There is no shortage of draft rankings available today. This aggregator helpfully collects 14 of them. So, why add another one?
The primary reason is my own interest. Perusing endless scouting reports and new stats tools (like those provided by McKeen’s and extraskater), it’s hard to resist the urge to gather one’s wits and all the available information in an effort to render some kind of judgment, to put my own stamp on the draft class and to refine what it is I think I know about projecting prospects into the NHL.
In part, I imagine all idiosyncratically compiled rankings are something of a journey of self-discovery. I don’t mean anything grandiose like young Werther discovering his sad little life is better abandoned to the garbage heap of history. Rather, I mean, when one sits down, attempts to bracket off biases, and attempts to come to some judgment concerning these jejune princes on ice, avoiding the knowledge of what one actually values is near impossible. As the internal debates over this or that player rages on, one unearths a nascent drafting philosophy. What one thinks matters about a player becomes clear.
For this reason alone, I think everyone interested in drafts and prospects ought to attempt the exercise. You can learn a great deal not only about the players on offer, but also about what it is you value in a prospect.
I publish this ranking, however, because I think it offers the reading public a novel and worthwhile view on the draft class of 2014.
Draft Ranking Philosophy
This is not a mock draft. I have taken no concern for draft order or team needs.
This is not a personal ranking. Or, this is not the order I would choose the players in if I were a GM. If I were a GM who cared little for the opinion of others and was content to allow my biases to roam free, I would take Leon Draisaitl first over-all and probably again with every subsequent pick. That kind of bias is indefensible from any reasonable point of view and moreover it is useless to others.
This is an attempt to compile a objective list the way a head scout would for his GM. Obviously, I do not see this ranking as bias-free or rising to the height of Gord’s-eye-view. Nothing can reach that level of objectivity. Nonetheless, this is an honest attempt to set aside, or at least vigorous interrogate, preferences and biases until I am satisfied a ranking can be justified on a variety of reasonable grounds.
So what are the reasonable grounds that inform my rankings?
In the advanced stats age of hockey, draft eligible players suffer from a epistemic deficit. We only have pieces of each man. We know so little about their usage, time on ice, shot generation, zone starts, linemates, etc., that we are often forced to take up the metrics of a by-gone age.
This doesn’t mean we need to restrict ourselves, however, to simple boxcars. Indeed, doing so can be deceiving. Have a look at this graphic TSN put up for their mock draft with Craig Button:
In this “head-to-head” graphic, TSN has listed the boxcars of two players but excluded some key information. While both players scored 105 points this season, Reinhart did so in 4 less games played. Knowing this helps us distinguish who had a better point-per-game scoring record.
But, we can do so much more than make this simple adjustment to boxcars. We can also split each player’s boxcars into discipline scoring (even strength, power play, penalty kill). Even strength scoring is the best way to evaluate and predict true talent. Power play scoring is highly reliant upon opportunity (a player not given a lot of PP TOI is going to see their boxcars suffer by comparison) and luck-based metrics like shooting and save percentage that tend to even out over larger samples.
In this ranking, I give a lot of weight to even strength scoring. This means not only splitting a player’s boxcars by discipline. It also means striping out empty net goals (as I reviewed the problem here, empty net goals can skew a player’s boxcars, goals for differential and IPP).
NHL equivalencies are an attempt to project a player’s boxcars (all disciplines) from a lesser league (be it KHL, SHL, CHL, etc.) into a full 82 game NHL season. These projections are based on historical data. It looks at how players from one league have performed the following year in the NHL in the aggregate. It is not meant to be fool-proof. It is meant to be an indicator of league strength and a player’s NHL potential.
I put a fair amount of weight on NHLE. Over time it has proven a reliable indicator of NHL scoring potential and it offers an intriguing insight into a player’s scoring talent and readiness for NHL work. For an 18 year old forward to make an impact at the NHL level, generally an NHLE of 35 or higher is a good line in the sand. Below that number and the player either won’t project as an impact forward, or needs more time to develop.
A player is draft eligible “who will be age 18 on or before September 15 in the year in which such Entry Draft is held.” So, the eligible players can be nearly a year apart and still belong to the same draft class. For obvious reasons, we should be cautious about over-interpreting the scoring of the extremely young and old players from their draft years.
I give a bonus to younger players on the reasonable assumption that we are seeing their scoring at an earlier stage of their development curve. They have more upside.
One of the problems with evaluating any player in any league is the problem of how to separate team success from player success. We often find ourselves praising or blaming players, irrespective of their individual talents and merits, for their team’s success or failure. Psychologically, this is nearly unavoidable.
There are, however, ways to mitigate this problem. One of the simple ways is to simply compare the player’s teams against one another in terms of winning percentage (I’ve listed the basic team data for all the CHL players). This only gets us so far.
A more promising route is to look at the goal differential (in all situations and at even strength) for each player’s team and then compare that to the differential when the player is on the ice. For example, if a team’s ES goals for percentage is 60%, but a player on that team’s on-ice ES goals for percentage is 70%, we can make some assumptions about that player’s scoring contribution relative to his team’s strength. Simply put, we can say that with that player on the ice at even strength, his team performs better than normal, or that he has a positive impact when on the ice relative to his team’s scoring strength.
By looking at players in this manner we can compare the contribution of all the players to their respective teams and normalize the question of team strength to some degree.
Another way we can look at players is by their individual points percentage (I’ve written about this here, here an here). IPP gives us a record of the contribution a player makes to the goals for scored while he is on the ice at even strength. It is an indictor not only of a player’s contribution to scoring (whether he drives the scoring bus on his team so to speak), but also of a player’s team and line strength. A player with a high IPP is heavily relied upon by his team to score points and more than likely receives little scoring support from his linemates.
I give a fair amount of contextual weight to these measures. They don’t tell us who is the best player, but they do indicate which players may have been benefited or hampered by the relative strength of their teams.
For all the on-ice goals for and against metrics, I’ve used extraskater’s data.
As a general rule of thumb, I value forwards, especially centers, above all else for draft prospects. A variety of people have looked at the problem of drafting goaltenders in the first round (see here, here, here and here). The problem there is development timeframe. Goaltenders take so long to develop, it is near impossible to project their talent to the NHL at the age of 18. The same is largely true of defensemen. A variety of people have also looked at this issue (see here, here and here). I’ve also looked at the problem of drafting and projecting defensemen here.
I am an empiricist.
A well crafted scouting report gives an empirical record of a player’s strengths, tendencies and range of skill. The trick in reading scouting reports is twofold in my estimation. First, it is important to read widely and attempt to find patterns from all the independent information. Second, it is important to try and filter out narratives. I classify a narrative as a piece of trivial information (i.e., it does not convey a record of events, or an empirical observation) that traffics in cliches, euphemisms, epithets, projections and biases of any form.
Examples of straight-forward empirical statements a scouting report might contain would be: “has a powerful first step”; “can make passes with his backhand and forehand”; “is aggressive on the forecheck”; “has a hard/accurate shot”; etc.
Examples of narrative driven statements a scouting report might contain would be: “he hates to lose”‘; “clutch performer”; “able to elevate his game”; “has poor body language/looks disinterested”; etc.
I give a fair amount of weight to scouting reports. They paint a valuable picture of a player’s attributes, style and range of skill. I am especially interested in learning about a player’s ability to manage the puck, play without the puck and create offence.
A Final Word on Market Inefficiencies
“… After his injury [Barry Fraser] retired to an office job with Ontario Hydro, but he maintained an interest in hockey by coaching young boys, and gradually the canny reports he sent in on prospects from his neck of the woods drew so much attention that he was swept into full-time scouting… it was on his word that the Oilers spent their valuable first-round draft choice on Coffey, had courted Anderson, and had gambled yet another draft choice on being able to talk Jari Kurri into coming to North America.”
~ The Game of Our Lives, Gzowski
For a variety of reasons, scouts and GMs under and over rate players. Chiefly, issues of size, character, nationality and playing style (watch for references to “European” or “Russian” or “flashy” style of play), come into play. For the most part, I see these concerns as so much silliness and I find NHL decision makers cut off their nose to spite their face on these matters. Recently, Steve Simmons of the Toronto Sun, wrote a profile 2014 draft eligible forward Josh Ho-Sang. Craig Custance of ESPN wrote on the same player and taking up similar themes here. While I don’t rate Ho-Sang as high as he or his agent rate him, it seems fairly clear to me that he is undervalued by the scouts and GMs of the NHL for a variety of tedious, trivial reasons.
While I don’t give players an edge on their peers because the market undervalues them (afterall this is a draft not a free agent signing), I also do not accept the concerns the market has for them. I don’t give much credence to issues of size and character and the like. While, it is certainly true that the bottle, for example, can derail a promising prospect just as much as injury, for the most part these kinds of concerns are way overblown.
Vital stats for CHL players are taken from their CHL team pages.
Boxcars and splits are taken directly from the CHL records. In the case of empty net goals, I have tracked them through the game reports of each player (I’ve found extraskater to be unreliable on this point).
Advanced stats such as estimated time on ice, points per 60, goals for percentage and IPP for CHL players all comes from extraskater. I’ll caution that these are estimates (the methodology of which is spelled out here) and not accurate records. And, I’ll note that I’ve found minor some discrepancies regarding the scoring metrics when cross-checked against the CHL’s game sheets. Despite this, I think the information is valuable and worth including.
All on-ice stats for SHL players comes from the SHL’s own records.
Vital stats for SHL players is taken from elite prospects.
Scouting reports are taken from a variety of sources and will be credited as mentioned.