2013 ENTRY DRAFT POST 3: EDM, LAK AND WALKABOUT

In yesterday’s look at the Oilers and their walkabout selections 2008-12, one of the things that became evident was that ‘gaps’ inside the top 100 are critical in breaking down off the charts selections. In 2007, Los Angeles selected D Thomas Hickey 4th overall, and Hickey was ranked #13 on Bob McKenzie’s list–that’s a major walkabout. With that in mind, I’m adopting Rom’s ‘graduated’ system for the talent pool and placing special emphasis on meandering inside the top 50.

So, we start by resetting the Oiler walkabouts.

URSA MAJOR (the big reach)

  • 2009-Anton Lander selected #40 overall, ranked HM (BM list went to 60, then 25 Honorable mentions)
  • 2010-Martin Marincin selected #46, ranked #71
  • 2012-Mitchell Moroz selected #32, ranked #56

I spoke with speeds yesterday and he made a strong point in regard to Moroz (which may also apply to Lander). When you’re selecting in the top 50 (in almost all draft years) there are still a plethora of prospects with a ‘wide range of skills’ available; so when Moroz is taken at #32, it’s a double whammy in that you’re reaching AND you are also leaving many assets with more universal appeal on the table. Mitchell Moroz may turn into an NHL player, but the range of spots in the order he can be slotted are not wide.

Lander is a similar case in that he’s a 3 or 4C as an NHL player (I don’t think anyone can argue he’ll ever score enough to supplant Gagner or similar on a skill line). For me, Marincin is a player with a nice range of skills so a less controversial selection.

Fair?

URSA MINOR (the smaller reach)

  • 2009-Troy Hesketh selected #71 overall, unranked
  • 2009-Cameron Abney selected #82 overall, unranked
  • 2009-Kyle Bigos selected #99 overall, unranked
  • 2010-Jeremie Blain selected #91, not ranked
  • 2012-Jujhar Khaira selected #63, not ranked

I’m more comfortable with splitting the group at 50, seems to me we can better see the reach picks that clearly happened when the talent pool was still enormous. This group is the secondary walkabout group, and the 2009 group is clearly one in which the Oilers scouting staff felt the BM list (or the draft overall) was not strong. I think Khaira doesn’t belong here, but this is where he lands.

The group of four above Khaira are the real story here. Edmonton took a shot for some reason (area scout, saw him good, etc) and it did not work out. I feel it’s important to state that a player like Bigos–taken at #99–is miles from the Moroz selection. The farther down each draft you get, the wider the range of draft lists. Bigos and Blain make the list, but should not be viewed as mortal sins.

schenn common

Brayden Schenn, Saskatoon Blades

For our purposes today, I’ll be using Rom’s outer marker

#1-3 picks = variance of 3 spots
#4-8 picks = variance of 5 spots
#9-15 picks = variance of 7 spots
#16-25 picks = variance of 10
#26-35 = variance of 10
#36-45 = variance of 10
#46-55 = variance of 10
#56-70 = variance of 15
#71-85 = variance of 15
#86-100 = variance of 15

  • 2008-Drew Doughty selected #2 overall, ranked #2
  • 2008-Colten Teubert selected #13 overall, ranked #13
  • 2008-Vyacheslav Voynov selected #32 overall, ranked #44
  • 2008-Robert Czarnik selected #63 overall, ranked HM
  • 2008-Andrew Campbell selected #74 overall, unranked
  • 2008-Geordie Wudrick selected #88 overall, ranked HM
  • 2009-Brayden Schenn selected #5 overall, ranked #5
  • 2009-Kyle Clifford selected #35 overall, ranked #55
  • 2009-Nicolas Deslauriers selected #84 overall, unranked
  • 2009-JF Berube selected #95 overall, ranked HM
  • 2009-Linden Vey selected #96 overall, unranked
  • 2010-Derek Forbort selected #15 overall, ranked #11 overall
  • 2010-Tyler Toffoli selected #47 overall, ranked #27 overall
  • 2010-Jordan Weal selected #70 overall, ranked #48 overall
  • 2011-Christopher Gibson selected #49 overall, ranked #42 overall
  • 2011-Andy Andreoff, selected #80 overall, unranked
  • 2011-Nick Shore, selected #82 overall, ranked #54 overall
  • 2012-Tanner Pearson, selected #30 overall, ranked #29 overall

URSA MAJOR (the big reach)

  • 2008-Vyacheslav Voynov selected #32 overall, ranked #44
  • 2009-Kyle Clifford selected #35 overall, ranked #55

The Voynov reach is a small one, and no one on planet earth would ever complain about it–that’s a helluva second round pick. The Clifford selection is interesting to me in that it has some similarities to the Moroz selection. Clifford had a strong 18-year old season (year after draft) and was in the NHL at 19. Moroz did not have a big season and will return for another year of junior, but I do think this is a comparable and does suggest organizations do jump ahead at times for this player type–and do so in this range.  Mitchell Moroz’ career path may never resemble Clifford’s when he turns pro, but on draft day I think they were comparable.

URSA MINOR (the smaller reach)

  • 2008-Andrew Campbell selected #74 overall, unranked
  • 2009-Nicolas Deslauriers selected #84 overall, unranked
  • 2009-Linden Vey selected #96 overall, unranked
  • 2011-Andy Andreoff, selected #80 overall, unranked

Not much to say about this group, interesting that the 2009 draft is front and center with the Kings, too. Vey looks like a player.

RUNNING THE NUMBERS

rnh1

  • Oilers overall: 22 selections, 8 have been outside the perimeter. 36.36% of the selections are beyond the outer marker.
  • Oilers Ursa Major: 12 selections, 3 have been outside the perimeter. 25% of the selections are beyond the outer marker.
  • Oilers Ursa Minor: 10 selections, 5 have been outside the perimeter. 50% of the selections are beyond the outer marker.
  • Kings overall: 18 selections, 6 have been outside the perimeter. 33.3% of the selections are beyond the outer marker.
  • Kings Ursa Major: 9 selections, 2 have been outside the perimeter. 22.2% of the selections are beyond the outer marker.
  • Kings Ursa Minor: 9 selections, 4 have been outside the perimeter. 44.4% of the selections are beyond the outer marker.

What does this tell us? Well, I asked for suggestions yesterday from you, and the Kings were the overwhelming choice. I used Rom’s guidelines (which I believe to be fair) and we ran those numbers for the Oilers and the Kings. The EDM/LAK 5 year drafts look pretty close to me: both go off the board for need and both do it more as the draft rolls along–I would guess these are ‘industry standard’ moves.

The Kings took a big guy out of order (Clifford) and the Oilers took a big player out of order (Moroz). In the season leading up to his draft, Clifford’s point per game (.467 in OHL) was in the range we’d expect to find a 4th liner (should he make the NHL). Moroz had similar numbers leading to his draft (.379 point per game in WHL) and certainly projects as a Kyle Clifford should he make the NHL.

The Kings took a defenseman with a nice range of skills (slightly) out of order and the Oilers did the same with Martin Marincin. I don’t think Marincin will ever be Voynov, but it’s a defensible selection.

The Lander selection we’ve covered previously.

My conclusion is this: I don’t believe it is wise (as a rule) to draft big forwards or stay at homes who are 6.09 out of order. The McKenzie list is basically an “NHL draft for Dummy GMs” guide–it’s all laid out for you! So when selecting Kyle Clifford or Mitchell Moroz NHL teams are trading the actual value of their selection in exchange for someone who is extremely unlikely to cover the ‘draft value’ bet.

They are, in fact, drafting for need. I humbly submit to you the following: when a team makes a “Moroz” selection it no longer reflects only on the scouting staff, but rather on the management who charge the scouts with delivering product. The Moroz selection runs ‘out of time’ with risk averse and stands out from the other selections (Pitlick, Musil) in that area, and for that reason I’m going to lay some of the credit/blame at the feet of Steve Tambellini.

These player types are more plentiful (and thus more readily available) after the draft and should not be acquired with early second round selections. In our next segment, we’ll talk about Craig MacTavish and his ‘early words’ on the draft and what he would like to see from the Oilers draft this summer. I expect one of the things he’ll want is strong assets in the second round.

written by

The author didn‘t add any Information to his profile yet.
Related Posts

85 Responses to "2013 ENTRY DRAFT POST 3: EDM, LAK AND WALKABOUT"

  1. spoiler says:

    LT….

    What does your conclusion have to do with the work above it?

    The bulk of the post is on how many religious visits a team goes on and whether the Oilers are more frequent in their puberty rites than other teams… And it turns out they are nearly identical to one off the better (luckier?) drafting teams in the league.

    Which really only proves that there’s a wide range of scout’s opinions in Bob’s scouting poll. Stunning that such circular logic would show up. Stunning.

    How do we conclude from all this that it is “not wise” to draft big forwards or stay at home defensemen? That’s a different topic and has nothing to do with the evidence presented.

  2. Hammers says:

    The important thing is what McT does . My guess is best player with great character as they can be traded . I never understood why teams took chances in there first 3 picks maybe even 4 as that’s the top 120 players .

  3. Lowetide says:

    Hammers:
    The important thing is what McT does . My guess is best player with great character as they can be traded . I never understood why teams took chances in there first 3 picks maybe even 4 as that’s the top 120 players .

    Risk averse is the wisest way to run a draft, I don’t think there’s a lot of doubt. The Hickey selection is a recent lottery example, but even in the early to mid second round–especially in a strong year–the best plan is to bet on the guy who established himself during the draft year.

    MacT’s influence on the draft is going to be very interesting.

  4. Woodguy says:

    This is really cool stuff.

    I generally stay pretty quiet about who to draft as projecting kids can be a mugs game and it is usually years before you know you are right or wrong.

    One thing that does grab my interest is the results of “sliders”

    It was mentioned in a thread yesterday that MTL had drafted a few kids one year well past their “generally accepted draft position”

    In this post, it looks like LAK did the same in 2010 with:

    2010-Derek Forbort selected #15 overall, ranked #11 overall
    2010-Tyler Toffoli selected #47 overall, ranked #27 overall
    2010-Jordan Weal selected #70 overall, ranked #48 overall

    Forbort didn’t slide too much, but as with the other metrics thrown around, the higher up in the draft, the less positions to take into account for a “slide” or a “walk about”

    Forbort just finished College this year and got in 6 games in the AHL before the season ended.

    Tyffoli produced at .879pts/gm in the AHL this year and has shown well in his NHL debut

    Weal produced at .523pts/gm in the AHL this year and has yet to make his NHL debut

    Interesting stuff.

  5. Romulus Apotheosis says:

    Wow!

    Sticking to that video for a second, it has a lot of interesting info:

    first that deal they mention (trading a second rounder to get the 13th pick) is wrong… unless I misheard it. Apparently, it went like this:

    June 20, 2008 – Acquired a first-round choice in the 2008 NHL Entry Draft (12th overall selection; later traded to Buffalo on June 20, 2008) from Anaheim for two first-round choices in the 2008 NHL Entry Draft (17th and 28th overall).

    June 20, 2008 – Acquired a first-round choice in the 2008 NHL Entry Draft (Colten Teubert) and a third-round choice in the 2009 NHL Entry Draft (later traded to Calgary on June 27, 2009) from Buffalo for a first-round choice in the 2008 NHL Entry Draft (12th overall).

    http://kings.nhl.com/club/page.htm?id=40851

    that’s a pretty active series of moves. but it certainly looks like they had Teubert on the mind and targeted him.

    Interesting that Pierre mentions Erik Karlson as the other D option…. imagine!

    And… BM again compares him to Weber!

    He must have been really highly regarded at the time.

    That’s got to show how unpredictable these things are, especially with D.

  6. SoxandOil says:

    The thing with the Moroz pick (at least to me) was it seems Edmonton was praying that Samuelsson dropped to them. When those pesky Coyotes selected him (at 27) Edmonton just seemed to go to the next Oil King on their list. Samuelsson would have been nice, you have to feel that Mac-man of action-T would have traded up to grab him.

    A thing of note when comparing these two lists, when LA reached they got a NHL player. Also LA took the consensus BPA (they seemed to specialize in draft day droppers) regardless of position, size, or need. When you remove 1st round picks for that time frame (and I know its early to judge) but LA comes out ahead (maybe) despite Edmonton having better draft position.

    Also Raanta Watch day three…Supposedly he went to the lake for a camp of some sorts. Doubtful that a announcement will be made till monday. If it turns out to be Edmonton its interesting that there has no leaks or mentions by the MSM.

    Would Raanta fill the back up role immediately? Or would he need some OK seasoning? Does that mean that regardless of Raanta another goalie will have to be brought in?

  7. Lowetide says:

    Rom: Teubert WAS highly regarded. That’s why imo it is so important to save the draft profiles and return to them over and over again. Because players develop a lot after their draft day. The one thing we do know, though is that Teubert’s skills would have been available to see in his draft year and that there was no ‘offensive kicker’ coming for him.

    In that way, I think the Teubert pick–like the Plante pick before–tells us choosing a stay at home defenseman is a bad idea in the first round. The big Russian is such a player this season.

  8. steveb12344 says:

    It’s looking pretty good on the Raanta front.

    A few tweets from a Finnish dude close to the situation…

    Antti Eteläaho‏@finanatsu13m
    @AndrewPJensen That I think it’s either EDM or WPG, which Raanta choose. All indicates to that.

    Antti Eteläaho‏@finanatsu27m
    @WpgJetshocktalk Yea, but then there is that factor that he could easily be starting goalie (won that job) for EDM when next season ends…

    Antti Eteläaho‏@finanatsu23m
    @WpgJetshocktalk Yes, but I wanted to say, it could seem for Raanta that Pavelec is future’s goalie in WPG, when EDM’s no 1. is still open.

    Interesting.

  9. Lowetide says:

    I’m sitting here waiting for the Raanta story to break, but would guess we find out tomorrow. Based on the fact this has been a secret for basically 40 hours, I don’t know if my money is on the Oilers. :-)

  10. Romulus Apotheosis says:

    This “range of skills” point is probably the biggest take-away for me.

    Though BM’s list mitigates it to some point, having Teubert that high despite his known lack of offensive upside at the time seems to show the perverse love of size/strength bleeds into his list too.

    Staying within BM’s list and picking big players without much offensive upside is somewhat dangerous, leaving the list and picking these players is nuts.

    An interesting question would be whether BM’s list — which as we discussed yesterday, ultimately just reflects the current thinking of various scouts — varies from year to year as these kinds of players fall into and out of fashion… following on MacT’s point that success breeds tunnel vision as teams keep trying to recreate the last team that won.

  11. Romulus Apotheosis says:

    Lowetide:
    I’m sitting here waiting for the Raanta story to break, but would guess we find out tomorrow. Based on the fact this has been a secret for basically 40 hours, I don’t know if my money is on the Oilers.

    A new regime?

    Somehow both more public and more secretive?

  12. Romulus Apotheosis says:

    Lowetide: In that way, I think the Teubert pick–like the Plante pick before–tells us choosing a stay at home defenseman is a bad idea in the first round. The big Russian is such a player this season.

    Looks like Smid is another good example.

    Love the player and he clearly is a solid NHLer, but 9th overall is crazy in hindsight considering his complete lack of scoring.

  13. Lowetide says:

    Rom: This is imo the culture changing. When Bill James began writing about baseball all sorts of things were clearly established. For instance, minor league stats were regarded about equally, despite the fact that minor league ballparks came in all shapes and sizes. James took the time to meausre how many singles, doubles and triples and home runs and walks and strikeouts took place in each park, and then let the air out of the tires of each prospect accordingly.

    He famously published a minor league book that projected Jeff Bagwell to win the NL batting title, despite the fact he was in New BRitian (iirc) in Boston’s system and had been traded to Houston over the winter.

    Anyway, a writer got hold of that minor league book, looked at James major league projections and wrote an article about James calling for Bagwell to win the batting title as arookie.

    Bagwell did not win the batting title, but did hit .294 with power in that big Astrodome. James said later he was scared that if Bagwell had failed his reputation would have been ruined.

    But the attitudes toward minor league stats changed over time. And so the Dodgers prospects who were always expected to hit 500 home runs a year because they played in Albuquerque (band box) were now seen in a more realistic light.

    Stay at home defensemen, goalies and coke machines will one day be moved back to the mid second round or below (unless the goalie is an impact prospect) but we’re still getting there.

  14. steveb12344 says:

    Lowetide,

    Maybe he hasn’t even told the team that he’s picked yet. It sounds like nobody knows anything at this point.

    I have to say though, if it is between the Oil, and the Jets. I can’t really think of any reason to go to Winnipeg. Probably the only NHL city less desireable than Edm. to players. lol

  15. sliderule says:

    I think what bothers me about the oiler scouting staff is not whether they go walkabout after the first round is they make bad picks.
    If you compare the drafts of the kings and Oil from 2005 to 2009 inclusive you find the Kings drafted 15 players who played at least 10 games in NHL.
    Surprise surprise the oil drafted 14 players with the same number of minimum games.
    The games played tell a different story.

    Kings 2680 games played for 179 per player
    Oilers 1662 games played for 118 per player.

    It’s when you step out of first round that you find a real disconnect.

    Kings second round and later 1559 games 9players 173per player
    Oilers second round and later 648 games 8 players 81 per player

  16. Bar_Qu says:

    My takeaway from this is how consistent the Oilers have been with the LAK in drafting (broad strokes) and the failure of prospects is less on poor drafting than it is injuries and luck – which I’ve heard somewhere before.

    I tried to comment yesterday on how it was a most enjoyable draft thread to follow,but #stupidphone. So, congrats gentlemen on some fine thinking and parsing of information to help us elevate discussion beyond “oh yeah?”

  17. Rondo says:

    What about the ability of the team to slow cook the player. ex. Logan Couture

    Some teams just rush their players.

  18. jonrmcleod says:

    Lowetide,

    Have you talked with Corey Pronman about evaluating a draft? I know he believes you shouldn’t judge it after the fact.

  19. godot10 says:

    Romulus Apotheosis: Looks like Smid is another good example.

    Love the player and he clearly is a solid NHLer, but 9th overall is crazy in hindsight considering his complete lack of scoring.

    This is relative. Smid isn’t a bad pick if you go back and look at that draft.

    A long career top 4 (though on the defensive side) is not a bad result for a 9th overall pick, in a weak draft (except for the top 2).

    You shouldn’t let the fact that we had to watch the sausage being made impact the fact that the resulting sausage actually turned out okay.

  20. Lowetide says:

    jonrmcleod:
    Lowetide,

    Have you talked with Corey Pronman about evaluating a draft? I know he believes you shouldn’t judge it after the fact.

    I think everyone has an idea about how long to let things boil before calling a draft, for me it’s 5 years after. I don’t really think we can go back and credit Kevin Prendergast for Colin McDonald emerging as a regular in NYC 9 years after the draft.

    That doesn’t really make sense to me.

    I certainly agree there’s no sense in judging a draft right after it takes place. However, we’re not so much judging the drafts 2008-12 as talking strategy and best ways to gather talent in the draft.

  21. vishcosity says:

    Re slow cooker, maybe averaging each team’s average player delay could provide a corrector. So on average, a Det prospect gets pn average 41 GP* in A while Edm gets on average 27* AHL GP.

    But then, Edm drafts higher on average per round, and further the 1 OV’s mess up the math.

    Ricki was saying yesterday about expected results based on probability, and maybe that too needs to consider 1 – 100 rather than average percentage per round.

    Bar Qu: I was posting yesterday (and today) on an iphone seemingly fine. Very odd.

    *pure speculation. No Internet at Gma’s.

  22. zoofarmcheater says:

    Very cool! If you ever do another one of these, want to do one with San Jose? From what I can remember, their name comes up the most when people talk about teams that are great at drafting.

  23. TheOtherJohn says:

    Bar_Qu:
    My takeaway from this is how consistent the Oilers have been with the LAK in drafting (broad strokes) and the failure of prospects is less on poor drafting than it is injuries and luck – which I’ve heard somewhere before.

    I tried to comment yesterday on how it was a most enjoyable draft thread to follow,but #stupidphone. So, congrats gentlemen on some fine thinking and parsing of information to help us elevate discussion beyond “oh yeah?”

    So we are similar to LAK except for “injuries and luck”. WTF?

    They have a boatload of players from the 2nd round or later who have played significant games in the NHL and we have guys struggling mightily in the AHL. Isn’t that the ultimate measure? Finding players who can make & help on a NHL roster

    One thing is NOT like the other!

  24. rickithebear says:

    Yesterday past draft history showed our collection of 2nd to 7th picks should
    yelde4.25 Players. (Peak playing age 26-31YR)
    Cornet, hartikainen, Lander, rajala, Roy, pitlick, Marincin, Martindale, Davidson,
    musil, Simpson, Rieder, Gernat, Moroz, Khaira, McCarron.

    Our 6 first rounders should have yeild 4 players.
    Eberle, MP, Hall, RNH, Klefbom, Yakupov.

    Yes we are not LA.!

    “Yes daddy i know i get one cookie after breakfast.
    Two is awesome!
    But I want Three!

  25. jonrmcleod says:

    Lowetide,

    Maybe I misunderstood him, but I think Pronman says you should NEVER evaluate a draft based on what happens afterward. Instead, he says a draft should be evaluated based on the information available on the day of the draft.

    Regarding the best strategy, I think that any of us could be in charge of a team’s draft in the first 3 rounds by simply using the consensus rankings of players. (Have you ever looked at what players the Oilers would have selected if they had always picked the highest ranked player on BM’s list?)

    I’d also be interested in seeing which teams, if any, consistently find useful players in the lower rounds. I think that’s how scouts really earn their money.

  26. gd says:

    LA’s drafting to me looks as average to mediocre as the Oilers (and the majority of the NHL).

    -Picking Bernier at 11th shows the Kings weren’t ahead of the curve in realizing goalies shouldn’t be picked in the 1st round unless they are dominant.

    -The Hickey walkabout at 4 is one of the 10 worst picks of the last decade.

    -Going WHL three years in a row for 1st rd picks after the first two flopped looks questionable. Schenn is a okay pick, but OEL would have been a homerun.

    -The best thing they did was stay bad enough that they got a top 4 pick in 2008, which looks like the greatest year for D ever. Picking Voynov was obviously a great move and makes up for the huge miss on Tuebert.

    -I think their strength were the non-draft moves. Scuderi was the perfect UFA signing at the right time. The Carter/Richards trades were very shrewd deals. Quick’s emergence as a top 3 goalie definitely is crucial to Lombardi’s legacy. If either Roy, Bunz or Perhonen became an elite goalie, MacT’s would be well on his way to legendary status.

    It looks to me like trying to figure out what teams have a successful draft strategy is like trying to figure out whose economics policies are best. You have to wait at least 5 years to properly evaluate and then the circumstances have changed to determine the right strategy going forward.

    I sense the most successful scouts going forward will be the guys who mix a good eye for talent, with an understanding of the right temperment to succeed, with a strong understanding of how to use statistical anaylsis. I hope the Oilers are effectively evaluating their scouts for both their successes and mistakes, implementing a good development program for the scouts to evolve and every year bring in a few young scouts to freshen up the ranks.

  27. jfry says:

    For me the big take away is that we would never have taken toffoli, hey or weal even though they were easy to watch and their numbers in junior looked to have some sustain.we very rarel let someone fall to us unless were bragging that we had gernat 35th. Detroit seems to always be there in the second and third round when “skill” players are falling.

    The more we look at the drafting record the more I’m disapointed in our local staff. We draft horribly out of the chl, which is a bit mind blowing.

  28. Lowetide says:

    Jon: I’ve read several posts on lower round gems and it appears to be completely random. No one team delivers consistently in later rounds despite what Red Wings fans will tell you.

  29. Romulus Apotheosis says:

    godot10: This is relative.Smid isn’t a bad pick if you go back and look at that draft.

    A long career top 4 (though on the defensive side)is not a bad result for a 9th overall pick, in a weak draft (except for the top 2).

    You shouldn’t let the fact that we had to watch the sausage being made impact the fact that the resulting sausage actually turned out okay.

    I don’t disagree at all.

    Smid is my favorite current Oiler. Big fan!

    That said, erasing the hind-sight that he actually made it, the question is on draft day, what is the value of players with little to no offensive upside and what is the risk they don’t pan out.

    Smid covered the bet.

    But my suspicion is that if you take a generic player in the Smid mould more often than not you end up with a bust than a boom (at least for players of this type taken in the top 15).

  30. Lowetide says:

    gd:
    LA’s drafting to me looks as average to mediocre as the Oilers (and the majority of the NHL).

    -Picking Bernier at 11th shows the Kings weren’t ahead of the curve in realizing goalies shouldn’t be picked in the 1st round unless they are dominant.

    -The Hickey walkabout at 4 is one of the 10 worst picks of the last decade.

    -Going WHL three years in a row for 1st rd picks after the first two flopped looks questionable. Schenn is a okay pick, but OEL would have been a homerun.

    -The best thing they did was stay bad enough that they got a top 4 pick in 2008, which looks like the greatest year for D ever. Picking Voynov was obviously a great move and makes up for the huge miss on Tuebert.

    -I think their strength were the non-draft moves. Scuderi was the perfect UFA signing at the right time. The Carter/Richards trades were very shrewd deals. Quick’s emergence as a top 3 goalie definitely is crucial to Lombardi’s legacy. If either Roy, Bunz or Perhonen became an elite goalie, MacT’s would be well on his way to legendary status.

    It looks to me like trying to figure out what teams have a successful draft strategy is like trying to figure out whose economics policies are best. You have to wait at least 5 years to properly evaluate and then the circumstances have changed to determine the right strategy going forward.

    I sense the most successful scouts going forward will be the guys who mix a good eye for talent, with an understanding of the right temperment to succeed, with a strong understanding of how to use statistical anaylsis. I hope the Oilers are effectively evaluating their scouts for both their successes and mistakes, implementing a good development program for the scouts to evolve and every year bring in a few young scouts to freshen up the ranks.

    I specifically asked people what team they wanted, and LA was the overwhelming choice. I think ALL teams have hits and misses and some draft reaches and some draft fallers and both of thse categories please and disappoint.

    It’s a very random thing.

    However, what we’re looking at here–taking Moroz with about 25 names of daylight between pick and projection–is a small item in the overall picture. Did it pay off for LAK? I’d bet they say it did and if Moroz turns out like Clifford I bet the Oilers will be pleased too.

  31. Woodguy says:

    I’d like to see an analysis of Montreal.

    They killed it from 2001 – 2007 (their top 3 in 2007 are ridiculous)

    Seem to have fallen off the map since 2007. Wonder if there were staff changes in drafting?

    Here’s MTL from 2001-2007

    2007 Entry
    Draft Num. Player GP
    2007 Entry 12 Ryan McDonagh 169
    2007 Entry 22 Max Pacioretty 246
    2007 Entry 43 P.K. Subban 202
    2007 Entry 65 Olivier Fortier
    2007 Entry 73 Yannick Weber 115
    2007 Entry 133 Joe Stejskal
    2007 Entry 142 Andrew Conboy
    2007 Entry 163 Nichlas Torp
    2007 Entry 192 Scott Kishel
           
    2006 Entry
    Draft Num. Player GP
    2006 Entry 20 David Fischer
    2006 Entry 49 Ben Maxwell 47
    2006 Entry 53 Mathieu Carle 3
    2006 Entry 66 Ryan White 89
    2006 Entry 139 Pavel Valentenko
    2006 Entry 199 Cameron Cepek
           
    2005 Entry
    Draft Num. Player GP
    2005 Entry 5 Carey Price 310
    2005 Entry 45 Guillaume Latendresse 341
    2005 Entry 121 Juraj Mikus
    2005 Entry 130 Mathieu Aubin
    2005 Entry 190 Matt D’Agostini 267
    2005 Entry 200 Sergei Kostitsyn 353
    2005 Entry 229 Philippe Paquet
           
    2004 Entry
    Draft Num. Player GP
    2004 Entry 18 Kyle Chipchura 262
    2004 Entry 84 Alexei Emelin 105
    2004 Entry 100 J.T. Wyman 44
    2004 Entry 150 Mikhail Grabovski 367
    2004 Entry 181 Loic Lacasse
    2004 Entry 212 Jon Gleed
    2004 Entry 246 Greg Stewart 26
    2004 Entry 262 Mark Streit 491
    2004 Entry 278 Alexandre Dulac-Lemelin
           
    2003 Entry
    Draft Num. Player GP
    2003 Entry 10 Andrei Kostitsyn 398
    2003 Entry 40 Cory Urquhart
    2003 Entry 61 Maxim Lapierre 463
    2003 Entry 79 Ryan O’Byrne 308
    2003 Entry 113 Corey Locke 9
    2003 Entry 123 Danny Stewart
    2003 Entry 177 Christopher Heino-Lindberg
    2003 Entry 188 Mark Flood 39
    2003 Entry 217 Oskari Korpikari
    2003 Entry 241 Jimmy Bonneau
    2003 Entry 271 Jaroslav Halak 220
           
    2002 Entry
    Draft Num. Player GP
    2002 Entry 14 Chris Higgins 523
    2002 Entry 45 Tomas Linhart
    2002 Entry 99 Michael Lambert
    2002 Entry 182 Andre Deveaux 31
    2002 Entry 212 Jonathan Ferland 7
    2002 Entry 275 Konstantin Korneyev
           
    2001 Entry
    Draft Num. Player GP
    2001 Entry 7 Mike Komisarek 519
    2001 Entry 25 Alexander Perezhogin 128
    2001 Entry 37 Duncan Milroy 5
    2001 Entry 71 Tomas Plekanec 598
    2001 Entry 109 Martti Jarventie 1
    2001 Entry 171 Eric Himelfarb
    2001 Entry 203 Andrew Archer
    2001 Entry 266 Viktor Ujcik
         

  32. sliderule says:

    gd,
    The kings drafting in second and later is worlds better than Oilers.

    In the 2007 draft they picked four players after first round who are or have been regulars in NHL.

    Even the infamous Hickey pick played regular for isles last season.

    Five players from one draft .the oilers could wish for that.

  33. G Money says:

    If Bob’s list is “Drafting for Dummies”, the analysis above could suggest that “GMing for Dummies” would have as the golden rule:

    Trades: BE BOLD
    Drafts: BE CONSERVATIVE

    Looking at recent Oiler history, one could make the case that our regime the last few years did the opposite.

    Romulus Apotheosis: A new regime?

    Somehow both more public and more secretive?

    This is a mark of well-run companies, no? Things that are confidential are kept in confidence, while things that are not confidential are actively out in the open.

    Again – one could argue the regime the last few years did the opposite.

  34. Racki says:

    Woodguy: Gauthier did axe 6 scouts in 2010 (before that draft). Not sure what other turnover they’ve had, but that’s a fairly significant change.

  35. gd says:

    sliderule,

    I think to say the Kings have been worlds better is an exagerration. While Simmonds/Martinez/King represent great picks, it’s hard not to criticize having a 4th overall pick not play a game for your team and to be released on waivers. The Kings definitely appear to be a well put together team right now, but to me to say the Lombardi era drafting is anything but slightly above average would be a huge overstatement.

    The Oilers second rounds are concerning, and will be until someone establishes themselves as a bonafide NHLer. I have no problem with them hopefully specializing in WHLers, but it concerns me that they appear to be making the wrong calls on a lot of them. I had hoped when they made the Moroz pick that they had “inside info” that he would develop better than he has so far. My problem with the Abney/Hesketh picks were they left Eakin and Connauton on the table who should have been guys they would know lots about.

  36. Racki says:

    I’m sure it’s been mentioned ad nauseum, but speaking of 2007, boy did we ever blow that draft. The Oilers have 3 first round picks… get a hit with Gagner, but strike out badly with their 15th and 21st selections (Plante and Nash respectively). But looking at the draft, it seemed to be pretty sporadic. Lots of teams went off the map. But how nice would it have been to escape the draft with Gagner, Blum, and Pacioretty? Of course, hindsight is 20/20… Pacioretty and Blum would have been small reaches.

    Katz needs to put all his money together to research/build a time machine.

  37. Lowetide says:

    The 2007 draft cost Prendergast his job. No question.

  38. rickithebear says:

    TheOtherJohn: One thing is NOT like the other!

    Understanding successful drafting is interesting!

    this LA team:
    13 of 23 roster positions are drafted by them. 56%
    6 are first rounders 26%
    Brown #13 03
    Kopitar #11 05
    bernier #11 06
    Lewis #17 06
    Doughty #2 08
    Paerson #30 12

    2nd Round 3 13%
    Voyonov #32 08
    Clifford #35 09
    Toffoli #47 2010

    3rd -7th 4 17%
    Quick #72 05
    Martinex #95 07
    King #107 07
    Nolan #186 09

    The other 10 involving LA assets:
    Mike richards (simmonds 2nd), (schenn First)
    Carter (jonson 1st), 1st picK
    Penner (tuebert 1st) (1st LA klefbom) (3rd Zharkov)
    Stoll
    Greene (Vishnovsky)
    Regher ( 2- 2nd picks)
    Ellerby 5th rd pick

  39. rickithebear says:

    marancin for nash?
    Miss?

    13-14 oilers;
    Hall #1 10
    Gagner #6 07
    Yakupov #1 12
    RNH #1 11
    Eberle #22 08 (pronger)
    MP #10 09
    XXX
    XXX
    Smyth (fraser)
    Hartikainen #163 08
    Lander #40 09
    rajalas #101 09
    XXX
    XXX

    Smid (pronger)
    Petry #45 06
    Belov UFA
    klefbom #19 11 (Penner)
    XXX
    J. Schultz UFA
    N. Schultz (Salo)
    Potter UFA

    Dubnyk #14 04
    XXX

  40. Joel Pepin says:

    So in yesterday’s “2013 Entry Draft Post 2: Are the Oilers Guilty of Too Much ‘Walkabout’?” post, I finally made the plunge to try to contribute something to this wonderful blog. Lowetide’s tremendous work on drafts, and his line in the sand of games played, made me wonder if there is a way to better evaluate draft success. I pulled the draft data from HockeyDB and made a monster spreadsheet. I did it twice to check my numbers (though if anyone wants a copy of the spreadsheet, I would be thrilled to have some 3rd party confirmation).

    Total Players Drafted (2008-2012):
    Defense: 366 (35% of players drafted)
    Forward: 578 (55% of players drafted)
    Goalies: 108 (10% of players drafted)

    Total NHL Games Played by all players drafted:
    Overall: 16,238 (15.44 GP per drafted player)
    Defense: 6,216 GP (16.98 GP per drafted player)
    Forwards: 9,768 GP (16.90 GP per drafted player)
    Goalies: 254 GP (2.35 GP per drafted player)

    A player drafted in 2012 would likely have played less than a player drafted 4 years earlier. To normalize this, I calculated for each player the maximum possible GP. So a player drafted in 2008 could have played a max of 376 games, whereas a 2012 player may have played 48 games max. I calculated this per position D/F/G (using data for all 30 teams). We get the following:

    Drafting Defensemen:
    Round (% of GP)
    1 (26.0%)
    2 (5.2%)
    3 (1.6%)
    4 (0.8%)
    5 (0.6%)
    6 (1.5%)
    7 (1.1%)

    Drafting Forwards
    Round (% of GP)
    1 (27.6%)
    2 (4.9%)
    3 (2.2%)
    4 (2.2%)
    5 (4.2%)
    6 (1.4%)
    7 (0.5%)

    Drafting Goalies:
    Round (% of GP)
    1 (0.0%)
    2 (1.7%)
    3 (0.3%)
    4 (1.5%)
    5 (0.5%)
    6 (0.1%)
    7 (0.9%)

    So overall, a forward drafted in the 1st round of 2008 would have been expected to play 103.8 games by now (27.6% of 376), whereas a defenseman drafted in the 3rd round of 2010 should be expected to play 3.3 games (1.6% of 212).

    Based on the round that players are drafted, to keep up with the rest of the league, each draft pick should be expected to yield a certain number of GP. A few caveats:

    - This doesn’t consider draft position (e.g., drafting 1st overall vs. 30th). However, we’ve got over 1000 data points, so I believe it’s a good start.
    - More nuanced analysis is close. My post yesterday shows that we can actually quantifiably see how blueliners develop over time (a D drafted in 2009 should have played 7.3% of Max GP, but a D from 2008 should have played 13.6% of Max GP).
    - Weak players on poor teams will play more games in regular season (e.g., Lander), but excellent players on poor teams will play fewer games in playoffs (e.g., Hall)
    - If a player is traded, their GP are still allotted to the original team that drafted them. E.g., LA gets a benefit from trading Teubert to a weak team like the Oilers with a sparse blue

    Here is the D, F, and G draft data for LA and the Oilers from 2008 to 2012:

    Edmonton
    (Position) – (Actual GP) – (Expected GP)
    D – 1 – 73
    F – 800 – 382
    F* – 479 – 276 *Data after removing actual & expected GP from Hall/RNH/Yak
    G – 0 – 4

    Los Angeles
    (Position) – (Actual GP) – (Expected GP)
    D – 490 – 285
    F – 482 – 200
    G – 0 – 9

    So even with the 3 #1 overalls removed, Edmonton still drafted forwards nearly as well as LA. LA drafted defensemen exceptionally well too (whereas the Oilers didn’t do very well in that area). Neither team dedicated a lot of good draft picks on goalies, nor did either team get any GP from drafted goalies in this time period.

    I’m going to make one more post, looking at draft leaders per position. I really hope this is driving the conversation in a good direction. Any feedback would be wonderful. Thanks Lowetide!!

  41. Lowetide says:

    Awesome stuff, Joel. I was going to look into the EDM D but I think that number will come around once Klefbom, Marincin and others begin to show up. LA”s blue (Doughty, Voynov) will never be caught though, outstanding.

  42. Captain Happy says:

    I’ve always believed each year’s draft should be evaluated on it’s own merit since the quality of players can vary so widely from year to year.

    While far from scientific, I thought a numerical indication based on draft position and NHL games played might be at least interesting.

    So, draft position multiplied by NHL games played:

    2007:

    LAK

    Thomas Hickey – (4X39) 156
    Oscar Moller – 4524
    Wayne Simmonds – 22387
    Alec Martinez – 13490
    Dwight King – 8720

    LAK 2007 Total – 49,277

    EDM

    Sam Gagner – 2,484
    Alex Plante – 150
    Riley Nash – 777
    Linus Omark – 6305
    Milan Kytnar – 127

    EDM 2007 Total – 9,843

    2008:

    LAK

    Drew Doughty – 728
    Colten Teubert – 312
    Slava Voynov – 3,264
    Andrei Loktionov – 10,701

    LAK 2008 Total – 15,014

    EDM

    Jordan Eberle – 4,290
    Johan Motin – 103
    Phillpe Cornet – 266
    Teemu Hartikainen – 8,476

    EDM 2008 Total – 13,135

    2009:

    LAK

    Brayden Schenn – 550
    Kyle Clifford – 7,175
    Jordan Nolan – 13,020

    LAK 2009 Total – 20,745

    EDM

    Magnus Paajarvi – 1,630
    Anton Lander – 2,680

    EDM 2009 Total – 4,310

    2010:

    LAK

    Tyler Toffoli – 270

    LAK 2010 Total – 270

    EDM

    Taylor Hall – 171

    EDM 2010 Total – 171

    2011:

    LAK

    0

    LAK 2011 Total 0

    EDM

    RNH – 102

    EDM 2011 Total – 102

    2012:

    LAK

    Tanner Pearson – 30

    LAK 2012 Total – 30

    EDM

    Nail Yakupov – 48

    EDM 2012 Total – 48

  43. Joel Pepin says:

    And finally, based on the differential of actual GP vs. expected GP, we have the following best 5 teams in terms of drafting:

    Defensemen Differential (GP vs. Expected GP):
    1) LA: +205
    2) NY Islanders: +197
    3) Florida: +181
    4) Toronto: +180
    5) Philly: +176

    27) Edmonton: -72

    Forwards Differential (GP vs. Expected GP):
    1) NY Islanders: +570
    2) Colorado: +477
    3) Edmonton: +418
    4) LA: +282
    5) NY Rangers: +250

    Forwards Differential (GP vs. Expected GP):
    * In this case, I remove the effect of Hall/RNH/YAkupov, but DID NOT remove Stamkos or Tavares from Tampa or NYI respectively
    1) NY Islanders: +570
    2) Colorado: +477
    3) LA: +282
    4) NY Rangers: +250
    5) Tampa Bay: +240
    6) Edmonton: +203

    Goaltending Differential (GP vs. Expected GP):
    1) Nashville: +54
    2) Washington: +48
    3) Minnesota: +17
    4) Florida: +16
    5) Ottawa & NY Islanders: +15

    19) Edmonton: -4

    I think we can rest easy that MBS has done a fantastic job drafting forwards, but we are having a rough time drafting blueliners. To put things in perspective, a differential of -70 isn’t even 1 full season of 1 D-man, so maybe the Oilers simply aren’t dedicating enough good draft picks to d-men? I don’t know.

    I think the only way to draft goalies is to draft lots and lots of them. Speaking of which, below are the resources spent per team (i.e., drafts in terms of expected GP) per position from 2008-2012:

    Defensemen:
    (Rank) – (Team) – (Expected GP):
    (1) Los Angeles -285
    (2) Minnesota – 245
    (3) Atl / Win – 235
    (4) Anaheim – 227
    (5) St. Louis – 224

    (25) Edmonton – 73

    Forwards:
    (1) NY Islanders – 434
    (2) Edmonton – 382
    (3) Atl / Win – 377
    (4) Anaheim – 364
    (5) Tampa Bay – 347

    Goalies:
    (1) Colorado- 24
    (2) St. Louis – 17
    (3) Florida – 15
    (4) NY Islanders – 14
    (5) Dallas – 12

    (18) Edmonton – 4

    It’s starting to look like when drafting forwards and defensemen, teams using a lot of resources on specific positions (LA on D and F, NYI and Edm on F) are being rewarded. There doesn’t seem to be any reward to throwing a lot of resources at goalies.

    Also, Atlanta/Winnipeg (for a weaker team) is really not drafting well.

  44. RickDeckard says:

    steveb12344,

    It’s funny because Pavelec is terrible. Pavelec’s best season (minimum 30 games) is the same as Dubnyk’s worst (min 30GP).

  45. Joel Pepin says:

    Captain happy… I think I know what you’re doing (getting more GP from a higher draft number should equal more success) and I agree that this needs to be considered.

    What I tried to do (for forwards) is plot the GP vs. inverse draft position (210-draft position) and I got the following (messy equation):
    y = 1E-08×5 – 5E-06×4 + 0.0008×3 – 0.0564×2 + 1.7164x – 11.376
    y = average games played
    x = inverse draft position (i.e., 210 – draft position)

    I need the help of someone smarter about getting equations out of masses of data. I don’t tend to trust anything as messy as the equation I just listed above.

  46. TheOtherJohn says:

    Joel Pepin:
    And finally, based on the differential of actual GP vs. expected GP, we have the following best 5 teams in terms of drafting:

    Defensemen Differential (GP vs. Expected GP):
    1) LA: +205
    2) NY Islanders: +197
    3) Florida: +181
    4) Toronto: +180
    5) Philly: +176…27) Edmonton: -72

    Forwards Differential (GP vs. Expected GP):1) NY Islanders: +5702) Colorado: +4773) Edmonton: +4184) LA: +2825) NY Rangers: +250

    Forwards Differential (GP vs. Expected GP):* In this case, I remove the effect of Hall/RNH/YAkupov, but DID NOT remove Stamkos or Tavares from Tampa or NYI respectively1) NY Islanders: +5702) Colorado: +4773) LA: +2824) NY Rangers: +2505) Tampa Bay: +2406) Edmonton: +203

    Goaltending Differential (GP vs. Expected GP):1) Nashville: +542) Washington: +483) Minnesota: +174) Florida: +165) Ottawa & NY Islanders: +15…19) Edmonton: -4

    I think we can rest easy that MBS has done a fantastic job drafting forwards, but we are having a rough time drafting blueliners. To put things in perspective, a differential of -70 isn’t even 1 full season of 1 D-man, so maybe the Oilers simply aren’t dedicating enough good draft picks to d-men? I don’t know.

    I think the only way to draft goalies is to draft lots and lots of them.Speaking of which, below are the resources spent per team (i.e., drafts in terms of expected GP) per position from 2008-2012:

    Defensemen:
    (Rank) – (Team) – (Expected GP):
    (1) Los Angeles -285
    (2) Minnesota – 245
    (3) Atl / Win – 235
    (4) Anaheim – 227
    (5) St. Louis – 224

    (25) Edmonton – 73

    Forwards:
    (1) NY Islanders – 434
    (2) Edmonton – 382
    (3) Atl / Win – 377
    (4) Anaheim – 364
    (5) Tampa Bay – 347

    Goalies:
    (1) Colorado- 24
    (2) St. Louis – 17
    (3) Florida – 15
    (4) NY Islanders – 14
    (5) Dallas – 12

    (18) Edmonton – 4

    It’s starting to look like when drafting forwards and defensemen, teams using a lot of resources on specific positions (LA on D and F, NYI and Edm on F) are being rewarded.There doesn’t seem to be any reward to throwing a lot of resources at goalies.

    Also, Atlanta/Winnipeg (for a weaker team) is really not drafting well.

    Fascinating read Joel.

    If one removes the 1st round from the equation is it your conclusion that “MBS has done a fantastic job drafting forwards” hold up? Because if I read your table right Eberle’s 195 NHL games played constitutes 51% of the 382 non HAll/Yak/RNH F games played. Or is the conclusion largely predicated on Eberle;s #’s?

    Because the area that LAK is pulling some very nice F’s in the draft is in the 2nd, 3rd, 4th rd picks. The Oilers: not so much

  47. Joel Pepin says:

    And LT, in response to your comment about never catching LA’s D… the Oilers shouldn’t expect to ever catch LA. In terms of draft picks used, LA spent more than any other team trying to get blueliners. In comparison, the Oilers are 25th in the league in terms of using draft resources to get d-men.

    If the Oilers had gotten ~ 50 GP out of drafted D from 2008-2012, we would have been matching pace with LA (~ 1/4 the GP).

  48. cabbiesmacker says:

    There needs to be endless discourse and hypothesis to discover the Oilers can’t draft if their lives depended on it outside of round one?

    Mediocre Bastard Stu. Those seconds, especially the early ones the past few years, are beauties bud. Brandon Saad says hello.

  49. Romulus Apotheosis says:

    Woodguy,

    I ran LT’s test on MONT, CBJ and ISLES yesterday in the thread from 2008-2012. here’s Mont:

    (drafted – player – BM’s Rank)

    2008
    56 – kristo – 39 (under range)
    86 qualier – unranked (who knows)

    2009
    18 – leblanc – 17 (under)
    65 – nattinen – unranked (walkabout)
    79 – bennett – unranked (walkabout)

    2010
    22 – tinordi – 23 (within range)

    2011
    17 – beaulieu – 11 (under)
    97 – didier – unranked (walkabout)

    2012
    3 – galchenyuk – 5 (within range)
    33 – collberg – 27 (under)
    51 – thrower – 39 (crazy under)
    64 – bozon – 50 (crazy under)
    94 – vail – 60 (stupid crazy under)

  50. Lowetide says:

    Joel Pepin:
    Captain happy… I think I know what you’re doing (getting more GP from a higher draft number should equal more success) and I agree that this needs to be considered.

    What I tried to do (for forwards) is plot the GP vs. inverse draft position (210-draft position) and I got the following (messy equation):
    y = 1E-08×5 – 5E-06×4 + 0.0008×3 – 0.0564×2 + 1.7164x – 11.376
    y = average games played
    x = inverse draft position (i.e., 210 – draft position)

    I need the help of someone smarter about getting equations out of masses of data.I don’t tend to trust anything as messy as the equation I just listed above.

    Stay on your own course, Joel. You’re doing fine. :-)

  51. Joel Pepin says:

    TheOtherJohn… let me take a look into counting only rounds 2-7

  52. Joel Pepin says:

    Thanks LT.

    TheOtherJohn… below is the GP differential – GP(Actual)-GP(Expected) for all 30 teams, from 2008-2012, for only forwards, when excluding the 1st round picks…

    RANK TEAM GP Differential (Rounds 2-7): GP Actual – GP Expected
    1 NY Rangers +318
    2 Colorado +269
    3 Los Angeles +266
    4 NY Islanders +224
    5 Chicago +193
    6 Ottawa +171
    7 Nashville +107
    8 Philadelphia +105
    9 Columbus +82
    10 New Jersey +72
    11 Dallas +61
    12 Tampa Bay +59
    13 Buffalo +52
    14 Anaheim +40

    15 Edmonton +31

    16 Washington +16
    17 San Jose +7
    18 Carolina -21
    19 Calgary -42
    20 Minnesota -43
    21 Montreal -45
    22 Detroit -47
    23 Vancouver -54
    24 Pittsburgh -60
    25 Florida -63
    26 Atl / Win -75
    27 Boston -77
    28 St. Louis -93
    29 Toronto -95
    30 Phoenix -97

  53. Captain Happy says:

    Joel Pepin:
    Captain happy… I think I know what you’re doing (getting more GP from a higher draft number should equal more success) and I agree that this needs to be considered.

    What I tried to do (for forwards) is plot the GP vs. inverse draft position (210-draft position) and I got the following (messy equation):
    y = 1E-08×5 – 5E-06×4 + 0.0008×3 – 0.0564×2 + 1.7164x – 11.376
    y = average games played
    x = inverse draft position (i.e., 210 – draft position)

    I need the help of someone smarter about getting equations out of masses of data.I don’t tend to trust anything as messy as the equation I just listed above.

    I’ve been very interested in what you’ve done…I was just providing a different way of looking at things.

    Teams that find real NHL players in later rounds of the draft seem to be very successful since it adds so much to organizational depth and provides the team with trade ammunition when the timing is right.

  54. TheOtherJohn says:

    Great read Joel.

    The added analysis moves the MBS away from a “fantastic job drafting forwards” to 15th. I expect it you removed Lander’s ridiculous 67 GP dressed in the NHL when he was not ready to play at that level we’d get the Oilers to the low 20′s

    You have, though, identified the true Magnificent Bastards when it comes to drafting F:

    1 NY Rangers +318
    2 Colorado +269
    3 Los Angeles +266
    4 NY Islanders +224
    5 Chicago +193
    6 Ottawa +171

    **** Great work!

  55. Hammers says:

    Has anyone looked at differences in any team that has changed there GM . Thinking over the last 4- 5 years . My guess is we may see major differences after a change . Bowman in Chi , Gillis in Van, etc and it maybe based on draft position . Right now McT will have a tough time beating Tambo because of 3 1st rounders plus Ebs and J Schultz & Magnus . That needs to be taken into account unless i’m missing something .

  56. Lowetide says:

    Hammers:
    Right now McT will have a tough time beating Tambo because of 3 1st rounders plus Ebs and J Schultz & Magnus . That needs to be taken into account unless i’m missing something .

    Still, if the Oilers draft well those players will eventually find their way to the NHL (with the Oilers or through trades, etc). Shaun Van Allen was drafted #105 overall in 1987, and posted some exceptional AHL seasons before finally getting a regular job in 1993.

  57. Joel Pepin says:

    Thanks Captain Happy and TOJ… I also ran the data for Rounds 2-7 for D-men.

    1 NY Islanders +285
    2 San Jose +150
    3 Colorado +89
    4 Calgary +82
    5 Dallas +69
    6 Los Angeles +67
    7 Minnesota +64
    8 Nashville +51
    9 Carolina +48
    10 Philadelphia +38
    11 Columbus +36
    12 Phoenix +31
    13 Ottawa +25
    14 Washington +15
    15 NY Rangers +8
    16 Anaheim -3
    17 Buffalo -6
    18 Florida -11
    19 Boston -12
    20 Tampa Bay -13
    21 Montreal -13
    22 Atl / Win -17
    23 Vancouver -28
    24 Toronto -33
    25 Detroit -37
    26 St. Louis -37

    27 Edmonton -39

    28 Pittsburgh -39
    29 New Jersey -56
    30 Chicago -67

    So overall, great work by NYI, NYR, Colorado and LA for getting quality F & D players from Rounds 2-7.

    Finally, to quantify Captain Happy’s comment about bias for crummier teams to draft players with more GP, I ran the numbers and that bias definitely exists. Using all data for all 30 teams, Rounds 1-7, 2008-2012, and plotting the GP differential (games played minus expected games played) vs. regular season points, I got the following relationship:

    (GP Differential) = 1309.6 – (2.8071)*(Total Regular Season Points)

    In other words, for every extra win (2 points) during those 5 seasons, on average, that meant that the team should have had a drafted player during the same period play 5.6 fewer regular season games. I tried to normalize this by adding to the GP differential 2.8071 per point of regular season hockey. The leaders are as follows:

    Team – Normalized GP Differential
    (1) NYI – 1756 (i.e., GP Differential of 782, +2.8071 x 347 Regular Season Points)
    (2) LA – 1691
    (3) Philly – 1601
    (4) NYR – 1576
    (5) Colorado – 1517

    (15) Edmonton – 1261

    (29) Calgary – 1060
    (30) Montreal – 1027

    I need more time to think about this. The linear relationship I listed has an R2 value of 0.31, so that’s pretty hokey. And despite the fact that this puts Calgary near the bottom of the heap (always great to see), I think incorporating actual draft position should be the holy grail of this kind of work (as opposed to just looking at a round-by-round basis). Also, more data should be incorporated (going back 5 years isn’t enough, since the good teams before the 2008 draft would have still gotten a lot of good miles from previously-drafted players.

    Thoughts or comments? Thanks everyone!

  58. melancholyculkin says:

    Joel Pepin,

    Is there any chance you could send me your spreadsheet? I’d be interested in seeing what I could make out of it.

    email: tomashavard@gmail.com

  59. melancholyculkin says:

    Some really good discussion here. The question that remains is whether there’s a demonstrable skill at finding players outside the first round, or if a team like LAK is just riding a 5 draft heater. I think we’ve got a good start to answering it.

  60. vishcosity says:

    The capt happy model may potentially over reward great late round picks while the Pepin model seems to be potentially skewed by one great pick.

    Fascinating measures, I’m kinda disappointed that I didn’t think of either one.

  61. Joel Pepin says:

    melancholyculkin,

    Thanks for looking into this. I emailed you the spreadsheet, and I have inserted some highlighted cells with brief descriptions of each tab. The Pivot Chart is really the key, and for us to get better data, I think we’re going to have to add more draft years of data to the first tab.

    Good luck!

  62. vishcosity says:

    Now that I think of it, in college we used noise reduction algorithms. It’s been too long to remember how they work, but I wonder of someone else may remember. The idea was to average the extreme data points with their neighbours and soften their effect. One was called boxcar averaging I think. I remember others too but not by title nor means of application.

  63. Joel Pepin says:

    vishcosity,

    Hey Vishcosity, I’m kind of surprised by how drafting does affect the teams. I’m now used to following the Oilers where each year, the draft yields another major piece of the puzzle… most teams don’t seem to operate that way. Consider this…

    During 2008-2012, Detroit’s drafted players contributed a total of 72 GP over ~4.5 seasons. During the same time, the NY Islanders were top of the heap with 1392 GP, while LA was second with 972 GP (the average was 541).

    Now consider this… if your team signed 3 great UFAs for the entire 2008-2012 seasons (e.g., a 3rd line C, a big top-6 winger, and a top 4 D-man), who were healthy only 90% of the time, they’ve just eaten up over 1000 GP. That more than compensates for any incredible drafting a team can do.

    Based on that scenario, I would argue that for the OIlers moving forward, MacT being able to sign good UFAs will completely overshadow any drafting they do in the next couple of years.

  64. melancholyculkin says:

    Joel Pepin,

    Thanks Joel! I can’t promise I’ll find anything, but we’ll see.

    Cheers.

  65. Captain Happy says:

    Joel Pepin,

    Just awesome stuff.

  66. VOR says:

    Melancholy Culkin,

    I remain utterly unconvinced that we have made any progress. First of all how can we say anything about the success of a scouting staff from 2008-2012?

    From the 2012 draft there are 6 players in the NHL (any games)
    From the 2011 draft there are 22 players in the NHL (any games)
    From the 2010 draft there are 40 players in the NHL (any games)
    From the 2009 draft there are 77 players in the NHL (any games)
    From the 2008 draft there are 96 players in the NHL (any games)

    The average exclusive of the 8 and 9th round in the previous five years was 105. So from the 2012 draft there are probably around 99 players left to arrive (though apparently Mitch Moroz is a bust so we shouldn’t be expecting him).

    Then Joel is making no allowance for draft opportunity – in 2008 and 2009 using the points system developed by Michael Schuckers the LA Kings draft choices had a value of 4061. There were 19 of them.

    The Oilers had 12 picks combined in those two years and a draft pick value of 2041. In other words in the first two years of the head scouting of MBS he had half the opportunity LA’s scouts had.

    That said lets go back and look at games played by LA draft picks from 2008 and 2009 combined they come to a total of 962. In the same period the Oilers drafted players who have played to date 480 games. Then if we correct for the difference in draft pick value (ie. opportunity) 4061/2041 * 480 = 955. So Stu corrected for opportunity gets 955 games versus 962 games for LA’s scouting staff. Stu must be total shit screwing up that badly.

    Since then by the way LA has a total of 10 games played from the next three drafts. The Oilers have 281. This idea that we should ignore the first overalls so as to handicap Stu is idiotic. We are going to drop Drew Doughty aren’t we? He makes up over a third of all LA games played in the last five draft classes. You are thinking he wasn’t a 1 but a 2. Schuckers work shows a 1OV has a value of 917 and a 2nd OV has a value of 871 and the basis of that we are keeping one guy and junking three.

    I think there may be an agenda here. We have sliderule analyzing the years 2005-2009 for example. No offence sliderule but small sample sizes make me nervous. Use Schuckers and take that out to 2000-2009 (ie. the Lowe years where there is enough time passed to have a clue about the draft) and the Oilers actually outperform LA relative to each teams opportunity. They both outperform league average.

    By the way the last three years LA’s opportunity has been shit and the Oilers have had massive opportunity. Maybe they tanked, maybe they didn’t. However, it is far too early to tell for the last three years. So Melancholy Culkin the truth is we are just chasing phantoms based on tiny sample sizes and impatience. Well, that and a desire to be negative in all things related to the Edmonton Oilers.

    Joel’s analysis is brilliant in concept but lacks any sort of statistical testing to demonstrate it has any significance whatsoever. It is hard to believe it could given how few players have made it into the NHL from those last 3 draft classes.

  67. Lowetide says:

    Joel Pepin:
    vishcosity,

    Hey Vishcosity, I’m kind of surprised by how drafting does affect the teams.I’m now used to following the Oilers where each year, the draft yields another major piece of the puzzle… most teams don’t seem to operate that way.Consider this…

    During 2008-2012, Detroit’s drafted players contributed a total of 72 GP over ~4.5 seasons.During the same time, the NY Islanders were top of the heap with 1392 GP, while LA was second with 972 GP (the average was 541).

    Now consider this… if your team signed 3 great UFAs for the entire 2008-2012 seasons (e.g., a 3rd line C, a big top-6 winger, and a top 4 D-man), who were healthy only 90% of the time, they’ve just eaten up over 1000 GP.That more than compensates for any incredible drafting a team can do.

    Based on that scenario, I would argue that for the OIlers moving forward, MacT being able to sign good UFAs will completely overshadow any drafting they do in the next couple of years.

    Along the same lines, consider how difficult the road up the mountain has become for Hartikainen. Signs with the Oilers in 2010, that season sees rookies Hall, Eberle, Paajarvi and Omark among F’s. Following season? Nuge. Last year? Yakupov.

    I expect some of these kids who are vying for skill jobs (Rajala, etc) may get pushed back a little as we move forward.

    Of course, Detroit’s prospects have the added difficulty of prospects who are ahead by AHL seasons waiting their turn.

  68. Lowetide says:

    VOR:
    Melancholy Culkin,

    I remain utterly unconvinced that we have made any progress. First of all how can we say anything about the success of a scouting staff from 2008-2012?

    From the 2012 draft there are 6 players in the NHL (any games)
    From the 2011 draft there are 22 players in the NHL (any games)
    From the 2010 draft there are 40 players in the NHL (any games)
    From the 2009 draft there are 77 players in the NHL (any games)
    From the 2008 draft there are 96 players in the NHL (any games)

    The average exclusive of the 8 and 9th round in the previous five years was 105. So from the 2012 draft there are probably around 99 players left to arrive (though apparently Mitch Moroz is a bust so we shouldn’t be expecting him).

    Then Joel is making no allowance for draft opportunity – in 2008 and 2009 using the points system developed by Michael Schuckers the LA Kings draft choices had a value of 4061. There were 19 of them.

    The Oilers had 12 picks combined in those two years and a draft pick value of 2041. In other words in the first two years of the head scouting of MBS he had half the opportunity LA’s scouts had.

    That said lets go back and look at games played by LA draft picks from 2008 and 2009 combined they come to a total of 962. In the same period the Oilers drafted players who have played to date 480 games. Then if we correct for the difference in draft pick value (ie. opportunity) 4061/2041 * 480 = 955. So Stu corrected for opportunity gets 955 games versus 962 games for LA’s scouting staff. Stu must be total shit screwing up that badly.

    Since then by the way LA has a total of 10 games played from the next three drafts. The Oilers have 281. This idea that we should ignore the first overalls so as to handicap Stu is idiotic. We are going to drop Drew Doughty aren’t we? He makes up over a third of all LA games played in the last five draft classes. You are thinking he wasn’t a 1 but a 2. Schuckers work shows a 1OV has a value of 917 and a 2nd OV has a value of 871 and the basis of that we are keeping one guy and junking three.

    I think there may be an agenda here. We have sliderule analyzing the years 2005-2009 for example. No offence sliderule but small sample sizes make me nervous. Use Schuckers and take that out to 2000-2009 (ie. the Lowe years where there is enough time passed to have a clue about the draft) and the Oilers actually outperform LA relative to each teams opportunity. They both outperform league average.

    By the way the last three years LA’s opportunity has been shit and the Oilers have had massive opportunity. Maybe they tanked, maybe they didn’t. However, it is far too early to tell for the last three years. So Melancholy Culkin the truth is we are just chasing phantoms based on tiny sample sizes and impatience. Well, that and a desire to be negative in all things related to the Edmonton Oilers.

    Joel’s analysis is brilliant in concept but lacks any sort of statistical testing to demonstrate it has any significance whatsoever. It is hard to believe it could given how few players have made it into the NHL from those last 3 draft classes.

    No one is adopting this as gospel, we’re just blue skying. I do agree though that taking out the #1′s changes the framework (we have to deal with the facts) but as a fun game I don’t see any harm.

    I think Stu MacGregor has done a fine job, nothing in these numbers is going to convince me today. However, 5 years from now we can assess the 2008-2012 draft and some of the things discussed here may give us better guidelines.

    It’s all good imo.

  69. godot10 says:

    Joel Pepin:
    vishcosity,

    Now consider this… if your team signed 3 great UFAs for the entire 2008-2012 seasons (e.g., a 3rd line C, a big top-6 winger, and a top 4 D-man), who were healthy only 90% of the time, they’ve just eaten up over 1000 GP.That more than compensates for any incredible drafting a team can do.

    Based on that scenario, I would argue that for the OIlers moving forward, MacT being able to sign good UFAs will completely overshadow any drafting they do in the next couple of years.

    I disagree. Improving an NHL team in a hard cap environment is all about value-for-money. Unrestricted established free agents (non-entry level) are almost never value-for-money. Name UFA’s are final pieces of the puzzle, not building blocks.

    The UFA’s which tend to be value-for-money are the 2nd tier more unknown ones toiling at the bottom of NHL rosters, or in the AHL.

    Paying full value for established UFA D is probably the only place I would dare tread for a rebuilding team. Incoming forwards should all be of the value-for-money variety on a rebuilding team.

    Or else one is really not getting ahead. One has to become playoff competitve mostly with a value for money philosophy. Paying full value is for the push from playoff competivenss to contention.

    But I’m willing to make exceptions for defensemen.

  70. VOR says:

    I actually agree Lowetide. I think Joel’s approach is brilliant and probably would work well applied to earlier drafts where we have all or nearly all players in the NHL that ever will be. In five years the same analysis will be closer to being valid for 2008-2012. In ten years it will be closer still.

    I am just unconvinced that any five year run by any one scout staffing will ever be valid proof that it isn’t all just luck. I think it is a classic small sample size problem. The noise probably far outweighs the signal at all times post draft for any one team for any given five year period.

    I am certain Joel’s approach also needs to be adjusted to actual draft order (rather than just by round) using something like Michael Schuckers work, which rickithebear posted in an earlier thread. To show its power as a tool consider that Joel has pointed out the Islanders are doing much better than Detroit over the last five years in terms of drafted players making the NHL and playing significant games.

    Schuckers lets us figure out how much more opportunity NYI may have had than DET. The Schuckers # for all of the last five year draft choices of the NYI is 8949 and it is heavily back loaded to 2008-2009. This is why the Islanders are stocked with prospects and they are arriving in the NHL. DET has a number of 4241. In other words the Islanders had more than twice the draft opportunity of the Wings. From 2008-2012 the Islanders actually have had more draft opportunity than the Oilers and should have a better outcome as a result if both scouting staffs are equal.

  71. melancholyculkin says:

    VOR,

    Can I assume this is the Shuckers paper you’re referring to? http://myslu.stlawu.edu/~msch/sports/Schuckers_NHL_Draft.pdf

    I am just unconvinced that any five year run by any one scout staffing will ever be valid proof that it isn’t all just luck. I think it is a classic small sample size problem. The noise probably far outweighs the signal at all times post draft for any one team for any given five year period.

    I’m in complete agreement. A lot of people point to whatever team happens to have had some draft success recently and proclaim them the new gods of drafting without offering any proof that their success is anything but luck.

    Ideally, draft success will be normally distributed among the teams and we can calculate standard deviation and set up confidence intervals and use all the other nice statistical tools at our disposal.

    Of course, you also want to take into account draft opportunity, and you’ve got expansion in the 90s buggering things up. Maybe do a study of the 30 team league from the 2000 NHL entry draft through until the 2008 draft. Of course that’s only 9 drafts (including 2008) and who knows if that’s a large enough sample.

    Then there’s the fact that GMs get hired and fired, and they’ve all got their own philosophies about the draft, so they hire and fire scouts accordingly. Also, who says a players success lies solely at the feet of the scouts? How important is a good farm system? Is the emergence of a late round pick credit to the scouts or to the development pipeline?

    Maybe there are some teams that have a skill at the draft table, but it’ll be a hell of a thing to prove. I just like Joel’s model and think it might reveal something if applied to a larger sample.

  72. Ryan says:

    I would tend to agree with this comment.

    In simple terms the problems boil down to:

    2009 – 40 Anton Lander
    2010 – 31 Tyler Pitlick
    2010 – 48 Curtis Hamilton
    2010 – 46 Martin Marcin
    2011 – 31 David Musil
    2012 – 32 Mitchell Moroz

    I’m not a draft guru, but out of 4 high and 2 mid 2nd round picks, I’d like to see a useful NHL player that slots in above 7d and 4c. Am I asking too much?

    Out of the six, the Pitlick, Hamilton, and Moroz picks are probably already blown. Musil has clay feet and Lander can’t produce at a 4rth line level in the bigs. Marcin’s the only hope.

    cabbiesmacker:
    There needs to be endless discourse and hypothesis to discover the Oilers can’t draft if their lives depended on it outside of round one?

    Mediocre Bastard Stu. Those seconds, especially the early ones the past few years, are beauties bud. Brandon Saad says hello.

  73. VOR says:

    Ryan,

    Marincin looks to be a hockey player. He needs to fill into his frame and learn to pick his spots better but nobody can seriously think he is a bust. He is an adventure certainly, but could well be, based on how things are going in the AHL, an impact NHL player.

    I think Hamilton is toast. I understand that part of it is injury after injury but even so it seems certain he will never see an NHL ice surface. A shame really because there was so much to like about him when the Oilers drafted him. I think Hamilton maybe needs to think about Europe where he might extend his career.

    Pitlick was tracking exceedingly well before suffering a very severe concussion while playing for Medicine Hat. He lost last year as a result – it was clear he wasn’t right somehow, very tentative. Then this year Pitlick was stuck a long ways down the depth chart until the lockout ended. Since then he has been improving quite steadily and during this year’s playoffs Pitlick has again looked like he might have an NHL future.

    Lets say on a scale of 1 to 6 (1 is sure fire NHLer – 6 is not a hope in hell) Marincin is a 2, Pitlick a 4, and Hamilton a 6.

    That brings us to Musil and Moroz. Musil is slow. Every scout knew that before he was drafted and even so every list had him +/- 30 OV. He was not a reach pick. He had a great season in junior this year. It just wasn’t a flashy one. Musil is a player who grows on you over time. He does the little things right. He will certainly, barring injury, get chances in the NHL. On top of which Musil has played well to brilliantly in international play. He is probably a 3, better chance of an NHL career than Pitlick less than Marincin. Hardly, however, the lost cause people keep claiming.

    Moroz is the real wild card. I get that everybody here hates him. Yet Moroz is stunningly good at what he does. That is check the crap out of the other teams’ top players. In fact, I think he is as good as any forward in the WHL in his own end of the ice. This is an elite skill we are talking about. On top of which Moroz is an above average forechecker and an above average fighter. Did I mention he is a plus skater?

    What Moroz doesn’t have is offence but he has never been put in a role where he would be expected or allowed to produce offence. Moroz won’t be the first forward to have a long NHL career because of their ability to shut other players down (Marcel Goc might be a good comparable). Nobody I have read or heard criticizing him has ever been able to offer any reason to think Moroz can’t play NHL level defence. If he can he will have a long and valuable NHL career on either a 3rd or 4th line. The hate for Moroz seems to be because people want him to be Milan Lucic based on an assumption that the Oilers had him on par with Samuelsson. For which there is no evidence.

    Perhaps the Oilers aren’t that stupid. Maybe they thought they were very weak on the third and fourth lines and saw a chance to take a 1000 NHL game checking forward who can fight. Like I said, if you think a Marcel Goc who can fight instead of comparing him to Lucic it is hard to argue he isn’t tracking very nicely indeed. The question with Moroz isn’t if he will have an NHL career. It is will he have enough offence for it to be on the 3rd line.

    What I don’t get is why you can’t all see Moroz is a 1, a certain NHLer? He works like a demon, limits his mistakes, understands and plays defence at an elite level, and is a willing and capable fighter. Didn’t we have this conversation all year about how the Oilers needed to upgrade the third and fourth line and be bigger and tougher. Yet when the Oilers spend a draft pick on a kid who could well be part of the solution to that long term problem you all shit on him. I really don’t get it.

    Goc was taken twenty over all and nobody thinks he hasn’t covered the bet. Why would you think Moroz’s defence won’t carry him equally far? Maybe you think teams don’t need shut down forwards who are tough as nails and don’t give an inch? In which case I would suggest watching a few Stanley Cup playoff games.

  74. vishcosity says:

    I think we can make statements about the 2008-2012 draft via LT’s and Rom’s system by measuring against consensus.

    Further, I think cpt happy’s idea is very worthy of further discussion if only because the math is so simple.

    The GP measure of Pepin’s system I think needs qualifiers a plenty such that its going to need verification through say 1991-1998 to see how we’ll it works. I’m nervous that extreme data points may be too affective, maybe those could be qualified by TOI or something. What to do with the 1 OV’s is another issue entirely.

    Still like the idea LT started with as it does give seemingly legit opportunity to measure the recent drafts and current scouting / GM decisions. That is the first I’ve seen of any system to measure current management which remotely resonates with me.

  75. VOR says:

    Vishcosity,

    I like all of these approaches. I think conceptually Joel’s system in brilliant. LT’s system is descriptive of how a team approaches drafting. I think Rom’s suggestions for improving the definition of a reach pick adds greatly to the descriptive power of LT’s idea. I do think Joel needs to use some system that looks not at the round a player was picked but the spot in which he was picked. I suggested one, there may be others.

    On the other hand I can’t help shaking the feeling that a lot of people here are already using these ideas to prove points that neither the concept nor the data set support. Am I wrong in hearing some of you saying that MBS is just BS because he makes all these reach picks and they don’t work out? You do all know it is far too early in Stu’s tenure to have the slightest clue whether he is Magnificent or Bastard? You do know the sample size if far too small for the results to mean anything?

    I just want to make sure you are all not getting the cart 1000s of miles before the horse.

  76. vishcosity says:

    A hybrid system using a late round multiplier like Happy model combined with the GP thing has real potential for draft 5 + years out. Not sure the formula needs E+5 as a variable, it can’t be fifth dimensional for any reason that I can fathom, I suspect that formula is a computer’s attempt to most closely fit the data.

    n = GP x square root of draft number may work, that would soften the outliers though a straight root may be an over correction. Something in the middle is surely possible.

    Box car averaging for anything beyond a standard deviation (if one is way out, average it with it’s two neighbours and re enter the outlier) would also reduce small sample effects.

    Probably the shift from round number to draft number should deal with the 1 OV’s (compared to Det at 22 or whatever) and a correction for TOI could be better all around.

    All that relates to drafts five years out, it would be outstanding to connect it to LT’s reach pick discussion and measure current regimes. That I don’t see just yet.

    There is a formula in this but I’m not going to be able to build it from this here phone whilst on the road. Anyone got the skills? I’m certain there are far better statisticians than me in this crowd.

    I hope some of that made sense, I’ve done about twelve things since I started writing this post and his it probably seems like I’ve been drinking.

  77. Joel Pepin says:

    Lots of good discussion and suggestions. I’ll try to read the Shuckers paper before crunching more numbers. Thanks for the link!

  78. vishcosity says:

    Another reasonable correction could include consideration for team depth at said position. For instance if Ebs Hemsky and Yak get 18 mins, 16 mins, and 14 mins the year prior, a draft pick pushing Hemsky down the list should get more value, although that would be less impressive if the pick pushed Belanger, Lander and VDV.

    So I don’t know about that. It would really need to consider overall player strength. Maybe the yahoo list would work for that. Except goalies because they rank them not legit so as to mess with those who don’t attend the live draft.

  79. VOR says:

    I want to return to LT’s original comment about defensible selections. If I understood him correctly Lowetide was suggesting the Moroz pick was not defensible. I think Lowetide is wrong and in being wrong invalidates many of the assumptions that underlie the models we have been talking about here.

    We act in a data desert when evaluating juniors. That alone means we can’t say Moroz wasn’t BPA at his slot in the draft. I watch a lot of Oil Kings games, going to nearly all the home games. I’d say by eye that in his draft year and then again this year Moroz played quite limited minutes (4th and then 3rd line) but against extremely tough competition. Moroz opponents alternate between the other teams most talented players and the other team toughest thugs. He and Ewanyk come out for the majority of the d-zone draws and thus are pushing hard ice, very hard ice.

    I wish I had data to support these statements but I don’t. But lets say my eyes aren’t lying. Perhaps the Oilers were drafting a player out of the consensus order because they value the ability to play defence more than other teams. They probably have access to the very data that would demonstrate Moroz’s defensive skills.

    By eye Moroz can forecheck very well. If there was a metric to measure that and confirm what our eyes see then we could say the Oilers drafted Moroz out of order because they value forechecking more than other teams. By eye Moroz’s Corsi numbers were out of this world and Moroz certainly had a very good plus/minus. If his corsi domination was confirmed by data we could say the Oilers like players who as juniors can push the river when the river is in flood. Again, it is worth noting the Oilers may have this data at their finger tips.

    His numbers tell us Moroz fights occasionally. Our eyes tell us he does it well. No stat can tell you that when Moroz is on the ice his team mates play with more courage. However, anybody who has watched a lot of Oil Kings games can tell you it is true. Some one has to watch Moroz live to see that he can really skate. It appears as you watch him play that Moroz gets stronger relative to his competition the longer he stays on the ice. Given the incredibly long shifts great offensive players get in the WHL that is a valuable skill and one that is confirmed by numbers.

    Moroz’s VO2 numbers at the combine tell us he has an amazing recovery time and tremendous ability to do high intensity cardiovascular work. Not in the top ten but certainly top thirty in the entire draft.

    You also have to see him life repeatedly to get that Mitch Moroz is Fing tough. Tough as nails. There is no stat for it but I would think we all get how important it is when you are being pulverized in the playoffs. You’d rather be the hammer than the nail but if you are on the receiving end you have to be able to shake it off and go right back to work. That is Mitch Moroz. Maybe the Oilers value that skill more than other teams. Maybe they simply know how consistently Mitch Moroz takes a hit and skates away exhibiting both toughness and discipline.

    I think the idea Moroz was an indefensible pick taken because of organizational need is far from proven. You all think he was taken because the Oilers were trying for a power forward. There is no proof that is what the Oilers were thinking.

    Maybe they wanted the best player available. A lot of things matter when determining the impact a player has in the NHL. An elite level checker who can fight, has NHL + speed, with above average cardiovascular capacity and who plays hard on the puck may well have more value than a scoring sensation whose game won’t transfer to the NHL.

    I also think that there is a widely held belief that players of this type grow on trees and the Oilers could have gotten someone similar or better later in the draft at less cost. The problem is that none of you ever test that theory. Feel free to point out actual examples of the players left in the draft when the Oilers picked Moroz who combined his ability to skate, to check, to fight, to score and had better numbers than Moroz at the combine.

    I am pretty confident finding such a player won’t be as easy a task as you all seem to think. That leaves you with the argument that these skills don’t matter. In which case I will give you my rebuttal in advance and yes I know I am repeating myself. Watch some playoff hockey!

    You can argue this way for most reach picks by every team. For one reason or another teams reach. If their reasons are sound then they aren’t reaching. The thing is we don’t know the reasons and if the player doesn’t work out we say it is because they were a reach pick. Maybe the reasoning was perfectly valid and bad luck wiped out an act of scouting genius. We have absolutely no way of knowing.

  80. DeadmanWaking says:

    Joel Pepin,

    Hey, Pepin, I’ll look at that sheet, too, if you’re sharing. Fire it at worehoarsewaging@yahoo.ca, which is fresh to the wagons. Using the R FDA package (Ramsay, Hooker, Graves) I can constrain the fit to be monotonic, and for the second derivative to have at most one zero between the 1st and 10th pick overall. We’ll see if I can work this quickly.

    Monotonic: No-one really believes the 40′th pick is better than the 35′th pick long term, whatever ripples we find in the data.

    2nd derivative: For most methods of scoring potency against draft position, we know the largest gaps occur at the beginning of the draft. The difference between 1st and 3rd is more than the difference between 10th and 12th (if the data says differently, I’d discard the data). We’d also expect the derivative to be monotonic. However, using games played as your potency proxy, you’ve hard-capped the top few picks, so this might create an inflection point that wouldn’t otherwise exist: in games played, Hall and Seguin can only be distinguished on health (and Ruffness as I recall Bruce paining over at Cult). Actually, I like the FDA package because curve fitting has a “roughness penalty” whereby one can squeeze down on “total curvature” (should one be so inclined). Total curvature is defined as the integral of acceleration squared. Seriously. I think this package even does confidence intervals on the curve fit (which will flare like a trumpet at the high end). To constrain the flare, I could additionally tell it that if there were Mario or Crosby sextuplets available at picks -5 to -10 they’d pretty much all get the same number of games played (games played is not a good metric for sifting a generational plethora).

    In other tangents, I’m reading two books right now freshly published. The Success Equation: Untangling Skill and Luck in Business, Sports, and Investing and Mating Intelligence Unleashed. These tomes are strangely synergistic, though surprisingly the second is far more technical in tone than the first. I think I’ll use them as mutual book-marks. Perhaps I’ll interpose The Many Shades of Luck cheek-to-cheek with She Totally Wants Me. Does anyone else leave their unfinished books lying around common-law akimbo?

    I’m also living large today (or at least walking tall): I pivoted my right-hand 1920×1080 panel into the portrait orientation. I tried something like this long ago, but found the video card was very sluggish driving the transposed display. What brought this on, all of a sudden? Well my lemon tree (aka my squeeze) forgot a necessary wire when she went off to horse-sit so she pops in this afternoon wearing shades to snag the cable and notices that she can only read one of my two monitors. The other one is black. “Weird,” she goes. “Ah, yes” I say … “my two LCD panels are polarized in the opposite orientations”. With polarized shades, my right panel goes black, unless you tilt your head. So I ask my lemon tree “How does the photon know how to be half visible when you tilt your head 45 degrees?” Clever little photons. Then it came to me, hey, I could program wearing shades if I rotated the right panel upright. So now I have this 21″ tall viewing area on my desk and I’m amazed how well my video card eats this up. But not many pages I visit are long enough to really take it for a proper test drive with the SmoothWheel Firefox add-on I loaded yesterday (because this new PDF viewer integrated into Firefox was scrolling like one of those pixelated “Not in Service” marquee displays on the bony prominence of an empty TTC bus barreling past Carlton and Church in the mid-eighties).

    Need more blacktop! So what happens, I wonder, if I head to Lowetide on a busy day? Well, I quickly discover that this here “wandering” thread goes the distance and guess what … I managed to fit Pepin’s longest post on a single screen, minus the last inch (this in a font a bifocal-denier fluent in inches can still read).

    It’s pretty weird with my displays in this T-bone configuration. If I’m absent-minded, my mouse goes into the corners and never comes out. The clever thing was getting two panels in different years with different pixel resolutions having exactly the same dot pitch, so my mouse doesn’t warp across the divide. Bring on the 200 post 1st period where Gagner disappears from the line-card two hours before the puck drops and Lander draws in at 2C. I’m loaded for bear.

  81. Marc says:

    Ryan:

    I’m not a draft guru, but out of 4 high and 2 mid 2nd round picks, I’d like to see a useful NHL player that slots in above 7d and 4c.Am I asking too much?

    Out of the six, the Pitlick, Hamilton, and Moroz picks are probably already blown.Musil has clay feet and Lander can’t produce at a 4rth line level in the bigs.Marcin’s the only hope.

    Statistically you should expect two of those six picks to play 200 games in the NHL in any capacity and four of them to bust. Lander should probably make it to 200 in a fourth line role and Marincin is tracking well so we’re seeing exactly what the stats tell us we should expect to see. If any of the others make it to 200 NHL games, even as a fourth liner, MBS will have done an above average job with these second round picks.

    There seem to be a lot of people on here that think that ‘only’ getting a fourth liner with a second round pick is somehow a failure. The reality is that getting any NHL player out of the second round beats the odds and should be seen as good drafting.

  82. Marc says:

    For a bit more context, between 1998 – 2008 there were a total of 324 second round picks. Only 13 – or about 1 out of 25 – have been an NHL Allstar.

    Going through the lists of second round picks, there were about 3-5 each season that jumped out as good ie. not bottom of the roster, NHL players – like Stoll or Kulemin or Petry. That works out to roughly 1 every 8 picks.

    Expecting MBS to have come up with 2 or 3 such players out of 6 picks seems to me to be setting an unrealistically high standard.

  83. vishcosity says:

    Vor -

    I’ll happily be the first to say that regarding Moroz, I hope MBS is he smartest guy in the room. Hope the kid makes it big.

    I believe the point of the exercise was to find the teams that reach against the consensus (and conversely grab the drifters) more often than others.

    To this effort, Moroz actually making the pros is sort of irrelevant, (except to later question those who’s opinion became the baseline/consensus)

    Thanks for the perspectives though, if my boy can add some of what you describe, he may have a chance of moving on to what he currently claims to want to do.

  84. oilswell says:

    melancholyculkin: Ideally, draft success will be normally distributed among the teams and we can calculate standard deviation and set up confidence intervals and use all the other nice statistical tools at our disposal.

    Stupendously minor point: I’m not sure it is wise to assume that drafting is normally distributed. If drafting later than pick 120 is mostly random noise, then an individual draft might have teams show a normal distribution for normalized value they get out of a draft, but overall it should be a uniform distribution. In the case that there is repeatable ability to draft after pick 120 I don’t think we have any good idea if that ability is normally distributed. Consider for example, actual NHL talent: look at any draft class ranked by GP or P/GP and it looks like a power-law distribution, and I wouldn’t be terribly surprised if drafting ability is likewise.

  85. oilswell says:

    VOR: Then Joel is making no allowance for draft opportunity – in 2008 and 2009 using the points system developed by Michael Schuckers the LA Kings draft choices had a value of 4061. There were 19 of them.

    Being a scholarly paper I find it odd that Schuckers didn’t cite the Dawson and McGee paper. which provides a pick value ranking system that accounts for the strength of the draft class.

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

© Copyright - Lowetide.ca