Canucks’ 2016 NHL Entry Draft: Deep Dive

Draft Review - 2016

With the summer winding down, we at Canucks Army are looking to fill the dead space between our Top 20 countdown and the start of the World Cup of Hockey. This week, I’ve decided to head back to the draft, and I’m going to be looking at the past, present and future on Canucks drafting and NHL prospects by going over the 2015, 2016, and 2017 NHL Entry Drafts and how they pertain to the Canucks.

Following yesterday’s rehash of the 2015 Canucks’ draft class, today is a deep dive into their 2016 haul. At this point, it looks considerably weaker than the 2015 group, but then again, the 2015 class looks a lot better now that it did on draft day.

Disclaimer: When it comes to pGPS numbers, you may notice that some have changed over time. That’s because I’ve been doing ongoing research and development of the system throughout the entire summer, and when I push a new update, individuals’ percentages often change (I spoke of this recently in my article on Olli Juolevi). This is particularly true with prospects in leagues with wider age ranges (college, and professional leagues even more so). This is designed to make the numbers more representative of the players by tightening the similarity of the factors involved. At this point, I think I’ve found a formula that I’m planning on sticking with for the foreseeable future, so pGPS scores should be fairly stable for a while.

Olli Juolevi – D – 1st Round, 5th Overall

Ranked Canucks Army’s 2nd Best Canucks Prospect

Olli Juolevi is the highest Canucks draft pick since the Sedins in 1999, so naturally there’s going to be a high level of expectation on the Finnish defenceman.

Juolevi is a puck moving defenceman with a boatload of hockey smarts and a knack for offensive creativity. He’ll be a power play quarterback one day, having already excelled in that role for both the WJC champion Finnish squad and the Memorial Cup Champion London Knights.

2016 OHL London Knights 57 9 42 2.8% 13.2% 19 41% 32 15.9

Juolevi’s pGPS is among the highest of all Canucks prospects (should be no surprise), though a 40 percent chance of graduation to the NHL is less than what we’d expect from a fifth overall pick. It was brought up in the comments of my Juolevi prospect profile that a projection this low for a player like Juolevi – who is nearly a sure thing to play 500 games in the NHL, let alone 200 – could indicate an error in the pGPS methodology. Rather, it is important to remember that pGPS deals only with numbers, and this is what previous OHL players with Juolevi’s age, size and statistics have accomplished.

In my response in the comments however, I took another approach. Since we know that scouting rankings combined with statistical analysis can provide more accurate projections that either of the two methods individually, we can combine them in a simple manner that would take into account scouting and perceived pedigree. Using draft position as a proxy for pedigree, we can compare Juolevi only against players drafted in the first round, or, to take it a step further, only against players drafted in the top ten.

Compared Against n s pGPS % pGPS p82 pGPS R Average Line Assignment
All OHL Database 32 13 41% 32 15.9 2nd Pair
First Round Only 10 7 70% 32 27.7 2nd Pair
Top 10 Only 5 5 100% 35 42.6 1st Pair

These projection numbers are probably more accurate, as the chances of Juolevi busting are extremely remote. This type of method is useful to provide context for high level players like Juolevi, and may weed out comparables that have obvious flaws noted by scouts, but that statistics cannot pick up on.

While the consensus has been that there weren’t likely any bona fide number one defencemen available at the 2016 draft, I still believe that Juolevi’s ceiling is as a number one – and that a projection as a top pairing, number two defenceman is just a safer bet.

William Lockwood – RW – 3rd, 64th Overall

Ranked Canucks Army’s 24th Best Canucks Prospect

Will Lockwood was a bit of a shocker pick from the Canucks, given that there were a bevy of higher rated players left on the board when the Canucks were picking at 64th overall. Given that most draft services had him ranked in the low hundreds, taking Lockwood at that spot could certainly constitute a reach.

The reasoning that the Canucks have given for taking Lockwood where they did is that they’d identified certain players that they were targeting, and they felt that Lockwood wouldn’t be available when the Canucks picked next at 140th. There’s some semblance of logic there, provided that Lockwood turns out to be as good as they hope he is, though it also comes as a stark reminder of how few picks the Canucks had in the top half of the draft, and for that they have themselves to blame.

2015-16 USDP U.S. National

U18 Team

59 13 33 5.2% 13.3% 12 6% 23 1.6
2015-16 USHL USNTDP Juniors 20 3 6 2.2% 4.3% 7 0% 0 0.0

Lockwood’s pGPS score leaves plenty to be desired, especially for a 64th overall pick. His USHL cohort yielded no matches at while, while his time with the U.S. National Team Development Program as a whole (his USHL season is actually a portion of his USDP season) was a little brighter and managed a single successful match.

As I mentioned in yesterday’s post, USHL players are disadvantaged to a certain extent, as the league’s rate of players produced has been steadily increasing over the past couple of decades. As a result, although players now have as good a shot as, or better than, the QMJHL of making the NHL, that wasn’t the case too long ago, and players that produced at the same rate (adjusted for era) in 1990 and 2015 are going to have very different likelihoods of success. One thing that we can do to combat that is to adjust pGPS for league – that is, compare a player against similar leagues using equivalencies. Here’s what Lockwood’s pGPS scores would have locked like if he had put up the same numbers (adjusted for league) in North America’s various junior leagues:

USHL 0.7% 2.3% 0.6%
USDP 3.0% 3.0% 3.2%

Here, Lockwood’s stats in each the USHL and USDP were converted to the other North American junior leagues. As you can see, there’s still plenty left to be desired – although they are at least better than flat zeros. This is still pretty intuitive: Lockwood produced points at rates of 0.30 per game in the USHL and 0.56 per game with the USDP team. Converted to other junior leagues (which is a pretty close conversion), those numbers won’t get you very far. Lockwood is committed to the University of Michigan next season. A year in the NCAA should provide some more distinct and reliable information.

Lockwood will long be judged with the 64th overall pick hanging over him, but in reality, where a player is picked should only be used to judge the staff who picked him. As for the player himself, I don’t believe Lockwood is dead in the water just yet – he has the potential to carve out a career as a useful bottom six forward in the NHL. As for the Canucks however, it’s likely that they lost value by leaving prospects with much higher upside on the table.

Cole Candella – D – 5th Round, 140th Overall

Ranked Canucks Army’s 12th Best Canucks Prospect

Cole Candella is an intriguing prospect, and probably has better value for his draft position than any 2016 Canucks’ pick outside of Juolevi. He’s an offensive oriented left-shot, puck moving defenceman, ticking a lot of boxes for the Canucks. A wrist injury kept him out of the lineup for a good chunk of the season and likely affected his draft stock. The Canucks were lucky enough to scout him before the injury occurred (rather than only during his recovery), which may have given them a different glimpse than other teams got. Hopefully this turns out to be an advantage.

Candella is “a strong puck-moving defenceman with a very good sense of the game in front of him, anticipating the play and reacting appropriately”, according to the scouting report by Future Considerations, which ranked Candella 99th. The report further praises his ability to gain the offensive zone, his shot, and his physicality. listed some of his issues on the defensive side of the puck, including a tendency to abandon his partner in pursuit of offensive opportunities, and that he occasionally stops moving his feet in one-on-one coverage in his own zone, which can make him a little too easy to beat.

Given his increase in production from 2014-15 to 2015-16, he probably deserved to go earlier than he did, so the Canucks were fairly lucky to get him at 140th.

2015-16 OHL Hamilton Bulldogs  37 4 20 2.0% 10.2% 14 14% 27 4.6

Candella’s 14 percent pGPS score puts him a decent ways above the expected value at his draft position (approximately 10 percent), making him one of the Canucks’ best value selections in 2016 through the eyes of pGPS.

Candella’s best comparable was easily P.K. Subban (though at 90.6 percent similarity, Subban’s draft season barely made the 90 percent threshold for a match). Dennis Wideman and T.J. Brodie were also on the list, though the majority of successful matches were second pairing defenders.

Candella’s is one of the prospects that I am most intrigued to follow this season. He was generally considered to be better than the draft position he eventually found himself in, so it will be interesting to see what sort of potential he can demonstrate with his injury long behind him.

Jakob Stukel – LW – 6th Round, 154th Overall

Ranked Canucks Army’s 21st Best Canucks Prospect

While a lot of Canucks fans were hollering for Jim Benning to grab Ty Ronning at the draft (I know I was), and were disappointed when that didn’t come to fruition, the Canucks didn’t ignore their backyard entirely. In the sixth round, they grabbed local boy Jakob Stukel of the Calgary Hitmen – and formerly of the Vancouver Giants.

Stukel is from Surrey and played his minor hockey for Cloverdale Minor (which is where I played for 15 years, and for some reason this connection makes me feel good about myself). He’s another overage pick, having turned 19 back in March, though a look at his history gives a reasonable explanation.

Stukel missed the entirety of the 2013-14 season with a knee injury, and played only 49 games in his first draft year, putting up a lackluster 16 points. The injury likely held him back and delayed his development. A couple of years later however, it looks like he’s turned it around.

What is particularly noticeable about Stukel in the video above is his separation speed. Stukel has the ability to burn defenders who are caught flatfooted and has the hands to finish breakaways.

After a slow start with the Giants this season, Stukel was dealt to the Calgary Hitmen, where he production skyrocketed – he produced at nearly a point per game during his time there, with 34 goals and 56 points in 57 contests.

2015-16 WHL Vancouver/Calgary 69 36 60 14.8% 24.8% 25 7% 39 3.5

Among Stukel’s successful matches were Brendan Morrow, Kris Versteeg, and Tomas Fleischmann. Linden Vey also made an appearance on the list. His successful comparables averaged nearly half a point per season in the NHL, mostly scoring at second and third line paces.

Stukel’s pGPS of 7 percent is a bit low, even for 154th overall, and there were definitely still higher percentage bets on the board at that time. However, the fact that he was producing at basically a point per game following the trade to Calgary could indicate that Stukel has more to give. He’ll head back to Calgary this season (he’s already dressed for pre-season games with the Hitmen) where he’ll look to continue that point per game pace, or, ideally, improve upon it.

Rodrigo Abols – C – 7th Round, 184th Overall

Rodrigo Abols is a name familiar to attentive Canuck fans, as he was invited to the 2015 Young Stars Tournament in Penticton. Abols performed very well at the showcase tournament and charmed a good chunk of the Vancouver faithful, who were still swooning from the Ronalds Kenins experience in 2014-15 and were completely on board with bringing a Latvian like Abols into the fold.

What a difference a year makes: Ronalds Kenins, once known affectionately in these parts as the Latvian Locomotive, failed to make the Canucks out of camp last season and never recovered – he’s now headed back to the Swiss National League. Abols meanwhile, had an entirely mediocre season in Portland, his first campaign this side of the Atlantic.

2015-16 WHL Portland Winterhawks 62 20 49 8.8% 21.5% 18 10% 25 2.9

Abols managed to wrangle a pGPS percentage of nearly 10 percent, which is a little over the expected value at his draft position, despite scoring well under a point per game in his draft-plus-two season. His matches are mostly bottom six and replacement level players, with Dave Scatchard probably being the best among them. Paul Gaustad and Lance Bouma also qualify as matches.

Last season, as fans were clamoring for a contract for Abols, I dug into the NHL’s Collective Bargaining Agreement to explain why that wasn’t a possibility – basically, it was because he was draft eligible and not eligible for free agency at his age due to the fact that he hadn’t played a season in North America to that point. Well, he is now – or he would be, had the Canucks left him on the board and selected a younger player instead. Abols, a 20-year old coming off an unimpressive season, only needed to sneak past 17 more teams to become a free agent, allowing the Canucks to sign him, in addition to acquiring an additional asset. For that reason alone, I am not a fan of the Abols selection.

As for Abols himself, he’s old enough to head to Utica this season, provided that they sign him to a contract (either an Entry Level NHL deal, or an AHL contract, which they can have Utica do while maintaining his NHL rights). Whether or not he goes there is another matter, given that Utica has stocked itself with free agent forwards, and the Canucks may feel that Abols could use another year in the WHL.

Brett McKenzie – C/LW – 7th Round, 194th Overall

Ranked Canucks Army’s 25th Best Canucks Prospect

With their last pick of the 2016 NHL Entry draft, the Canucks took yet another over-aged player, totaling three in this class alone. That’s generally something that history would suggest to avoid, but in Brett McKenzie’s case, it looks like the Canucks made a wise decision.

The Vars, Ontario native plays both centre and the wing, he’s solid in the faceoff dot, and he’s strong in all three zones, with hockey IQ being one of his strengths. Hockey Prospect’s 2016 Black Book had this to say about his abilities:

McKenzie is a big-bodied centre who plays hard at both ends of the ice. He has a powerful shot and good positioning in the offensive zone. He is a physical player along the wall and does a good job winning his share of battles. Defensively he competes, works hard to get into passing lanes and blocks shots on a consistent basis. This has allowed him to play in all game situations. His skating is a little awkward and he would benefit from cleaning it up.

McKenzie increased his goal total from 11 to 26 between his first draft season and his draft-plus-one year. Like Carl Neill last year, he could be a player that turns himself into a prospect but making up for a poor draft year and producing like he should for his age.

2015-16 OHL North Bay Battalion 66 26 53 10.8% 22.1% 21 15% 28 5.0

A pGPS score of almost 15 percent is impressive enough for a later round pick, but for a 194th selection, it’s excellent. At about 10 percent above expected pGPS, McKenzie was also the best value selection by the Canucks at the 2016 draft.

McKenzie will return to North Bay for his fourth year of OHL hockey and look to improve upon a solid 2015-16 campaign.


At first blush, 2016’s haul looks considerably less impressive than what the Canucks walked away with in 2015. Granted, our opinion of the 2015 class has increased tremendously over the past 14 months. It’s not entirely off the table that the same thing couldn’t occur with the 2016 class, and we could be singing its praises by next May.

That seems a little bit unlikely at this point, for a couple of reasons. One is that the top pick, Olli Juolevi, is going to have a huge amount of pressure on him. By all accounts, Juolevi may have been the best available defenceman, and selected him at fifth overall is not so egregious that is can constitute a “reach”. However, it’s very unlikely that it could ever be considered a “steal” at that point, and certainly not within the next calendar year, in the manner than grabbing Brock Boeser at 23rd overall can now be considered a steal. In other words, we’ll be happy is Juolevi lives up to his draft position, whereas Boeser has already surpassed our expectations for a draft-plus-one year. In that view, Boeser (and by extension, the 2015 draft class) has a clear advantage.

For another reason, there are less players – six as opposed to seven – inherently limiting the opportunities for surprises. Furthermore, half of them were overage selection, and one (Abols) was even in his third year of eligibility and just a short time away from being a free agent. Overage selection typically have less upside, and have a tougher time impressing fans and analysts.

Finally, while there were question marks about both Brisebois and Lockwood as early third rounders, Brisebois was at least picked where the industry consensus had him ranked, whereas as Lockwood was picked a full round earlier than his consensus ranking average. That means that we have to assume that the Canucks know something that a lot of the scouting services don’t (not out of the realm of possibility, but not exactly a given).

The following graph charts the Canucks’ draft efficiency according to the pGPS metric, (with Expected pGPS for each draft position derived from the pGPS of all players from the 2016 draft class):

VAN 2016 Draft Efficiency

The positions on the graph are merely a visualization of what we discussed in the individual sections. Juolevi in particular is being undersold, for reasons already mentioned. That latter section of their draft picks show the Canucks breaking even if not gaining value on their draft picks, with Brett McKenzie being the best value for selection.

If the Canucks are going to make this draft class look better than it does today, it’s probably going to rely heavily on Will Lockwood doing in the NCAA this year what Adam Gaudette did last year; and that is produce much more than his USHL career would give him credit for. Other factors might be whether Juolevi can repeat some of last year’s success by winning World Junior gold and another Memorial Cup (neither is out of the question for the Finns or the Knights), and the lower selections having strong seasons in terms of production – especially for their age. For example, we’d like Jakob Stukel and Brett McKenzie to have seasons that look good for draft-plus-two players, not good for players that just got drafted.

If some of these things can happen, then our opinion of the 2016 class can increase drastically. Of course, it’s going to need a near miracle to contend with a 2015 class that looks like it could be the best Canucks draft class in a decade. However, since they’re all going into the same prospect pool, judging the drafts individually may be a bit fruitless. At least its clear that the depth in increasing.

  • Bob Long

    Its called a steal because it doesn’t happen very often. To expect that each draft is pretty unrealistic.

    Juolevi imo is a great pick – he already has the poise needed to be a 1/2 D by all accounts, something that for a lot of guys doesn’t show up until their mid-20s. We’ll find out at camp how he handles himself.

    The rest of them are attempts to fill the Gillis draft gap (or is it a chasm?) with some over-age long shots that seem to be very good character people.

    One thing I find really funny about the pGPS fancy stats graph is its literally upside down from how you would rank the potential of each guy from a classic hockey pov

    • In no way am I expecting a steal in each draft. What I merely was trying to get across was that it is easier for Boeser to look impressive, given where he was taken, than it will be for Juolevi, from whom people will automatically expect greatness.

      I’m a big fan of Juolevi. Don’t automatically assume I’m casting aspersions on him.

      re. the graph, that’s kind of the point, it’s a chart of potential relative to selection, with the caveat that it undersells Juolevi (something I’ve already gone over in a couple of articles). If it was just an article for projection, then yeah. Juolevi would be the highest, and even being “undersold”, he’d still have a value nearly triple the next best pick.

  • chinook

    What is not clear is whether the 2015 or 2016 draft class is stronger. From what I recall, the 2016 draft was not as highly regarded as the 2015 draft. As such, I think its a little bit of a stretch when you say that Lockwood was somewhat of a reach at 64. I think at some point past the first round, most teams just pick players that they have scouted and interviewed opposed to seeing who has the best numbers in whatever league. You don’t have to look any further than the 2 Canucks that were not drafted but invited into the prospects camps (I forgot their names) who were in top 10 in scoring in the Q.

    • I don’t at all think it’s a stretch to suggest that he was a reach. He was ranked far lower by every single ranking service, and each analyst they interviewed agreed that he was picked early.

      It’s not an insult, it’s just relative to where he was projected. He could still turn out to be a great pro. It happens.

  • chinook

    Absent from your discussion is comparison of the overall depth and strength of the 2015 and 2016 drafts. I’ve read (or inferred) 2016 did not have the depth of 2015, hence selection of over-agers in later rounds. For example, Toronto was roundly criticized for picking over-agers – more than Vancouver I recall.

    Question for you: how many over-agers were taken by other teams in 2016? Was it only Toronto and Vancouver? (I’m suggesting NHL scouts might be smarter than you give them credit)

    • It’s a very interesting notion, and it’s something that I’m working on for the future. A couple of comments:

      1) Yes, Toronto picked a lot of over-agers. I thought they had a pretty underwhelming draft overall, which brings me joy. No, I do not like Toronto just because they do “analytics things”. I hate them, just as I always have and always will, and I take joy when they make poor choices.

      2) Yes, lots of teams picked over-agers. 2016 had an inordinate amount of over-agers. Interesting trend. That said, the reason I criticize some of those choices is that where plenty of players in their first year of eligibility that were both ranked highly by scouts AND performed well vis-a-vis statistical models. Those players will have an article to themselves later this week.

      3) In the graph at the bottom, expected pGPS is determined by the projections of each player in THIS DRAFT CLASS, so Y-axis relatively on the graph is measured against 2016 class, not drafts in general. Whether it was weaker than 2015 (I agree that it was) is roughly neutralized.

      4) This article of course focuses only on Vancouver and judges them according to their selections, not what other teams have done. I have plans to go over each team’s draft haul, but that will take time. Have patience.

      • chinook

        Thanks Jeremy for the clarification, I think I get it. About your #3, it is the over-ager’s pGPS in their overage year that is plotted – correct? Have you discussed the change in pGPS from draft-eligible to draft+1 years? (maybe you have and I missed that refinement)

        Think I recall Calgary selecting overagers too, so I look forward to that future article.

  • TheRealPB

    Looking to next year, the Canucks have a first, two seconds, a third, and a fourth round pick still intact (if I understand the Philip Larson deal correctly). The second-round pick received from the Blue Jackets will be quite valuable if the BJs do poorly again this year.

    In other words, the Canucks have the opportunity for a banner drafting year next June. I’m hoping none of the picks will be traded, so that Benning can showcase his scouting acumen.

    • Ryan Biech

      They currently possess one second round pick

      Columbus has the choice of surrendering their 2017 or 2018 2nd round pick – which must be decided before June 1, 2017

      • Bob Long

        The Larsen pick is a 4th if he meets some kind of ‘performance measures’ that were not released, but its probably making the team, if not Edm gets a 5th.

        So for 2017 JB will likely have a 1st, 2nd, 3rd, and a 5th if Larsen makes it, 1st, 2nd, 3rd, 4th if not. So thats OK either way.

        If CBJ does well, they probably give us the 2017 2nd rounder, if not they’ll keep it since they’ll have a higher pick and 2017 is supposed to be a weaker draft than 2018 apparently so I hope they do keep it for our sake.

        So best case scenario from a draft pov is Larsen is beat out for a job, and CBJ does well and makes the playoffs and prefers to draft in 2018, so we’d then have 1st, 2-2nds, 3rd, 4th.

        • Bob Long

          I’d be happy if we traded Sbisa + their Torts pick back for Hartnell. I think that would complete the veteran depth needed for guys like Boeser to move into the system and learn how to play on a competitive team.

          • Ryan Biech

            You could probably do it with Sbisa alone given the Jacket’s cap situation, but Pedan and Larsson have to step up – and Groot has to show he is legit. I don’t think WD/JB want three inexperienced guys (and Biega) in the third pairing. JB has repeatedly said that he wants to have more d depth, so I actually see him potentially adding another D (Schlemko, Harrison, Gryba type) that can play in the AHL and be a reasonable call up.

          • Dirty30

            Don’t forget that there needs to be one D available for the Vegas draft and if you trade Sbisa you need to expose someone else — likely Biega.

            So that’s okay — but if you pick up an additional forward who can be exposed in the draft, who gets shuffled out as an alternative when you protect the guy you just acquired?

            Someone needs to get picked by Vegas from the Canucks, and my choices would be Sbisa or Dorsett.

            And I wonder if the reason Bartowski got played over Pedan had something to do with eligibility at the Vegas draft?

            Ugh, too early to be speculating!

          • chinook

            I agree, there could be management (manipulation?) of games played for those potentially subject to the expansion draft. I hadn’t thought of Pedan. I speculated in an earlier post about Gaunce. Either one needs to have a break-out season for it to become an issue. Might Vancouver have another Ben Hutton this year?

            As for trading Sbisa plus a 2nd for Hartnell…. c’mon! Its Columbus should include a 2nd to make the deal. And even then I wouldn’t do it, Hartnell’s contract is that bad.

          • Bob Long

            I’d protect: Sedin, Sedin, Eriksson, Sutter, Sven B, Horvat and then Hartnell because he has a NMC. That exposes one of Hansen, Rodin or Granlund, maybe Dorsett as top 4 forwards to get picked. It would suck to lose Hansen to Vegas, but not devastating.

            On D, I’d protect Tanev, Guddy, and Edler.

            We don’t need to have Sbisa on the roster to leave a D exposed, theres still Biega, McEneny, Larsen and Pedan for that.

          • The problem with Hartnell is his NMC. He has to be protected in the expansion draft. He can’t be bought out if need be, or sent down to the AHL. Although he may be a good fit on our 2nd line, he’s also slow footed, 34 years old, and wear and tear due to his style of play. His cap hit is high $4.75M. That’s a lot going against getting him, but again, he would be ok on our 2nd.

            Also, Canucks get Columbus 2nd round pick in 2017 or 2018. It’s up to Columbus to decide which one. My guess, 2018.

  • Chungus

    After getting screwed in the draft lottery, I really have no interest in the draft apart from Juolevi… who was most peoples backup pick.

    Unless one of these prospects turn into a number one C there is really not a whole lot there to be excited about (except Juolevi obviously).

    • TheRealPB

      They may not be exciting, but there is some good value in getting players like McKenzie and Candella as late as the Canucks did. Honestly, after the second round getting anything above a replacement level player is a win in my books.

  • chinook

    From reviewing many previous draft years from 10 or 20 years later the best forwards are picked in the first round, usually the top half of the first round. Very few “misses”. But with defenders its much more hit-or-miss. Its common for top defenders to fall into the second (or lower) rounds. Being undersized is only part of the reason, lots with good size too. Seems to be the nature of the position, that there is higher risk picking a D early in the first round. I’m not critical of taking Juolevi, a team still needs to balance organizational need and trust their own assessment of talent, but as a rule-of-thumb – take forwards in the first round and defenders in the second.


    “With their last pick of the 2016 NHL Entry draft, the Canucks took yet another over-aged player, totaling three in this class alone. That’s generally something that history would suggest to avoid…”

    I’m curious about this statement. Are you referring to the history of conventional scouting/eye test methods? As in traditional draft methodology tells us to avoid over-ages players?

    Or are you referring to actual statistics for the “hit rate” on these types of picks (versus draft position)?

    I’m just wondering how this relates to the recent work, such as below, regarding market inefficiencies and the undervaluing of over-aged players in the draft:

    Do the models used at Canucks Army draw different conclusions regarding this subset of players?

    The Toronto Maple Leafs have drawn much praise for using this very methodology to select over-aged players with 5 of their 11 picks. Analytics wunderkind Kyle Dubas is being called the Billy Beane of hockey for “gaming the system” by betting on over-aged players outperforming their draft position:

    The Leafs took 5 overagers with 11 picks. The Canucks took 3 out of 6.

    And based on pGPS%, it would appear that Vancouver’s three over-aged choices have the best statistical probability to outperform the average returns for their draft position (of the six 2016 Canucks selections).

    So what is it that history tells us about avoiding this?

    It would seem that over-aged players, as a whole, have had quite favourable returns versus their draft position (especially into the middle and later rounds). And the pGPS numbers (based on a historical model) reflect quite favourably on the individual selections.

    Am I missing something?

    • Mostly I was referring to the Canucks own record with drafting players in their second or third years of eligibility, which is brutal (plenty of thanks to the Mike Gillis years there).

      Overall, over-agers tend to succeed at a much lower rate than players in their first year of eligibility.

      I have that DEV article before, and I don’t necessarily agree with it. The Last Word on Sports article that praises the Leafs’ draft is then just using the DEV article as “proof” that the Leafs did smart things. There is a lot of dissent in the hockey community that that is the case. The Leafs draft has typically been met with confusion and malcontent, both from traditional hockey minds and stats pushers. How it ends up will require time. It could work out well. But calling Dubas, Hunter, and crew the Billy Beane of hockey for “gaming the system” at this point is patently ridiculous. It’s not like they’re the first team to ever draft over-agers.

      In spite of all this, there are still some redeeming qualities to over-agers, and there were some that went undrafted that absolutely should have been picked. Brayden Burke is probably the best example.

  • TheRealPB

    I still can’t believe they picked Lockwood just one spot ahead of Abramov! The other picks are fine and the Juolevi pick alone makes things seem pretty good. I don’t really understand what they see in Lockwood but as you say a year and actual on-ice performance can make a huge difference in our assessments (see Gaudette or Cassels)