HFBoards

Go Back   HFBoards > General Hockey Discussion > The History of Hockey
The History of Hockey Relive great moments in hockey history and discuss how the game has changed over time.

Adjusted points

Reply
 
Thread Tools
Old
07-28-2010, 03:26 AM
  #101
Uncle Rotter
Registered User
 
Uncle Rotter's Avatar
 
Join Date: May 2010
Location: Kelowna, B.C.
Country: Canada
Posts: 1,713
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
Actually, there's a general trend that defies common sense. If you look at the PPO for and PPO against of teams, for some reason the better teams often have a mediocre or poor PPO/PPOA ratio, while some of the worst teams often have a good PPO/PPOA ratio.

It's not just the Oilers, look at the Islanders during their glory years. Now look at teams like the Oakland, Pitt and Kings in early days of expansion... or Washington and KC in their first seasons.

I think it's this simple: Good team gets out to a big lead over a bad team, so refs give the bad teams a couple extra power plays to hopefully make it a closer game (especially if home team is losing).
I think you're right. For example, 1981-82, 1982-83 & 1983-84 saw 3 Edmonton division rivals in the top 8 for PPO (LA, Vancouver, Winnipeg). LA was near the top of the league for most of the decade. From 81-82 thru 87-88 in terms of Power Plays Against Edmonton was 4th, 2nd, 2nd, 3rd, 3rd, 5th & 7th.

Uncle Rotter is online now   Reply With Quote
Old
07-28-2010, 03:29 AM
  #102
Uncle Rotter
Registered User
 
Uncle Rotter's Avatar
 
Join Date: May 2010
Location: Kelowna, B.C.
Country: Canada
Posts: 1,713
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
Actually, Gretzky along with the Oilers themselves were not an overly spectacular PP team.
Their league rankings for PP%:
1981-82 5th
1982-83 1st
1983-84 2nd
1984-85 2nd
1985-86 1st
1986-87 7th
1987-88 7th

They were good on the power play-they just didn't get very many of them!

Uncle Rotter is online now   Reply With Quote
Old
07-28-2010, 03:32 AM
  #103
Uncle Rotter
Registered User
 
Uncle Rotter's Avatar
 
Join Date: May 2010
Location: Kelowna, B.C.
Country: Canada
Posts: 1,713
vCash: 500
Quote:
Originally Posted by TheDevilMadeMe View Post

Maybe it had something to do with the fact that Edmonton loaded up on goons to protect their stars - so if the other team fouled, Edmonton made it a matching minor? A 4-on-4 would suit Edmonton just fine with Gretzky, Kurri, and Coffey all on the ice at once.
They changed the rule to eliminate 4-on-4s because of Edmonton. I can't remember if it was 1984 or 1985.

Uncle Rotter is online now   Reply With Quote
Old
07-28-2010, 12:24 PM
  #104
Infinite Vision*
 
Join Date: Apr 2008
Location: Ontario
Country: Canada
Posts: 2,862
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
Actually, Gretzky along with the Oilers themselves were not an overly spectacular PP team.Strictly imo, I always believed it was because they never had that big shot on the point making their PP a lil too one dimensional, relying on the down low plays too much.
I don't think it was a coincidence that Wayne's PP numbers got better with the arrival of Blake and his cannon with the Kings.
Gretzky just seemed to have more room with the Kings on the PP than he did with the Oilers.
Your the first person I've heard say that. Just because they didn't get many opportunites doesn't mean they weren't good, have you honestly watched many if any 80's Oilers games? Look at the stats above. Just as I thought their powerplay was one of the best, and rightfully so.

Infinite Vision* is offline   Reply With Quote
Old
07-28-2010, 12:40 PM
  #105
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 9,670
vCash: 500
Quote:
Originally Posted by Ryan87 View Post
Your the first person I've heard say that. Just because they didn't get many opportunites doesn't mean they weren't good, have you honestly watched many if any 80's Oilers games? Look at the stats above. Just as I thought their powerplay was one of the best, and rightfully so.
Yeah, I was just going on what I remember of them and their PP never seemed to be a focal point or something they relied on to win.
To be honest after seeing their lack of opportunities, I would have to agree that is more likely why it didn't seem as potent as say Calgary's was at the time and of course Big Al loading up from the point always tended to stick in your mind

Rhiessan71 is offline   Reply With Quote
Old
07-28-2010, 12:45 PM
  #106
BraveCanadian
Registered User
 
BraveCanadian's Avatar
 
Join Date: Jun 2010
Country: Canada
Posts: 6,857
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
Yeah, I was just going on what I remember of them and their PP never seemed to be a focal point or something they relied on to win.
To be honest after seeing their lack of opportunities, I would have to agree that is more likely why it didn't seem as potent as say Calgary's was at the time and of course Big Al loading up from the point always tended to stick in your mind
Yeah Calgary was really "known" for their powerplay.

I don't think you have much of a problem with Coffey as a point man though.. that guy didn't have a MacInnis rocket but apparently he could also score once in a while just the same.

BraveCanadian is offline   Reply With Quote
Old
07-29-2010, 04:44 PM
  #107
Sampe
Registered User
 
Join Date: Feb 2002
Location: Vantaa
Country: Finland
Posts: 2,648
vCash: 500
I like the top 10% rule in that it ignores lesser players and yet simultaneously makes the sample size fairly large in the modern era. But if we want to adjust stats across all eras, it would really harm the pre-1967 players: basically, the 46th-50th best scorers in a 30-team league are now supposed to be as good as the 10th best scorer in a 6-team league? Is there any reason to assume that expansions took place due to players getting more talented across the board?

I personally don't think so, wich why I would prefer to avoid samples based on a fixed percentage and rather use a small absolute sample size. This method isn't exactly perfect either, but I like it more as a starting point because:

1) We usually compare first lines in these discussions
2) There were only 18 first line forwards in the original six era
3) Some of those 18 were likely defensive-minded players and thus scored less than the best 2nd liners

One important point to consider however is the amount of American and European players in the NHL these days. They obviously affect the overall level of scoring whenever they are included in a sample size and since they didn't exist in the early decades of the NHL, I would personally eliminate their stats from the equation and only use Canadian players for adjustment. And then once the seasons have been adjusted, it's obviously possible to look at any first liner's stats regardless of their nationality and see how they compare across eras on paper.

Another point is the exceptional players and their stats. When the sample size is this small, they obviously have a huge effect on the overall scoring numbers. Thus I would definitely prefer to ignore the top scorers of each season. I don't have any tried and tested number available here, but eliminating the top 3 sounds too small to me so I would choose to ignore the entire top 5.

So to summarize:

1) I prefer taking the 6th-12th (or 6th-10th/15th/18th) best scorers among Canadian players from each season and counting their point totals to using a fixed percentage for adjusting the data. Without checking the stats I'm also pretty confident that every season in NHL history has featured at least this many Canadian first liners or more.

2) Adjusting points this way is nothing but a starting point; there are tons of other factors involved as more detailed research has showed.

3) I would love to use a larger sample size, but I have no idea how to apply it as fairly as possible.

Sampe is offline   Reply With Quote
Old
07-29-2010, 06:11 PM
  #108
matnor
Registered User
 
Join Date: Oct 2009
Location: Boston
Country: Sweden
Posts: 483
vCash: 500
Quote:
Originally Posted by Sampe View Post
1) We usually compare first lines in these discussions
2) There were only 18 first line forwards in the original six era
3) Some of those 18 were likely defensive-minded players and thus scored less than the best 2nd liners
This is exactly the reason why I chose the top 10% instead of a fixed number. The top 10% ensures that we basically get the first liners regardless of era. Taking a fixed number (say 50) we get even third liners from the O6-era which is not what we want. Maybe I misunderstood you?

Quote:
Originally Posted by Sampe View Post
One important point to consider however is the amount of American and European players in the NHL these days. They obviously affect the overall level of scoring whenever they are included in a sample size and since they didn't exist in the early decades of the NHL, I would personally eliminate their stats from the equation and only use Canadian players for adjustment. And then once the seasons have been adjusted, it's obviously possible to look at any first liner's stats regardless of their nationality and see how they compare across eras on paper.
I completely agree with only comparing players against Canadians in general (like comparing trophy-voting and top-10 finishes) which is somewhat of a proxy for the talent level. However, only using Canadians when adjusting is problematic, similar to the argument above, in that we risk including second- and third-liners. The fact that the expansion occured simultaneously with the euro-influx of players somewhat offset this but it's still a bit unsettling.


Quote:
Originally Posted by Sampe View Post
Another point is the exceptional players and their stats. When the sample size is this small, they obviously have a huge effect on the overall scoring numbers. Thus I would definitely prefer to ignore the top scorers of each season. I don't have any tried and tested number available here, but eliminating the top 3 sounds too small to me so I would choose to ignore the entire top 5.
The effect of outliers is almost negligible. When taking the average of around 40-50 players even an outlier like Gretzky has small importance. Nevertheless I would prefer to use the 10% percentile for this reason but chose top-10% instead to make it a bit more transparent. I doubt it makes any real difference though.

matnor is offline   Reply With Quote
Old
07-29-2010, 11:39 PM
  #109
Hardyvan123
tweet@HardyintheWack
 
Join Date: Jul 2010
Location: Vancouver
Country: Canada
Posts: 11,365
vCash: 500
Taking the 6th-12th stats from Canadian players still would not get around the problem that in modern times a lot of non Canadian players played in the league and where significant factors in scoring as oppose to a sample from the late 20 or 30's where those players did not affect the season at all. I just took a random year 2000-01 as I had it up for another purpose and 14 of the top 20 scorers that year where not Canadians. Taking those 14 guys out of any equation actually alters what happened in that particular season and I don't see the point in this.

If we take an earlier season in the 20's or 30's the top 20 guys overall would be Canadian with maybe a few exceptions of foreign born players but Canadian trained. How does adjusting for the 6th-12th Canadian players give us a better picture? We are still left with 2 entirely different comparisons here.

What's wrong with hockey reference.com, it's a fairly good starting point and very accessable. It takes into accounts various factors and is a good base, although it's not perfect, for comparing players from different eras stats wise.

Like it has been stated before, one needs to look into the reasons why any player did what they did stats wise in any year as well to give a more complete picture.

In the end any adjusted system would have it's weaknesses and this is why other factors have to be taken into consideration

Hardyvan123 is offline   Reply With Quote
Old
07-30-2010, 01:06 AM
  #110
Sampe
Registered User
 
Join Date: Feb 2002
Location: Vantaa
Country: Finland
Posts: 2,648
vCash: 500
Quote:
Originally Posted by matnor View Post
This is exactly the reason why I chose the top 10% instead of a fixed number. The top 10% ensures that we basically get the first liners regardless of era. Taking a fixed number (say 50) we get even third liners from the O6-era which is not what we want. Maybe I misunderstood you?
We do get first liners, but doesn't that also mean that we're comparing Howe's stats to, say, what the 12 most offensively gifted players of his era scored on average and Crosby's stats to the 60 most offensively gifted players of today? I would imagine that the first group is considerably tougher to match, unless you think that the 56th best scorer in today's NHL is just as naturally gifted as the 12th best scorer in the 1950s.

I would personally argue that the natural talent level of the elite Canadian players has not changed that much over the last 90 years or so. But back in the early decades they just didn't have a first line spot available for that 56th best guy.

Not sure if I understood your method correctly though.

Sampe is offline   Reply With Quote
Old
07-30-2010, 01:27 AM
  #111
Canadiens1958
Moderator
 
Canadiens1958's Avatar
 
Join Date: Nov 2007
Posts: 10,446
vCash: 500
Obvious Question

Quote:
Originally Posted by Sampe View Post
We do get first liners, but doesn't that also mean that we're comparing Howe's stats to, say, what the 12 most offensively gifted players of his era scored on average and Crosby's stats to the 60 most offensively gifted players of today? I would imagine that the first group is considerably tougher to match, unless you think that the 56th best scorer in today's NHL is just as naturally gifted as the 12th best scorer in the 1950s.

I would personally argue that the natural talent level of the elite Canadian players has not changed that much over the last 90 years or so. But back in the early decades they just didn't have a first line spot available for that 56th best guy.

Not sure if I understood your method correctly though.
Bolded. Then what about the natural talent level of the elite players from other hockey playing countries over the same period?

Canadiens1958 is offline   Reply With Quote
Old
07-30-2010, 03:23 AM
  #112
Sampe
Registered User
 
Join Date: Feb 2002
Location: Vantaa
Country: Finland
Posts: 2,648
vCash: 500
Quote:
Originally Posted by Canadiens1958 View Post
Bolded. Then what about the natural talent level of the elite players from other hockey playing countries over the same period?
I would imagine that it has gone up, given how much other countries struggled against Canada's amateur players in the early World Championships for instance. Hockey has always been popular in Canada so the elite players have always been extremely good. And thanks to the huge amount of players, generational talent differences are much smaller than in many of the the other big hockey countries. WW II era is an exception to the rule of course.

That said, population growth might have affected even Canada's numbers. And 'natural talent' can mean many things so you could argue that human being X is still an elite talent even if he will never even play hockey. But that's not what I mean here.

Sampe is offline   Reply With Quote
Old
07-30-2010, 04:49 AM
  #113
matnor
Registered User
 
Join Date: Oct 2009
Location: Boston
Country: Sweden
Posts: 483
vCash: 500
Quote:
Originally Posted by Sampe View Post
We do get first liners, but doesn't that also mean that we're comparing Howe's stats to, say, what the 12 most offensively gifted players of his era scored on average and Crosby's stats to the 60 most offensively gifted players of today? I would imagine that the first group is considerably tougher to match, unless you think that the 56th best scorer in today's NHL is just as naturally gifted as the 12th best scorer in the 1950s.

I would personally argue that the natural talent level of the elite Canadian players has not changed that much over the last 90 years or so. But back in the early decades they just didn't have a first line spot available for that 56th best guy.

Not sure if I understood your method correctly though.
If I would limit the number of players used to something like 10 players then I would agree with you. But if I use 50 players the ice time and the role of the bottom 25-something players would be significantly smaller in the O6-era.

matnor is offline   Reply With Quote
Old
07-30-2010, 04:54 AM
  #114
matnor
Registered User
 
Join Date: Oct 2009
Location: Boston
Country: Sweden
Posts: 483
vCash: 500
Quote:
Originally Posted by Sampe View Post
I would imagine that it has gone up, given how much other countries struggled against Canada's amateur players in the early World Championships for instance. Hockey has always been popular in Canada so the elite players have always been extremely good. And thanks to the huge amount of players, generational talent differences are much smaller than in many of the the other big hockey countries. WW II era is an exception to the rule of course.

That said, population growth might have affected even Canada's numbers. And 'natural talent' can mean many things so you could argue that human being X is still an elite talent even if he will never even play hockey. But that's not what I mean here.
This is an extremely important point which I feel should be discussed more on this board. I have no idea how many people played hockey in Canada in the 20s and 30s. Maybe there are some actual numbers out there that can be used to ***** the quality of the players playing in that era. It's clear that players like Shore and Morenz dominated their peers but to put them in an all-time perspective we would want to know how popular hockey was in Canada at the time. I have no idea and would like to hear from people that do know but I would guess that poor people just couldn't afford to play hockey at that time.

matnor is offline   Reply With Quote
Old
07-30-2010, 05:46 AM
  #115
Sampe
Registered User
 
Join Date: Feb 2002
Location: Vantaa
Country: Finland
Posts: 2,648
vCash: 500
Quote:
Originally Posted by matnor View Post
If I would limit the number of players used to something like 10 players then I would agree with you. But if I use 50 players the ice time and the role of the bottom 25-something players would be significantly smaller in the O6-era.
Yeah, I agree that you can't use a fixed number for the O6 era if you specifically want a large sample size. But does the larger sample size really help or does it just skew the results? I'm not entirely sure how you count the average scoring levels but the way I understood it is that you add the point totals of each player within that 10% and then divide the total by the number of players. To me that looks like you get a more talented 10% group (on average) from the O6 era than from later eras (even if we count European players) and thus Crosby & co will get an unfair advantage compared to Howe & co.

This of course is all based on my assumption that the 10 best scorers in the 50s were more gifted on average than the 50 best scorers today.

Sampe 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 09:31 PM.

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.