HFBoards

Go Back   HFBoards > General Hockey Discussion > By The Numbers
By The Numbers Hockey Analytics... the Final Frontier. Explore strange new worlds, to seek out new algorithms, to boldly go where no one has gone before.

Scouting, Statistics, and St. Louis

Reply
 
Thread Tools
Old
08-24-2012, 03:19 PM
  #101
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Stats

Quote:
Originally Posted by Iain Fyffe View Post
Wow. So you obviously agree that using stats aids in the scouting process. And yet, my suggestion that such is true is met with complaint after complaint. You're a piece of work, you are.

My entire point is that with a fairly simple system using only very basic stats, you can get pretty good draft rankings. I can only imagine what can be done with a much more thorough system with access to better data.
My point has always been that stats are a small part of the overall scouting system package. Your point seems to be that stats are a stand alone indicator that work well without any contribution from all the other components that at the arena scouting brings to the decision making process.

The subpoint is that your stats package is underperforming compared to the NHL overall package. Certain elements have merit from what I have observed - the age factor is given a bit too much weight. At the Novice level the difference between a first year Jan and December birthday may be as much as 365 days or upwards of 20%.By the midget/junior levels the difference may still be as much as 365 days but the actual devleopmental difference is down to app 5-7%.

Your appreciation of Quebec Midget AAA stats,Quebec/Eastern Ontario Junior A stats is not in tune with the league realities - expansions,rule changes jurisdictional issues and team structures. These impact performance and the evaluation of same.

You overrate Quebec/QMJHL goalies especially in the second half of the nineties onwards because the stats simply did not show how the game had changed at that position.

Pretty good does not cut it in the NHL. Access to better data? Interesting question. Change your approach to cultivating possible resources.

Canadiens1958 is offline   Reply With Quote
Old
08-24-2012, 03:46 PM
  #102
wgknestrick
Registered User
 
Join Date: Aug 2012
Posts: 1,474
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
My point has always been that stats are a small part of the overall scouting system package. Your point seems to be that stats are a stand alone indicator that work well without any contribution from all the other components that at the arena scouting brings to the decision making process.

The subpoint is that your stats package is underperforming compared to the NHL overall package. Certain elements have merit from what I have observed - the age factor is given a bit too much weight. At the Novice level the difference between a first year Jan and December birthday may be as much as 365 days or upwards of 20%.By the midget/junior levels the difference may still be as much as 365 days but the actual devleopmental difference is down to app 5-7%.

Your appreciation of Quebec Midget AAA stats,Quebec/Eastern Ontario Junior A stats is not in tune with the league realities - expansions,rule changes jurisdictional issues and team structures. These impact performance and the evaluation of same.

You overrate Quebec/QMJHL goalies especially in the second half of the nineties onwards because the stats simply did not show how the game had changed at that position.

Pretty good does not cut it in the NHL. Access to better data? Interesting question. Change your approach to cultivating possible resources.

How in the world do you even know "who" to scout? You have to primarily rely on stats to even tell you who you should be watching.

I am thankful that the NHL has been backwards thinking for a while now because it allowed Jagr to fall to my team at number 5. I would be pretty upset if my team had selected Nolan, Nedved, Primeau, or Ricci in front him.

wgknestrick is offline   Reply With Quote
Old
08-24-2012, 04:16 PM
  #103
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Scouting

Quote:
Originally Posted by wgknestrick View Post
How in the world do you even know "who" to scout? You have to primarily rely on stats to even tell you who you should be watching.

I am thankful that the NHL has been backwards thinking for a while now because it allowed Jagr to fall to my team at number 5. I would be pretty upset if my team had selected Nolan, Nedved, Primeau, or Ricci in front him.
Did it at the Bantam/Midget/Junior level for over 20 seasons. Still "birddog". Rather simple. Get the rosters. Watch the warm-up. Watch the game. Eliminate the non-prospects - weaker skaters for the level and so on down the line. Grade the rest. Quantify and qualify the prospects.

Canadiens1958 is offline   Reply With Quote
Old
08-24-2012, 05:06 PM
  #104
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
Did it at the Bantam/Midget/Junior level for over 20 seasons. Still "birddog". Rather simple. Get the rosters. Watch the warm-up. Watch the game. Eliminate the non-prospects - weaker skaters for the level and so on down the line. Grade the rest. Quantify and qualify the prospects.
Let's play the same game with your method that you're playing with Iain's. How many of your "star" graded prospects have succeeded in the NHL? And how many have failed? And how many stars did you scout, but miss?

I'll of course ask for something published several years ago, and not just a retrospective thing that you whip up today.

(I've always found it interesting that you have had the same reaction to this as a lot of traditional scouts had to the "Moneyball" concepts when they were first proffered - now it's easier to see why. You have a vested interest in the "old guard".)

Chalupa Batman is offline   Reply With Quote
Old
08-24-2012, 07:56 PM
  #105
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
What I Did

Quote:
Originally Posted by Taco MacArthur View Post
Let's play the same game with your method that you're playing with Iain's. How many of your "star" graded prospects have succeeded in the NHL? And how many have failed? And how many stars did you scout, but miss?

I'll of course ask for something published several years ago, and not just a retrospective thing that you whip up today.

(I've always found it interesting that you have had the same reaction to this as a lot of traditional scouts had to the "Moneyball" concepts when they were first proffered - now it's easier to see why. You have a vested interest in the "old guard".)
I just applied methodology passed down over the years modified to fit circumstances dictated or modified by rules and regulations. Examples - changes in the hockey calendar year, bumps in age groupings, eligibility issues. Flattered that people think anything was my method.

Not what I did. Clearly stated Bantam / Midget / Junior which is far short of scouting/grading for the NHL. Basic objective was to get every deserving player a fair opportunity without sway from the politics of youth hockey. Usually there is a consensus of the top 10 Midget AAA players in a region. Getting all worthy players a chance to compete for the remaining spots is the challenge

My only interest is that hockey is better tomorrow than it was yesterday or today. That certain aspects of pre O6 hockey are coming back fascinates me. There are very positive programs in place since the last five years which have brought hockey in Quebec back to the private schools,a few public schools, colleges. Also the former pros including former NHLers getting involved in youth hockey is a step back to the future.No different than what was happening in the fifties and sixties.

Canadiens1958 is offline   Reply With Quote
Old
08-24-2012, 08:13 PM
  #106
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
So you aren't prepared to hold yourself to the same standards that you're grilling Iain with. I suppose it's easier that way.

Chalupa Batman is offline   Reply With Quote
Old
08-25-2012, 10:17 AM
  #107
Blue Blooded
Registered User
 
Blue Blooded's Avatar
 
Join Date: Oct 2010
Location: Handicap spot
Country: Sweden
Posts: 2,471
vCash: 504
I find the criticisms thrown at Iain's model in this thread are rather odd, why slate it for not having Thomas in the top 30 when he was a very late pick? If it should be criticized it should be for failing to recognize high picks that panned out like Yzerman being out of the top 30 in 1983 and Sundin being 22nd in 1989. In these cases the scouts obviously knew something that the stats couldn't pick up.

But if you read through the series it is obvious that these misses have been acceptable so far with the incredible roster put together without top 10 picks.

No updates since April worries me since I'm very interested in seeing how it works in an era where it could possibly have existed 1999-present. Will there be more updates on the Westerns?

Blue Blooded is offline   Reply With Quote
Old
08-25-2012, 06:30 PM
  #108
Master_Of_Districts
Registered User
 
Master_Of_Districts's Avatar
 
Join Date: Apr 2007
Location: Black Ruthenia
Country: Belarus
Posts: 1,746
vCash: 500
Quote:
Originally Posted by Ogopogo View Post
Eyes are biased and forgetful. Scoring stats have a large role to play in evaluation - eyes help too but they are not the be-all, end-all.
Interesting.

Is the same true for evaluating goaltending talent?

Master_Of_Districts is offline   Reply With Quote
Old
08-25-2012, 08:56 PM
  #109
Ogopogo*
 
Ogopogo*'s Avatar
 
Join Date: Apr 2005
Location: Edmonton
Country: Canada
Posts: 14,216
vCash: 500
Quote:
Originally Posted by Master_Of_Districts View Post
Interesting.

Is the same true for evaluating goaltending talent?
Perhaps but, stats are completely worthless evaluating goaltending talent so you better have the best eyes available.

Ever notice how drafting goaltenders is a complete roll of the dice? Even the best eyes can mess it up and the stats don't help at all.

Ogopogo* is offline   Reply With Quote
Old
08-26-2012, 09:36 AM
  #110
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Somewhat

Quote:
Originally Posted by Ogopogo View Post
Perhaps but, stats are completely worthless evaluating goaltending talent so you better have the best eyes available.

Ever notice how drafting goaltenders is a complete roll of the dice? Even the best eyes can mess it up and the stats don't help at all.
Somewhat accurate. All the stats available will not help when it comes down to recognizing fundamental flaws. Example a goalie whose skates leave the ice when descending into the butterfly.Goalies can get away with such flaws and have impressive junior or NCAA performances but they will be exposed by the much faster/mature NHL game.

Conversely scouts and GMs micro analyze goalie stats. Example SV% is broken down into provenance of shots in the defensive zone while goalies are broken down by type - low,medium, high, left, right, center etc. Then you have the combine, testing data, the the pre draft workout data etc.This data is used to evaluate and develop pre and post draft.

Canadiens1958 is offline   Reply With Quote
Old
08-26-2012, 11:35 AM
  #111
wgknestrick
Registered User
 
Join Date: Aug 2012
Posts: 1,474
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
Did it at the Bantam/Midget/Junior level for over 20 seasons. Still "birddog". Rather simple. Get the rosters. Watch the warm-up. Watch the game. Eliminate the non-prospects - weaker skaters for the level and so on down the line. Grade the rest. Quantify and qualify the prospects.
So what happens on the games you miss or are on the same day/time as another evaluation? Stat evaluation is not bound by logistics and evaluates every single player, on every single day, in every single country. This is something that is just not possible by the old guard.

wgknestrick is offline   Reply With Quote
Old
08-26-2012, 02:46 PM
  #112
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
What Level?

Quote:
Originally Posted by wgknestrick View Post
So what happens on the games you miss or are on the same day/time as another evaluation? Stat evaluation is not bound by logistics and evaluates every single player, on every single day, in every single country. This is something that is just not possible by the old guard.
What level are you writing about?

At the levels I was working at there were 2-3 people in each organization assigned to the task. If there was any doubt the player would be invited to the appropriate level for the next try-out camp. The way evaluations work missing a few games either due to player injury or evaluator absence does not matter. 36 game season,player graded out of /100 per game.So if a player is grading at 80 over 35 games, miss one and what is the worst that happens? He grades a 100 it becomes a 80.55. Balanced by the fact that the best and worst games are factored out so he is not penalized at all.

At the higher levels leading to the NHL entry or beyond you have access to the game films, organizational birddogs,independents who want to break into the loop, central scouting, combines,etc.
Players are not disadvantaged nor are the teams.

On the other hand fundamental flaws in a player's game are never caught by stats regardless of their sophistication or provenance.

Canadiens1958 is offline   Reply With Quote
Old
08-26-2012, 03:07 PM
  #113
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
On the other hand fundamental flaws in a player's game are never caught by stats regardless of their sophistication or provenance.
Scouting's perfect (even though we see draft mistakes all of the time)?

You seem to have all of the answers, but you're not willing to hold yourself to the standards that you insist on Iain. Why is that?

Chalupa Batman is offline   Reply With Quote
Old
08-26-2012, 03:37 PM
  #114
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Standards.

Quote:
Originally Posted by Taco MacArthur View Post
Scouting's perfect (even though we see draft mistakes all of the time)?

You seem to have all of the answers, but you're not willing to hold yourself to the standards that you insist on Iain. Why is that?
My standards were set by the various associations, teams, leagues and jurisdictional bodies that I was responsible to and those standards remain part of my background, reflected in my comments.

Iain's standards, while high, are different, like the difference between IIHF and NHL standards. Yet both governing bodies manage to work for the good of the game.

Draft mistakes have to be qualified. A distinction has to be made between a draft mistake, mistakes in the post draft development of the player, or systematic changes in the organization.

Drafting offensively talented players then hiring a coach like Ken Hitchcock will impact on the players offensive stats. Such factors have to be considered in the evaluation of a draft - eg comparing Mike Modano to Mats Sundin.

Sharing an AHL farm club with another NHL team impacts the development of young #1 goalies since if both NHL teams send down potential #1's then the playing time is split and neither develops #1 stamina if he is playing less than in junior.

Missing "Do Not Draft" indicators is a drafting mistake but I have yet to see stats recognize such flaws.


Last edited by Canadiens1958: 08-26-2012 at 03:39 PM. Reason: typo
Canadiens1958 is offline   Reply With Quote
Old
08-26-2012, 04:00 PM
  #115
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
My standards were set by the various associations, teams, leagues and jurisdictional bodies that I was responsible to and those standards remain part of my background, reflected in my comments.

Iain's standards, while high, are different, like the difference between IIHF and NHL standards. Yet both governing bodies manage to work for the good of the game.
All that, and yet you don't seem to be working "for the good of the game". All you're doing is cherry-picking numbers to fit your agenda, and lobbing bombs from your bunker.

The goal of this subforum is to see where we can go using analytics. With your background, I can see you being quite helpful towards that end (if you choose to be).

Chalupa Batman is offline   Reply With Quote
Old
08-26-2012, 08:18 PM
  #116
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,172
vCash: 500
Quote:
Originally Posted by overpass View Post
Looking at your article from 2009, it looks like the system would put him higher. Correct me if I'm reading this wrong, but his "Projectinator estimate in the last column (10YE)" is 0.94, or a close third behind John Tavares and Brayden Schenn, ahead of Duchene, Kane, Kadri, Glennie, and Cowen.
That was an early version of the system. The current version puts him at 7th I believe that season, which is essentially the same as the 11th he was drafted.

Quote:
Originally Posted by Canadiens1958 View Post
Your point seems to be that stats are a stand alone indicator that work well without any contribution from all the other components that at the arena scouting brings to the decision making process.
As I said, you're a piece of work. Please actually read what I write instead of what you would like me to be saying. I have stated explicitly on many occasions, both at other sites and in this very thread (in direct response to you even), statistical analysis can be used to improve "eyes only" evaluation, which was the point which began this discussion in the first place.

Stats used as a stand-alone indicator can, in fact, produce decent results. That is true. But I've never once said that they produce better results than scouting, or that they should be used instead of scouting. They should be used in conjunction with scouting. That's all I've ever said, and if you keep pretending that I haven't all I can do is point out that you're plainly and simply wrong.

Quote:
Originally Posted by Canadiens1958 View Post
The subpoint is that your stats package is underperforming compared to the NHL overall package.
What, the stats package developed by one guy using only commonly-available data underperforms what hundreds of professional scouts accomplish? No kidding. That should bloody well be expected. You're a piece of work.

Quote:
Originally Posted by Canadiens1958 View Post
Certain elements have merit from what I have observed - the age factor is given a bit too much weight.
Prove it, thoroughly and systematically.

Quote:
Originally Posted by Canadiens1958 View Post
At the Novice level the difference between a first year Jan and December birthday may be as much as 365 days or upwards of 20%.By the midget/junior levels the difference may still be as much as 365 days but the actual devleopmental difference is down to app 5-7%.
Wait, you mean at levels that the system is not designed to operate, it doesn't operate very well? You don't say. The purpose of the system is absolutely explicit: to evaluate 17- to 19-year-old players based on the numbers they record at that age. You're a piece of work.

Quote:
Originally Posted by Canadiens1958 View Post
Your appreciation of Quebec Midget AAA stats,Quebec/Eastern Ontario Junior A stats is not in tune with the league realities
I'd be interested in how you know that, considering that I haven't said anything about Midget AAA or Jr A stats? You're a piece of work.

I'd love to incorporate Junior A stats, and have done some work on the BCJHL. But most historical Jr A stats are not readily available, which makes model-building for them pretty difficult. For example, there is source I know of for complete Alberta Jr A stats, but charges a fee for them. Since I'm no professional, there is no way to justify paying for such stats.

Quote:
Originally Posted by Canadiens1958 View Post
You overrate Quebec/QMJHL goalies especially in the second half of the nineties onwards because the stats simply did not show how the game had changed at that position.
Goalie is the least reliable position to project based on stats. (And based on scouting).

How did the scouts do with goaltenders, say in 1998? The top 13 goalies drafted played an average of 23 NHL games each. The two drafted in the first round played 12 games total. You're still doing this wrong.

Quote:
Originally Posted by Canadiens1958 View Post
Pretty good does not cut it in the NHL. Access to better data? Interesting question. Change your approach to cultivating possible resources.
I'm not in the NHL. I'd love to have access to better data, and more time to analyze it. But I do have a life, and there's only so much time one can spend on such a thing if one is no being compensated for it, and one has a family to feed.

Quote:
Originally Posted by Taco MacArthur View Post
Let's play the same game with your method that you're playing with Iain's. How many of your "star" graded prospects have succeeded in the NHL? And how many have failed? And how many stars did you scout, but miss?

I'll of course ask for something published several years ago, and not just a retrospective thing that you whip up today.
Indeed. I'm not holding my breath.

Quote:
Originally Posted by Blue Blooded View Post
I find the criticisms thrown at Iain's model in this thread are rather odd, why slate it for not having Thomas in the top 30 when he was a very late pick? If it should be criticized it should be for failing to recognize high picks that panned out like Yzerman being out of the top 30 in 1983 and Sundin being 22nd in 1989. In these cases the scouts obviously knew something that the stats couldn't pick up.
Exactly. I'm wondering if some combination of the system's results and, say, the Central Scouting Rankings could be used to produce some more-accurate ranking. Something to consider.

Quote:
Originally Posted by Blue Blooded View Post
No updates since April worries me since I'm very interested in seeing how it works in an era where it could possibly have existed 1999-present. Will there be more updates on the Westerns?
Yep, just a matter of time. Assuming the site is still interesting in publishing them of course...

Quote:
Originally Posted by Ogopogo View Post
Ever notice how drafting goaltenders is a complete roll of the dice? Even the best eyes can mess it up and the stats don't help at all.
I wouldn't say that. They are, by far, the most difficult position to project, but it's not completely random and stat do help a bit, if you're careful with them.

Quote:
Originally Posted by Canadiens1958 View Post
Missing "Do Not Draft" indicators is a drafting mistake but I have yet to see stats recognize such flaws.
They do catch some that would have been very useful in, say, 1998. 6-foot-5 forward who scored 3 goals in his age-17 season? DO NOT DRAFT. At the very least, not in the first round.

Iain Fyffe is offline   Reply With Quote
Old
08-29-2012, 10:27 AM
  #117
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,172
vCash: 500
Dragging this discussion back to this thread where it belongs.

Quote:
Originally Posted by Canadiens1958 View Post
Merk Jankowski, 2012 NHL Entry Draft first round draft pick - Calgary,Stanstead College, born Sept 13, 1994 was Midget AAA eligible:
So your counter to my point that players from Midget AAA are not drafted in the NHL draft is to provide an example of a player who could have played in Midget AAA but did not?

The Projectinator only cares about what the player actually did on the ice, not what he could have done.

Quote:
Originally Posted by Canadiens1958 View Post
Perhaps it is time to review your Midget AAA views.

Point is that you present them as definitive counters to my points.
You cannot have it both ways.
Cannot have what both ways? I don't Projectinate players from the Australian league either. If I never do that means the system is technically incomplete from your point of view, even if no player from that league ever gets drafted by the professionals.

Same with Midget AAA. If that level becomes a source of Entry Draft players, then it will be Projectinated. Until such time, it falls outside the scope of the system.

I could certainly do the work on it, at least for Quebec where the data is available. But that would be a waste of time at the moment, since that league is not a source of players for the Entry Draft.

Iain Fyffe is offline   Reply With Quote
Old
08-29-2012, 12:02 PM
  #118
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Eligibility

Quote:
Originally Posted by Iain Fyffe View Post
Dragging this discussion back to this thread where it belongs.


So your counter to my point that players from Midget AAA are not drafted in the NHL draft is to provide an example of a player who could have played in Midget AAA but did not?

The Projectinator only cares about what the player actually did on the ice, not what he could have done.


Cannot have what both ways? I don't Projectinate players from the Australian league either. If I never do that means the system is technically incomplete from your point of view, even if no player from that league ever gets drafted by the professionals.

Same with Midget AAA. If that level becomes a source of Entry Draft players, then it will be Projectinated. Until such time, it falls outside the scope of the system.

I could certainly do the work on it, at least for Quebec where the data is available. But that would be a waste of time at the moment, since that league is not a source of players for the Entry Draft.
Point is that Quebec and other Midget AAA players are eligible if they are 17 year olds that meet NHL Entry draft criteria.

Similarly there are Jr A leagues in Canada that never had players drafted in the NHL Entry Draft but you wish to have the information.

Until Mark Jankowski was drafted, Stanstead College was not a source either nor was their league.

Canadiens1958 is offline   Reply With Quote
Old
08-29-2012, 12:10 PM
  #119
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,172
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
Point is that Quebec and other Midget AAA players are eligible if they are 17 year olds that meet NHL Entry draft criteria.
Point is that not covering a league which (effectively) never produces a player drafted in the NHL Entry Draft is not exactly a gaping hole in the system.

Quote:
Originally Posted by Canadiens1958 View Post
Similarly there are Jr A leagues in Canada that never had players drafted in the NHL Entry Draft but you wish to have the information.
No, I want the leagues that are important with respect to the Entry Draft, principally the western and Ontario leagues. Others may be desirable, depending on their history in the draft.

The system can, in theory, work with any junior league, it's just a matter of devoting the resources necessary to calibrate the calculations for that league. If the league is very unlikely to produce a draft-worthy player, then the limited time available is better spent on another league.

Quote:
Originally Posted by Canadiens1958 View Post
Until Mark Jankowski was drafted, Stanstead College was not a source either nor was their league.
There will always be outliers. There's no way around that.

Iain Fyffe is offline   Reply With Quote
Old
08-29-2012, 01:05 PM
  #120
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Ultimate Irony

Quote:
Originally Posted by Iain Fyffe View Post
Point is that not covering a league which (effectively) never produces a player drafted in the NHL Entry Draft is not exactly a gaping hole in the system.


No, I want the leagues that are important with respect to the Entry Draft, principally the western and Ontario leagues. Others may be desirable, depending on their history in the draft.

The system can, in theory, work with any junior league, it's just a matter of devoting the resources necessary to calibrate the calculations for that league. If the league is very unlikely to produce a draft-worthy player, then the limited time available is better spent on another league.


There will always be outliers
. There's no way around that.
The ultimate irony. You are now using all the elements of the "old guard" position when the Montreal Canadiens drafted John van Boxmeer in the 1st round, 14th overall in the 1972 NHL draft.

http://www.hockey-reference.com/play...vanbojo01.html

Conversely the progressive organizations, GMs, scouts, found ways to mine the Jr A leagues to the point where today they cannot be ignored at all.

Previously on this board, I posted about the growing shift to the private high schools in Quebec for hockey development.

Call it what you wish but progressive GMs, scouts, etc have to be pro-active but your efforts based on the bolded above is re-active.

Canadiens1958 is offline   Reply With Quote
Old
08-29-2012, 01:14 PM
  #121
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
The ultimate irony. You are now using all the elements of the "old guard" position when the Montreal Canadiens drafted John van Boxmeer in the 1st round, 14th overall in the 1972 NHL draft.

http://www.hockey-reference.com/play...vanbojo01.html
Your reaction leads me to believe that you aren't properly following along.

Iain is suggesting that it's not time-efficient to optimize his algorithms for every single possible input source to the NHL Draft. Although he didn't state it, I'd imagine that he would have a corollary that if he were solely using his method (*), he wouldn't draft players from those sources where it is exceptionally unlikely to be worth the effort.

(*) Which he's also stated that he wouldn't do in practice.

It's conceivable that a NHL superstar would be discovered out of the intramural leagues at UCLA. That doesn't mean that we need to use finite resources to incorporate the UCLA intramural leagues into Iain's methods.

Your example is actually just the opposite of what Iain is saying. Are you following?

Chalupa Batman is offline   Reply With Quote
Old
08-29-2012, 02:07 PM
  #122
Iain Fyffe
Hockey fact-checker
 
Iain Fyffe's Avatar
 
Join Date: Feb 2009
Location: Fredericton, NB
Country: Canada
Posts: 2,172
vCash: 500
Quote:
Originally Posted by Taco MacArthur View Post
Your reaction leads me to believe that you aren't properly following along.

Iain is suggesting that it's not time-efficient to optimize his algorithms for every single possible input source to the NHL Draft. Although he didn't state it, I'd imagine that he would have a corollary that if he were solely using his method (*), he wouldn't draft players from those sources where it is exceptionally unlikely to be worth the effort.

(*) Which he's also stated that he wouldn't do in practice.

It's conceivable that a NHL superstar would be discovered out of the intramural leagues at UCLA. That doesn't mean that we need to use finite resources to incorporate the UCLA intramural leagues into Iain's methods.
This is very well-put. If I had unlimited resources (or even a significant amount of resources) of course I would incorporate every possible source of players in the system. Being that I'm one guy doing this in his spare time, this simply isn't realistic, and if an NHL team were actually pursuing this line of analysis, they would be in a much better position than I am to capture as much data as possible.

Canadiens1958 is right to an extent, that the system has to be re-active in some sense, because it relies on historical data. If there is a shift towards one line of development that had previously been relatively untapped, this will be slow to be reflected in the system. Of course, it will also take years to know whether scouts are right to be focusing more on this new line of development. It's one thing to claim it's more important, it's another to demonstrate that a significant number of NHL-worthy players actually emerge from it.

However, even if you completely ignore a relatively minor source of players, you'll still do very well with such a system. Covering major junior, the NCAA and European leagues gives you the vast majority of draft-worthy prospects, and if you can add Junior A in there as well so much the better. Not ranking every single player is not a big deal; you'll miss the occasional player (just like the scouts do) [and this assumes you're only using such a system to draft, which I don't recommend] but will do well on the great majority of them.

Iain Fyffe is offline   Reply With Quote
Old
08-29-2012, 02:42 PM
  #123
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
The New Canadiens

Quote:
Originally Posted by Iain Fyffe View Post
This is very well-put. If I had unlimited resources (or even a significant amount of resources) of course I would incorporate every possible source of players in the system. Being that I'm one guy doing this in his spare time, this simply isn't realistic, and if an NHL team were actually pursuing this line of analysis, they would be in a much better position than I am to capture as much data as possible.

Canadiens1958 is right to an extent, that the system has to be re-active in some sense, because it relies on historical data. If there is a shift towards one line of development that had previously been relatively untapped, this will be slow to be reflected in the system. Of course, it will also take years to know whether scouts are right to be focusing more on this new line of development. It's one thing to claim it's more important, it's another to demonstrate that a significant number of NHL-worthy players actually emerge from it.

However, even if you completely ignore a relatively minor source of players, you'll still do very well with such a system. Covering major junior, the NCAA and European leagues gives you the vast majority of draft-worthy prospects, and if you can add Junior A in there as well so much the better. Not ranking every single player is not a big deal; you'll miss the occasional player (just like the scouts do) [and this assumes you're only using such a system to draft, which I don't recommend] but will do well on the great majority of them.
See the direction that the Canadiens are taking under Marc Bergevin:

http://www.montrealgazette.com/sport...344/story.html

Hiring of Donald Audette and the growing importance of the Quebec Midget AAA league:

http://www.montrealgazette.com/sport...852/story.html

The European countries and the USA have caught on as well. Likewise for other regions of Canada and the USA as well as the other NHL teams.

Canadiens1958 is offline   Reply With Quote
Old
08-29-2012, 02:44 PM
  #124
Chalupa Batman
Mod Supervisor
 
Chalupa Batman's Avatar
 
Join Date: Sep 2005
Posts: 21,945
vCash: 500
Can you explain the relevance to this discussion, or are you just randomly bolding sentences in quoted posts followed by links to whatever topic is top of mind to you?

Chalupa Batman is offline   Reply With Quote
Old
08-29-2012, 03:15 PM
  #125
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Relevance

Quote:
Originally Posted by Taco MacArthur View Post
Can you explain the relevance to this discussion, or are you just randomly bolding sentences in quoted posts followed by links to whatever topic is top of mind to you?
The relevance is that NHL teams have realized that the key to drafting is below the junior level. This is in sync with the old sponsorship days where teams would sign players, at times before their 16th birthday - Bobby Orr was 14. A 17/18 year old drafted out of the CHL either goes back to junior or wastes time in the NHL. A 17/18 year old drafted out of school, Midget AAA or the USHL, Jr A gives the drafting NHL team more developmental flexibility and time.

The key to modern NHL scouting is the early identification of viable players and tracking them for seasons before their year of NHL Entry Draft eligibility. Significant change from a generation ago.

Canadiens1958 is offline   Reply With Quote
Reply

Forum Jump


Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off



All times are GMT -5. The time now is 02:02 AM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2014 All Rights Reserved.