71932026PM_129_Bruins_Canadiens.JPG

Many fans view a top-10 pick in the annual NHL Entry Draft as a sure thing.  The players selected in the first 1- picks are the best prospects in the world, the players seen the most by NHL teams and coveted for years in advance.  They are the players that teams plan to build their franchise around for years to come.

The reality, however, is a little different.  I went back and looked at all of the top-10 selections between 1993 and 2003, and discovered something a little surprising – less than two-thirds of the players selected (64.3%) have played or are projected to play 500 games at the NHL level (and some of those players, like Boyd Devereaux, will have long careers but only in a bit role).  Busts aren’t terribly shocking, they’re a fact of life, even with a top-10 selection.

It’s hard for me not to look at that number and wonder if NHL scouting and drafting could use some improvement.

In any case, I also ran a few other numbers and came up with success rates by position and by league.  Here are the success rates for players taken in the top-10 picks by position, with “success” defined as 500GP (actual or projected) at the NHL level):

Position # of Picks <100 GP 100GP 500+ GP Success Rate
Forward 61 8 12 41 67.2%
Defence 29 2 6 21 72.4%
Goalie 9 3 3 3 33.3%

That goaltending number shouldn’t be unexpected – NHL teams just aren’t very good at drafting goaltenders.  That’s obvious simply by looking around the league: players like Ryan Miller, Mikka Kiprusoff, Tomas Vokoun, and countless others were late picks.  Brian Finley, pictured above, played all of four games in the NHL, and the three goalies selected in the first round (Finley, Maxime Ouellet, and Ari Ahonen) have less than 20 NHL games between them.  Meanwhile, players like Miller, Craig Anderson and Alex Auld had to wait until after the first round for their names to be called.

Goalies are riskier picks too.  They’re more prone to injury, as we can see from players in the sample above; Dan Blackburn was pushed out of hockey altogether, injuries stunted Brent Krahn’s development, and even guys with obvious NHL talent like Rick DiPietro and Kari Lehtonen have failed to turn into franchise goalies because their teams can never be sure they’ll be healthy enough to play.  On top of that, goalies take longer to develop, and while there are 13-14 roster spots for a forward and seven or eight for a defenceman, a goalie generally needs to squeak in to one of two positions.  Given the abundance of not just good starters but star goaltenders late in the draft and via free agency, I don’t understand why teams continually wager high picks on goaltenders.

I also thought it might be interesting to see how successful scouts have been with Major Junior picks.  With that in mind, here’s a comparison between all the players taken in Canadian Major Junior versus all the players taken from everywhere else:

League # of Picks <100 GP 100GP 500+GP Success Rate
Major Jr. 66 11 16 39 59.1%
Everywhere Else 34 2 7 25 73.5%

The oddest thing about this chart is that it’s precisely the opposite of what any rational observer would expect, based on how NHL teams deploy their scouts.  Canadian Major Junior is the most heavily scouted league in the world and the primary source of players for the NHL.  Despite the fact that NHL teams allocate more scouts there than anywhere else, and despite the fact that the CHL is widely regarded as the best route to the NHL, the success rate for those players is less than it is for players from other leagues.

If we were to disregard the QMJHL (which sent only six of the 66 players selected) the numbers get even uglier, with the percentage rate for the OHL and WHL combined dropping to 56.7%.  Given that more players were taken from those two leagues than all the others combined, that’s a shocking number.

The question now is ‘why?’, but I’ll admit I don’t have an answer; just a belief that the system needs to be improved.

Comments (18)

  1. As long as the NHL continues to draft 18 year olds, there’s going to be an issue with projecting them.

    I wonder how it compares to NHL and NBA drafting (where they take guys 2-3 years older), as well as MLB (who draft all over the place).

  2. Pretty weak post by a good blogger. Where is the comparison to later picks? I think fans define a top 10 pick as sure thing compared to picking lower than that.

    So, if you disregard the goalies the success rate is 70%? And I would guess that injuries cause at least 5% of top 10 picks to not succeed. I would think that when you are drafting kids for a job that lasts 10-15 years or more, a 75% success rate is really quite good.

    And with the other (better) chart, my guess is that risk of the player not adapting to North American life, cost of scouting and drafting what you know (most, is not all GMs, are products of the North American development system), are the some of the reasons for this.

  3. How about the quality of the players though? CHL had a 59.1% success rate, but the players in that percentile could be scoring upwards to 90 points a year whereas the players selected everywhere else have a better chance of playing 500+ games but there numbers could be less than stellar. Not sure if this is the case but would be cool to know. If it were the case id def take my chances drafting from the CHL if the reward could be that great

  4. Erik:

    Don’t worry, I’m going to go more indepth here. Draft studies are a fairly time-consuming hobby for me, so I have to do them in smaller bits or I’d be doing one post a week ;)

  5. Shawn:

    Just thinking back to the sample, I’m pretty sure that isn’t the case. Plenty of impact players out of NCAA and Europe, and plenty of non-impact guys (Boyd Devereaux and the like) out of the CHL. I think games played does a fairly accurate job of reflecting that, although that might be something worth looking at later on here.

  6. Hawerchuk:

    That’s a good point. A really good point, actually, and something I should have checked. The counterpoint would probably be that an NHL game from a 24 year old is worth more than an NHL game from a 20 year old (and I think that’s probably the case) and that the late entry of college/European players is actually a blessing in disguise, but that’s assuming a lot, particularly without numbers to back it up (and I’m not sure it could be proven, given that bigger talents get moved to the NHL sooner).

  7. I dunno, Jon. I’m not sure you found something here. If you look at those drafts and you restrict the dataset to games played through age 24 (the youngest age in the full sample) – here’s the percentage who played 164 games or more:

    ON/W: 38/49
    Other: 38/51

    Or 246 games or more:

    ON/W: 34/49
    Other: 28/51

    I think there’s some definite draft preference for immediate impact vs total career impact, which you won’t get from an NCAA player and at the time, might not have gotten from a European with potential military service and contract issues holding up his arrival.

  8. Or, more accurately, I think you found the NHL’s preference for immediate impact over long-term. Productive seasons at age 18 and 19 are worth a lot to a bad team.

    Btw, my numbers up there are wrong – I forgot to take out the goalies…

  9. I haven’t looked at the numbers like you have but I’m going to quibble with a couple of your points.

    1. 500 GP to consider a goalie a success seems a pretty high bar. NHL.com shows that only 55 goalies in NHL history have played more than 500 games (plus some current goalies that are close to that mark.) So with a more realistic bar for GP the goalie success rate would probably be 50% or better.

    2. I’m not sure I’d agree with your logic that the better scouted leagues should produce more top talent. The only time you’d risk a high pick on a poorly scouted league is when it is clear that the talent is at a superstar level. And so a greater percentage of those picks are going to have good pro-careers because they had to be so good to begin with to get picked that high.

  10. Mike – Thanks for the comment.

    1. The question there is ‘should the bar be lowered to reflect the average career expectation of a goalie’ or ‘should the bar stay where it is because franchise goalies are so rare’. I lean towards the latter but the former’s an intriguing point. Generally I think that goaltenders are heavily overvalued by NHL teams.

    2. Another fine point. I’d submit that NHL scouts should have a better chance of identifying a future franchise player if they see him more times. If Player X and Y are both future stars, and I see X five times and Y 50 times, I’ve got a better chance of identifying Y as one than I do X, I would imagine.

  11. I think it’s clear that Major Junior is heavily overscouted. The more a scout sees a player, the more they lose sight of his deficiencies. It’s the opposite of what is expected, but it seems to play out that way.

  12. JW: re: Goalies. Franchise goalies are rare but so are franchise skaters. 500 GP is much more achievable for a middle of the road skater than goalie. Don’t forget also that a GP for a goalie is ~60 minutes played, for a skater just ~15-20. Goalies share the load differently. A skater carrying exactly his share of the load may play 82 GP; a goalie just 41. So sample sized must be treated differently.

    Put it this way. Looking at the 20 seasons 1989-2010, 621 skaters have played 500+ games, just 23 goalies. That’s nearly 30:1. In reality 2 of 20 guys who dress every night are goalies (even if only 1 of them gets credit for a GP). Also in reality about 1 draft pick in 10 is a goalie (9 of 99?? in your sample). So set a threshold closer to a 10:1 ratio. I suggest 250 GP, of which there are 72 goalies that qualify in the same sample size as the 621 skaters. To me that’s a fairer representation.

  13. The one other thing I see that hasnt been mentioned is what if the pick/ player dosnt work out because of his/her attittude, like a Radulov or Filitaov I cant see either of them coming back, at the end of the day I cant see that being put on the scouts. On the goalie debate I wouldnt base it on 500 games I would say around 250 but then I would throw in actual win % and if they where starters for the team and if not then did they have one year where they played a ton ala Raycroft, or even a guy like Clemmensen where he’s only played a ton fo games like once when Marty when down… Just my thoughts tho.

  14. You have to remember that Major Junior Hockey is where scouts generally go to scout role players and grinders as well. When scouts go over to Europe, Russia, etc. they are only scouting quality high skill players for the most part. So when you are drafting such a large number of players out of the WHL, OHL, QMJHL, and to a lesser extent the BCHL then of course the success rate drops so the statistics of that are biased like all statistics.

    In Europe and Russia they’re looking at players like Filatov, Ovechkin, Jagr, the Sedins, Bure, Mogilny, etc. because international players have the rep of being very skilled with the puck but due to a different style of play they do not play the NHL’s blue collar dump and cycle style. Most European players and teams play a more wide open style due to the bigger ice and just because of the different culture and philosophy in international hockey than North American Hockey.

    I’m surprised that the OHL is considered the best trainiing grounds though, because I would think the WHL would be the best training ground for future NHL players because the WHL is known as a more defensively tough league and they play a style similar to the NHL’s dump and cycle. I would say the OHL is almost a hybrid of the WHL and QMJHL in that it’s a little more wide open offensively but not as run and gun as the QMJHL.

    Something should change though as players like Pahlsson, Ruutu (Tuomo and Jarkko), and Ovechkin have proven European players can be drafted to fill roles as physical power forwards or grinding/checking players. So the preconceived notion that Europeans can’t fill bottom 6 forward roles is definitely false so NHL scouts should be over there more to see how the checking/role players overseas stack up the to the North American Juniors.

  15. On the other hand, Andrew, scouts aren’t drafting role players in the top-10. Or at least they shouldn’t be.

  16. JW:

    Don’t you think that no matter where most picks are taken from that the success rate of that source would drop when the number of players selected from that pool increased?

    I don’t think it’s possible that tons of draft worthy talent is hiding outside the CHL and no one realizes it.

  17. I am not sure that using stats is a good way to point issues with scouting however i agree
    that scouting needs to be improved.I also believe that nhl has up there draft age to 19 years as that extra year in junior helps those players in development.I also think that when a top ten draft pick is being scouted too much is put on his goal scoring abilities. I think more should be
    put on his overall game.I also think Canada could develop a better international system for players such as having play for team canada for 2 years .this help develop those players. I
    think the nhl has a big enough farm system they should not need these players before they are 20 years of age.I see alot of players in the nhl that should not be playing at that level

  18. and management of to Tramadol have pain the can wykry

Leave a Reply

Your email address will not be published. Required fields are marked *