoring and NHL success, any evaluator would be a fool not to pay serious attention to the numbers. However, one has to be careful about putting too much emphasis on a players numbers. In baseball, this issue is sometimes explained via the credo you cant scout a box score. Falling in (or out of) love with a prospect because of his stat line, without giving the particularized evaluation and analysis that you have to do for each case, is a major mistake.
Lawson Crouses case
NHL teams certainly look at a players production closely. Teams have wrapped up their midterm scouting meetings the past few weeks, and a number of NHL talent evaluators (but not all) have told me that Crouses production was a negative on him during their evaluations.
One NHL scout said on this issue, "Ideally, you want guys like [Mitch] Marner and Dylan Strome, where the production backs up what you are seeing on the ice. We all understand that stats arent the end-all be-all, but it does matter."
The following table shows data from the past 10 seasons in the OHL and includes players who were a top-10 draft pick and played in the OHL in their 17-year-old season. 2015 draft-eligible players are bolded, and the stats are as of games ending Feb. 8. Note that the goals created (GC) metric cuts out 50 percent of the value given to assists:
Code:
Age-17 Season Production Of Future Top-10 Picks
Player Games Goals Points Pts/Gm GC/Gm
Connor McDavid 30 25 76 2.53 1.68
Sam Gagner 53 35 118 2.23 1.44
Mitch Marner 47 36 94 2.00 1.38
John Tavares 59 40 118 2.00 1.34
Steven Stamkos 61 58 105 1.72 1.34
Dylan Strome 50 33 94 1.88 1.27
Tyler Seguin 63 48 106 1.68 1.22
Nail Yakupov 65 49 101 1.55 1.15
Sam Bennett 57 36 91 1.60 1.11
Jeff Skinner 64 50 90 1.41 1.09
Ryan Strome 65 33 106 1.63 1.07
Taylor Hall 63 38 90 1.43 1.02
Michael Dal Colle 67 39 95 1.42 1.00
Matt Duchene 57 31 79 1.39 0.96
Logan Couture 54 26 78 1.44 0.96
Cody Hodgson 68 40 85 1.25 0.92
Sean Monahan 62 33 78 1.26 0.90
Mark Scheifele 65 22 75 1.15 0.75
Jordan Staal 68 28 68 1.00 0.71
Bo Horvat 67 33 61 0.91 0.70
Lawson Crouse 37 19 30 0.81 0.66
Nazem Kadri 68 25 65 0.96 0.66
Nick Ritchie 41 18 35 0.85 0.65
Josh Bailey 69 22 61 0.88 0.60
Gabriel Landeskog 61 24 46 0.75 0.57
Alex Galchenyuk 2 0 0 0.00 0.00
The obvious outliers here are Sam Gagner, who played on one of the greatest CHL lines in modern times, and Alex Galchenyuk, who played in just two games in his age-17 season. For the most part, sorting by GC per game seems to show a reasonable picture of how prospects have developed.
What offensive stats obviously fail to capture is defensive value. If you see some of the highly successful, low-scoring forwards from that chart, such as Gabriel Landeskog and Jordan Staal, another success story such as Ryan O'Reilly (drafted outside the top 10) and potential future success such as Bo Horvat, the common theme is their great all-around play in junior; for the first three, this type of well-rounded play was evident almost immediately in the NHL. This is relevant because of Crouses significant defensive value.
Some might argue against Crouse, saying this is the same old routine of old, cranky scouts falling in love with "size" and grit; that an NHL evaluator will see a big player, and focus too much on that attribute instead of the other data that predict success more often, such as production. I wont pretend that bias is never in play, but its unlikely to be a huge influence, and its too simplistic in describing this particular case because Crouse is a multidimensional player.
[+] EnlargeLawson Crouse
Claus Andersen/Getty ImagesPlaying for Canada as an underage (or double underage) player is a rare feat.
Crouse has many positives to his game other than size. He skates well in general, and very well for a big man. The same can be said for his puck skills, even if hes not a dynamic offensive player. Hes a very advanced two-way thinker who is one of the best defensive forwards available in this years draft class. And yes, he has an elite physical game, as well, thanks to his size, strength, work ethic and physicality that allows him to win a lot of battles.
ESPN colleague Keith Law once wrote that the point of scouting is to develop a mental database over time to identify trends in players, as well as particular skills and physical tools that have predictive value. The idea is that scouting hundreds if not thousands of players during a long time period creates value for the scout in identifying these particular cases. Although scouting data might not be binary like certain stats, it can still be used in a predictive manner. For example: A player who skates at X level, with Y hands and Z hockey sense tends to slot into a certain role in the NHL.
In the case of Crouse, the scouting analysis shows us that we have a winger who has skating and puck skills that are good (but not great), is great defensively, has high-end hockey sense and has an elite physical game. That combination of skills projects a reasonable ceiling of that player being an above-average scoring NHLer who clearly isnt top-line caliber in scoring but plays very well versus the opponents best forwards and can be one of a teams top penalty killers. If that outcome is realized, this is a very valuable asset for any NHL team and certainly for any of the teams looking to be picking in the top 10 of this year's draft.
Crouse has also shown he can play well to earn his accolades. He played his way onto the under-18 team as an underage player last spring and onto the top line of the Ivan Hlinka tournament team and was a rare double underager for Canadas under-20 team at the world junior championship. Its very rare for Canadian prospects to do all of the above on merit, but he has always earned what he has received in this regard.
Crouse is worthy of a top-10 pick
Its the job of the evaluator to understand the statistical evidence, be able to properly identify when a particular case differs from the long-term trends and balance these competing factors. Lawson Crouse is one of those cases when the balance has to be done very carefully.
The lack of production this season is a notable concern, and it keeps me from pushing him into the top 5-7 range of my rankings. But balanced against the merits of his physical abilities, work ethic and hockey IQ, he remains a great prospect.
Any argument now that Crouse is not a first-round prospect -- or even a top-20 pick -- is entirely unreasonable. I have him slotted as the 10th-best prospect in the draft, and theres a reasonable case to be made for his being taken even earlier than that.