Dispelling the Myth: Where to Find a Goaltender in the Draft

By David A. Rainer

Different leagues play a different style of game — this is recognized across the globe. If you compare the style of game played in Finland with that which is played in Sweden, you will notice that the Finnish game is much more physical while the Swedish game has an increased focus on skill. Similarly, when looking at the CHL and NCAA hockey, there are striking differences in style. For example, the QMJHL traditionally is a high scoring league with more attention on getting up and down the ice while the WHL is more structured and the NCAA, with the elimination of fighting, is more attractive to smaller, skilled players.

Given these differences, can one league’s style of play produce a better player at a specific position than other leagues? Does the QMJHL produce a better goalie because of the number of shots faced? Does the NCAA produce a better forward because they emphasize skilled play? Does the WHL produce a better defenseman because of the attention to team play and structure? This study will examine this question and begin by focusing on goaltending. When an NHL General Manager is at the draft table, and with all other variables being relatively equal, can he lean towards one player over another simply by the league that player developed in with the confidence that the particular league has developed an NHL style of play?

When asked, General Manager of the Los Angeles Kings Dave Taylor alluded to the perceived differences in leagues. “The QMJHL is the biggest example or area that you can point to. A lot of that has been because of Patrick Roy and the popular butterfly goaltending style that so many young goalies try and emulate. I think you see good players out of a lot of different areas and at a lot of different positions. A lot of people think good, strong defensemen come only out of the Western Hockey League. There is some consideration that Europeans come with skill, while the Canadian players come with more of a physical game. There are certainly exceptions in different countries.”

Are these perceived differences reflected in statistical performance and are they accurate? The analysis shall begin with a look at the historical numbers. If an anomaly with the numbers surfaces, we can then consider possible reasons why the performance of one league can differ from another.

The first step in the analysis is to determine the best available neutral statistical category for success in the NHL. It is important to find a neutral category – that is, one that will not emphasize or de-emphasize a player based on style, period of time playing or other variables not within the player’s control – so that we can ensure oranges are compared with oranges. For example, the choice of goals scored would tend to unfairly de-emphasize the importance of a “stay at home” defenseman. Similarly, goals against average would show an unusually high quality in goaltenders during an age where goal scoring is depressed and unfairly devalue those goaltenders that toiled during the high-scoring years of the 80’s and 90’s.

Save percentage would be a more neutral category as the goaltender is more able to control his own fate in turning away shots and therefore is a more adequate valuation of his individual worth. However, there is a theory that save percentage is largely affected by the quality of team playing in front of the goaltender. This can greatly de-emphasize the quality of a goaltender playing on a bad defensive team.

It seems, that the only neutral category that can be identified is games played. Although not a statistic of star-power, to a large degree, games played reveals the relative worth of the player to the NHL. This is based on the premise that, irrespective of the type or amount of skills the particular player has, if he manages to play in an NHL game, he has some redeeming quality for which an organization is willing to pay. Therefore, the more games played by a player, the more or longer lasting his value has been to his team.

With the statistical category of games played in mind, let us set the parameters. Every goaltender drafted within the first nine rounds from the 1979 Entry Draft through the 1995 Entry Draft was recorded and categorized by the league the goaltender was drafted out of. Every goaltender drafted by an NHL organization during those years has been grouped into one of six categories: WHL, QMJHL, OHL, NCAA, USHS and Europe (all European leagues taken in conjunction). These groups were pitted against each other to determine which has been the most productive in terms of developing quality goaltending.

First, what percentage of draftees actually make it to the NHL? By this we mean the percentage of drafted goaltenders who play in at least a single NHL game in their career.

%League
63.5QMJHL
57.6WHL
53.0OHL
44.7Europe
28.9NCAA
19.6USHS

Notice that the three CHL leagues top the list and the QMJHL out performs the USHS by nearly three to one and the NCAA by over two to one. When looking simply at which leagues manage to develop a goaltender at least to the point where he is considered on an NHL roster, the QMJHL has to be the clear leader in production. To those that recall the perceived love affair by organizations with French-Canadian goaltenders, this should come as no surprise. But what might be even more telling is that the WHL places a close second behind the QMJHL while at the same time producing a higher overall total of NHL goaltenders (33) than the QMHJL (31).

Next, let us look at the average games played by those goaltenders that in fact reached the NHL. In other words, once the goaltender reaches the NHL, what is his expected career span.

Ave GPLeague
261.3WHL
196.2USHS
194.6QMJHL
189.3Europe
184.4OHL
127.5NCAA

As you can see, the WHL produces the highest average of games played by the goaltenders they develop. Not only does the WHL outperform goaltenders from the NCAA by over a two to one margin, but goaltenders from the WHL average nearly 70 games more than the next closest league.

Also note that the group with the lowest percentage of success (USHS) jumps up to No. 2 in games played average. This is due largely to the low total number of goaltenders reaching the NHL (10), one of which was Mike Richter who played in an impressive 666 career NHL games.

What do these numbers tell us? On the surface, it is apparent that the three leagues of the CHL outperform all other leagues in percentage making it to the NHL and in average number of games played in a career. A deeper look reveals that the QMJHL and the WHL can rival each other as the best at producing NHL caliber goaltenders. Let us take a closer look at these two leagues and break down the numbers by round to see if a preference can be had depending upon the round the selection occurs in.

(Total Reaching the NHL) / (Total Number Selected)

League1st2nd3rd4th5th6th7th8th9th
WHL4/47/86/73/75/73/62/31/22/7
QMJHL6/63/310/123/34/101/30/12/31/1

Average Games Played by Round

League1st2nd3rd4th5th6th7th8th9th
WHL467.0217.9451.7227.323.6207.7452.0108.0102.0
QMJHL279.0416.7247.547.740.8265.00.043.78.0

As you can see, when comparing each league by round, the WHL significantly outperforms the QMJHL at nearly every point. The lone exception would be the average games played by those chosen in the second round where the QMJHL dominates by nearly a two to one ratio. Besides that single category, the WHL either holds its own or defeats the QMJHL in both the number of draftees reaching the NHL and in average career length.

It should also be noted that the WHL leads in average games played in the crucial first three rounds where the vast majority of elite goaltenders are found. Further, from the fifth round on, the WHL maintains a high level of consistency in both number reaching the NHL and in games played average while the QMJHL significantly drops off. From that, it can be inferred that the pool of NHL caliber goaltending prospects produced each year by the WHL, on average, is deeper than that of the QMJHL.

In summation, based on an analysis of NHL games played by goaltenders selected in the entry draft, the three CHL teams outperform other developmental leagues in both percentage making it to the NHL and in average career length. Further, when comparing the WHL and the QMJHL, the WHL is far more successful in developing NHL caliber goaltenders than the QMJHL.

While considering goaltenders at the draft table, it might behoove the astute general manager to select one from WHL or the QMJHL at the top of the draft while leaning towards WHL goaltenders in later rounds. If the future performance of a goaltender cannot be predicted with 100 percent certainty, a selecting GM can at least acknowledge the odds and try to keep them in his favor as much as possible.

One side of the argument was succinctly put forth by Director of Amateur Scouting for the Los Angeles Kings Al Murray. “We have absolutely no bias in that area,” he said. “People might think that we’ve taken more Europeans or more college guys, but we’ve certainly had enough major junior guys at the tops of our lists on a given year. We just want the best player. We don’t really care where he plays.”

And this view is backed up by Dave Taylor. “That is part of the overall process. We try and rate them on their ability when we interview them and when we have a chance to speak with them. Our area scouts will speak with their coaches and teachers and try and gather as much information as we can. We don’t move a player up or down because he is from one county on another.”

This is the common claim throughout the NHL — that because you cannot predict the future performance of a player based solely on the league he is developing in, the area scouts need to view the individual player and ensure that he has all the tools that an organization looks for when labeling him “the best player available.” This view may not be wholly inconsistent with the argument that you should at least factor in the league the player is developing in. No one would argue that a scout need only look at the league the player is developing in and nothing more to determine the future success of a player. The argument is that knowing what league the player developed in will help predict future success a little bit more accurately. It should not be the only consideration involved or even the most important. It should only be one of the many considerations involved for which the weight of the consideration is left to the general manager to decide.

There is some belief amongst NHL area scouts that you can look to the league the player is developing in. Not necessarily because the particular league automatically produces the most NHL ready prospects, but because a particular league might install the intangible qualities in an individual player better than another league.

One area scout covering the OHL has to say, “We take great interest in the league in which a prospect is developing, as well as the coach he plays for. The players from different areas show signs throughout their careers of the roots they developed playing junior, college, tier 2 or in a European atmosphere. You are naturally all about your influences and surroundings. We see different attitudes as well as skills and styles within the three major junior leagues, let alone from the Euros. Face it, a kid from the west, born and raised on a farm is bound to have different character and knowledge of responsibilities than a rich Toronto boy.”

He goes on to state that particular differences in the leagues, such as toughness, can often lead to a different type of player developing. “Players in the west commonly are more aggressive and seem to have a certain amount of desire than we see overall from the OHL and/or Quebec, though each certainly has its players that would be an easy fit in the Western League.”

If nothing else, certain tendencies between leagues can be witnessed. These can be differences in players that are viewed as “intangibles” and are not easily quantified. But if you look to the historical performance (the statistical data of past performances), one might be more able to better grasp these unquantifiable intangibles and be more able to accurately predict future success than others.

Another area scout covering the New England area admits to taking the particular league and even team into account when assessing talent, but balks at the idea that it can be broken down by position, claiming the belief that the QMJHL produces elite goaltenders is just a myth.

“Yes, we do look at the league the particular player is playing in. Not so much by position, but the league in general. We also look at what team the player is playing for, and how that could potentially influence a player’s development down the road. I don’t believe you can break it down by position as to which league is better for a certain type of player. But, I do think that certain leagues, or for that matter certain teams in certain leagues can throw up red flags.”

Al Murray brings the debate to full circle. “We feel it goes in cycles from year to year. This year Ontario is particularly strong while last year was a great year for Quebec. This year colleges are down. Next year might be a big one. The US U-18 team has their best team ever and is going to have more players taken high off that team than ever before. It just goes in cycles. There’s no way to predict it beforehand except to watch the players as under-agers and know who’s coming up where.”

While it might go in cycles, the question is whether a particular league is better at being at the top of the cycle than other leagues. In other words, does one league complete the cycle more often? To understand this, we would have to look at the historical data. And as the data has played out above, the WHL has historically produced the better NHL goaltender more often and in later rounds. If it goes in cycles, the WHL has a shorter cycle than any other leagues and it might be of benefit to a general manager to at least understand this, if not take it into consideration.

If nothing else, the tendencies as reflected in objective statistical data, might be worth factoring in to some degree when deciding which player to draft. By no means should it be the only factor.

Guy Flaming and John Logue contributed to this article. Copyright 2005 Hockey’s Future. Do not reprint or otherwise duplicate without permission of the editorial staff.