HFBoards

Go Back   HFBoards > General Hockey Discussion > The History of Hockey
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
Notices

The History of Hockey Relive great moments in hockey history and discuss how the game has changed over time.

Adjusted Points

Closed Thread
 
Thread Tools
Old
10-28-2010, 05:29 PM
  #1
habsfan87
Registered User
 
Join Date: Mar 2008
Posts: 129
vCash: 500
Adjusted Points

Just wondering, how exactly do you calculate "adjusted points"? I'm not that great at math, but I find it really interesting. I'd be especially interested in figuring out what a lot of guys '92-'93 season looked like.

habsfan87 is offline  
Old
10-28-2010, 11:11 PM
  #2
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
I think it's ([Length of that season]/82)([NHL total GPG in 2005-06]/[NHL total GPG that season]) * Number of points player had that season.

So it basically adjusts the exact percentage difference in the league's average scoring and the season's length.

So if each team played 41 games that year (it's based on number of games the league had that year, so it doesn't adjust for injuries) and the average team had half as many goals per game, you'd quadruple the points total. If the season has the same number of games and 50% more goals per game, your adjusted points total would be two thirds as much (1/1.5 = 0.67)


Last edited by Derick*: 10-28-2010 at 11:16 PM.
Derick* is offline  
Old
10-29-2010, 02:38 AM
  #3
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,837
vCash: 500
The "basic" formula is:

Adj. P = P * (82 / Games in season) * (X / League GPG in season)

Where X is the GPG level to which you are adjusting. X could be an historical average over many seasons, or any number really (often 6.0 or similar number is used). X doesn't matter in this case, because all players will be affected proportionately.

A more advanced formula adjusts goals and assists separately and then adds them to yield adjusted points. Goals follow the formula above, while league APG is often calculate by multiplying league GPG and league Assists per Goal (which has drifted up steadily over the past 50+ years from ~1.60 to as high as 1.75).

Hockeyreference factors in roster size and deducts an individual player's stats from league totals before calculating GPG and APG. I find the latter especially flawed, as it highly favors players who played when there were fewer teams and also distorts the differences between higher and lower scoring players.

Czech Your Math is offline  
Old
10-29-2010, 03:04 AM
  #4
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
I'd be interested in a system that just adjusts for the 18 highest scoring players' point totals instead of goals per game.

The current systems still favor modern players too much because the increase in powerplays means star players have a higher percentage of the goals, and more assists are recording, inflating point totals relative to goals scored.

I wonder if anyone can guess why I picked the number 18 specifically. I bet seventieslord will be able to.

Derick* is offline  
Old
10-29-2010, 03:13 AM
  #5
TheDevilMadeMe
Registered User
 
TheDevilMadeMe's Avatar
 
Join Date: Aug 2006
Location: Brooklyn
Country: United States
Posts: 45,407
vCash: 500
Quote:
Originally Posted by Cognition View Post
I'd be interested in a system that just adjusts for the 18 highest scoring players' point totals instead of goals per game.

The current systems still favor modern players too much because the increase in powerplays means star players have a higher percentage of the goals, and more assists are recording, inflating point totals relative to goals scored.

I wonder if anyone can guess why I picked the number 18 specifically. I bet seventieslord will be able to.
18 first line slots in the original 6 era?

Matnor modified "adjusted points" to the Top 10 scorers of the era fairly recently, and the results for top scorers was more intuitively correct than standard adjusted points. The problem with such a method, however, is that it is really only applicable to players who are actually near the top of the league in scoring.

TheDevilMadeMe is offline  
Old
10-29-2010, 03:17 AM
  #6
nik jr
Registered User
 
Join Date: Sep 2005
Country: Congo-Kinshasa
Posts: 10,798
vCash: 500
Quote:
Originally Posted by Cognition View Post
I'd be interested in a system that just adjusts for the 18 highest scoring players' point totals instead of goals per game.

The current systems still favor modern players too much because the increase in powerplays means star players have a higher percentage of the goals, and more assists are recording, inflating point totals relative to goals scored.

I wonder if anyone can guess why I picked the number 18 specifically. I bet seventieslord will be able to.
hockey outsider has tried adjusting scoring based on average of 2nd-4th in goals.

http://hfboards.com/showthread.php?t...justed+scoring

nik jr is offline  
Old
10-29-2010, 03:41 AM
  #7
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
Quote:
Originally Posted by TheDevilMadeMe View Post
18 first line slots in the original 6 era?

Matnor modified "adjusted points" to the Top 10 scorers of the era fairly recently, and the results for top scorers was more intuitively correct than standard adjusted points. The problem with such a method, however, is that it is really only applicable to players who are actually near the top of the league in scoring.
Well that's usually who were analyzing regardless, right?

Right. Looks like I overestimated the question I figure that way we have the biggest sample size possible that doesn't exceed the number of first line forward slots when the league was the smallest, unless it had less than six back in the super crazy old days (to use the technical term for that era).

Quote:
Originally Posted by nik jr View Post
hockey outsider has tried adjusting scoring based on average of 2nd-4th in goals.

http://hfboards.com/showthread.php?t...justed+scoring
Interesting system. I like the adjustment to 50 and 75, definitely gives it a more intuitive feel than the ratio does. I also think it's better to give the adjustment to concrete numbers than to adjust it to 2006, which seems arbitrary, although Thornton did get exactly 125 that year, which is a strange coincidence.

I think 2nd-4th is still too small though. I'd like to do one based on 3rd - 18th. There was never more than two transcendental scorers at any particular time so it's not going to be thrown off by competition as much. Presumably the lower we go down the list the less the scoring rate is affected by when we happen to have the most talented players and the more it's affected by era alone, but if we go too far down we're no longer looking at first line scorers, so 3rd - 18th seems like the ideal sample.

Other things that may help is dividing everything into eras of 3 - 8 years so we have a larger sample size and basing it on the highest scorer's PPG (with >##GP or total points of course) so an injury to a star player or a bunch of simultaneous cold years doesn't throw it off.

What really shook me in that thread are Gretzky's adjusted assists. We so much take it for granted that he's the best overall player and remember them as the high scoring 80s so much that we really forget that he was best at playmaking above all, and by the largest margin. No one else is even close.

For instance, in 2006, Thornton had more primary assists than anyone else had total assists, and led second place in assists by more than 25% (96 to 71). That's also more than 10 more than Crosby's highest assist total. He did the same type of thing in 2006-2007. That shows how dominant his playmaking ability was. Yet Gretzky's adjusted peak assists blows Thornton's away, 127 - 81.

If I'm understanding this correctly, if peak Gretzky played in this era, he'd be able to get more assists than Thornton/Crosby/Ovechkin/Sedin have points

Derick* is offline  
Old
10-29-2010, 03:58 AM
  #8
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
Adjusted Peak Assist Leaders 1968 - 2008
Gretzky: 127
Orr: 103
Mikita: 87
Esposito: 86
Lemieux: 83
Howe: 84
Thornton: 81
Jagr: 80

Based on the thread above. I'm surprised there weren't more players higher than 80 seeing as it's supposed to be based on 2nd - 4th being adjusted to 75.

Gretzky, then a big gap, then Orr, then a big gap. Encourages my opinion that Thornton's the best playmaker since Lemieux and Gretzky, though

Actually, I don't understand how those numbers are possible. In 2005 - 2006, the assists were Thornton 96, Spezza 71, Savard 69, Richards 68. That's an average of 69.something assists for 2nd-4th, so why is Thornton's adjusted peak assists lower than his actual? If it's based on 75 shouldn't his adjusted assists be higher than his actual? Or did I misunderstand it and "peak" is an average of several seasons, not a single season?


Last edited by Derick*: 10-29-2010 at 04:05 AM.
Derick* is offline  
Old
10-29-2010, 04:14 AM
  #9
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,837
vCash: 500
I think adjusting vs. the top 10 or top 18 (or top X) is a step in the right direction, but not the best solution. For one thing, it ignores that the quality of top scorers can vary quite a bit.

Intuitively, the quality of the top scorers was likely lower during and immediately after WWII and after the first expansion, and higher towards the end of the O6 era, after the WHA merger (especially after a few years w/o expansion) and with the major exodus of overseas talent to the NHL in the '90s (which was disproportionately composed of scoring forwards).

Also, keeping X constant over time ignores the affect of additional opportunity in a league with a greater number of teams.

While I don't believe there is a 100% perfect solution to this issue, it seems best to not use the very top scorers when measuring baseline scoring. However, improvement over using league GPG is still possible. Using the tier of players that are just below the very top scorers allows for a fairer and steadier baseline, since this talent pool will vary much less than that of the very top scorers.

I would suggest some combination of the following:

1. "2nd N" or "3rd N" players, which if talent was balanced between teams, would be the second and third top scorers on their respective teams. In a 6 team league, "2nd N" would be an average of the 7th-12th top scorers in the league, while "3rd N" would be an average of the 13th-18th top scorers. In a 30 team league, "2nd N" would be the top 31st-60th top scorers, while "3rd N" would be the top 61st-90th top scorers.

2. Next tier players of a constant number. For example, always taking the average of the 13th-24th or the 19th-36th top scorers, regardless of how many teams are in the league.

The two measures could be combined arithmetically, geometrically, harmonically, or in some other manner. Exactly which tiers to use and how to combine them is somewhat subjective, but I think someone with a practiced expertise in statistics may have some additional insight into how to combine the two measures.

Czech Your Math is offline  
Old
10-29-2010, 04:29 AM
  #10
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
Quote:
Originally Posted by Czech Your Math View Post
I think adjusting vs. the top 10 or top 18 (or top X) is a step in the right direction, but not the best solution. For one thing, it ignores that the quality of top scorers can vary quite a bit.

Intuitively, the quality of the top scorers was likely lower during and immediately after WWII and after the first expansion, and higher towards the end of the O6 era, after the WHA merger (especially after a few years w/o expansion) and with the major exodus of overseas talent to the NHL in the '90s (which was disproportionately composed of scoring forwards).

Also, keeping X constant over time ignores the affect of additional opportunity in a league with a greater number of teams.

While I don't believe there is a 100% perfect solution to this issue, it seems best to not use the very top scorers when measuring baseline scoring. However, improvement over using league GPG is still possible. Using the tier of players that are just below the very top scorers allows for a fairer and steadier baseline, since this talent pool will vary much less than that of the very top scorers.

I would suggest some combination of the following:

1. "2nd N" or "3rd N" players, which if talent was balanced between teams, would be the second and third top scorers on their respective teams. In a 6 team league, "2nd N" would be an average of the 7th-12th top scorers in the league, while "3rd N" would be an average of the 13th-18th top scorers. In a 30 team league, "2nd N" would be the top 31st-60th top scorers, while "3rd N" would be the top 61st-90th top scorers.

2. Next tier players of a constant number. For example, always taking the average of the 13th-24th or the 19th-36th top scorers, regardless of how many teams are in the league.

The two measures could be combined arithmetically, geometrically, harmonically, or in some other manner. Exactly which tiers to use and how to combine them is somewhat subjective, but I think someone with a practiced expertise in statistics may have some additional insight into how to combine the two measures.
If I understand you correctly, you're making the league size/talent pool equivocation fallacy. The league has five times as many teams but that doesn't mean the average player is five times as good. Presumably the players in the league at the time would have been equal to the top 20% of the players in the league today, as the top players at any particular time make the NHL and it's not just a random selection. We could suddenly decide to double the league size next year but the 31st - 60th best centers won't magically become better because they're now first liners, and all the AHL callups won't be better because now the NHL has let them join.

An increase in league size would probably increase the talent pool slightly because more players are called up and have the opportunity to improve, so someone like Sedin or Datsyuk may have never been "discovered" in the O6, but there's no way the talent increase would be close to proportionate to the league size increase, so you can't say 7th - 12th then equals 31st - 60th today.

When you start considering global talent pool size and the emigration of European players etc. the whole idea of adjusting for era really becomes a headache. It's impossible to know, especially mathematically, what's different because of era and what's different because there just happens to be more or less talent.

Derick* is offline  
Old
10-29-2010, 04:48 AM
  #11
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 11,326
vCash: 500
As I said in the other thread, Adjusted stats are all well and good for most players and will give you a ROUGH idea through the era's.

However, when you try and apply what is essentially an averaging formula to anomalies like Gretzky, Orr and Lemieux who are so far above even just the average of the top 5% it fails badly.

For example, it's already been presented in this thread that Gretzky's highest assist total would of been only 127 instead of 163....I'm sorry but that's ridiculous.

Not only can you not rely on the rest of the league to dictate their numbers but you would be better off just eliminating those guys from the equations before hand and you'll get much more accurate and believable results.

Take Gretzky's 183 point year for example, the next highest scorer had only 108 and after adjusted stats are applied, all those guys hovering around 100 points lose like 4-6 total each but somehow Gretzky loses like 30-40 points.
In what world does that seem remotely accurate or right?

Seriously, anyone that thinks Gretzky in his prime playing today couldn't at least get within ear shot of 200 points is fooling themselves imo.

Lemieux less than a decade ago after recovering from cancer, major back problems, at the height of the DPE at the age of 36 was still able to score at almost a 150 point pace.
Have adjusted stats explain that one.


Last edited by Rhiessan71: 10-29-2010 at 05:05 AM.
Rhiessan71 is offline  
Old
10-29-2010, 05:19 AM
  #12
habsfan87
Registered User
 
Join Date: Mar 2008
Posts: 129
vCash: 500
Thanks guys!

habsfan87 is offline  
Old
10-29-2010, 05:29 AM
  #13
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,837
vCash: 500
Quote:
Originally Posted by Cognition View Post
If I understand you correctly, you're making the league size/talent pool equivocation fallacy. The league has five times as many teams but that doesn't mean the average player is five times as good. Presumably the players in the league at the time would have been equal to the top 20% of the players in the league today, as the top players at any particular time make the NHL and it's not just a random selection. We could suddenly decide to double the league size next year but the 31st - 60th best centers won't magically become better because they're now first liners, and all the AHL callups won't be better because now the NHL has let them join.

An increase in league size would probably increase the talent pool slightly because more players are called up and have the opportunity to improve, so someone like Sedin or Datsyuk may have never been "discovered" in the O6, but there's no way the talent increase would be close to proportionate to the league size increase, so you can't say 7th - 12th then equals 31st - 60th today.

When you start considering global talent pool size and the emigration of European players etc. the whole idea of adjusting for era really becomes a headache. It's impossible to know, especially mathematically, what's different because of era and what's different because there just happens to be more or less talent.
As I said, there is no perfect solution. There are two primary factors at work, the top end talent pool and the opportunity available. Of the two, I would agree that the talent pool is generally much more important, but they both have significant effects on high end scoring. While I agree that the 31st-60th best players today are not likely as talented as the 7th-12th best players in the post-WWII O6 era, I would also assert that the best 7th-12th players today are better than were the 7th-12th players in the O6 era. Perhaps the 13th-18th or 19th-24th best forwards today are approximately equivalent to the 7th-12th best forwards in the O6 era, it's difficult to say with any certainty.

Using tiers of a fixed number of players outside the very top few scorers works very well until the '90s. At that time, there is a major increase in the talent pool due to overseas talent. Not only that, but the talent is disproportionately comprised of higher end forwards (first and second liners)... along with some defensemen and a few grinders and goalies. One would expect this to significantly increase the league scoring level, but instead the opposite occurred from the early '90s to the present (excepting brief spikes mostly attributable to more power plays).

Many mistakenly attribute the lack of goals from the mid-90's to present on lack of talent (i.e. Gretzky and Lemieux aren't playing in their prime together any more), when the opposite is actually true. The depth of offensive talent was greater than ever and the increase in such talent was much greater than that on the defensive side, so it should have resulted in increased league scoring. Until this paradox is recognized and resolved properly, it will be difficult to properly assess methods of fairly adjusting inidividual point production.

Czech Your Math is offline  
Old
10-29-2010, 05:42 AM
  #14
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,837
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
As I said in the other thread, Adjusted stats are all well and good for most players and will give you a ROUGH idea through the era's.

However, when you try and apply what is essentially an averaging formula to anomalies like Gretzky, Orr and Lemieux who are so far above even just the average of the top 5% it fails badly.

For example, it's already been presented in this thread that Gretzky's highest assist total would of been only 127 instead of 163....I'm sorry but that's ridiculous.

Not only can you not rely on the rest of the league to dictate their numbers but you would be better off just eliminating those guys from the equations before hand and you'll get much more accurate and believable results.

Take Gretzky's 183 point year for example, the next highest scorer had only 108 and after adjusted stats are applied, all those guys hovering around 100 points lose like 4-6 total each but somehow Gretzky loses like 30-40 points.
In what world does that seem remotely accurate or right?

Seriously, anyone that thinks Gretzky in his prime playing today couldn't at least get within ear shot of 200 points is fooling themselves imo.

Lemieux less than a decade ago after recovering from cancer, major back problems, at the height of the DPE at the age of 36 was still able to score at almost a 150 point pace.
Have adjusted stats explain that one.
I agree, it's best to factor out the very top players.

Lemieux is a truly exceptional case. Of course he had tremendous health challenges, but otherwise he had several factors working in his favor that year:

- He was perhaps the most naturally talented player ever
- He turned 35 around the start of the season, but had only played 715 regular season NHL games entering the season
- The games he did play were mostly during a less physical era
- He had 3 1/2 seasons off and so was quite well rested
- It was the only time besides most of the '97 season that he played on a line with Jagr

Granted, for anyone but a top 10 all-time talent, such a feat at such an age would have been just about impossible, but Lemieux was so talented, and relied (relatively) so little on practice and training, that to score at that pace for a half season, while truly remarkable, is not as miraculous as it would be for almost any other player in hockey history. I don't mean to diminish what he did, it's just that he did similar things before, so even at 35 it was less surprising than it should have been. In adjusted terms, his PPG was the 4th best of his career, about as close to his 1st-3rd best seasons as his 5th-8th best, still quite amazing. However, it's also important to remember that he wasn't usually healthy during his prime either, and may have actually been healthier in 2001 than he was (on average) during his prime.


Last edited by Czech Your Math: 10-29-2010 at 05:55 AM.
Czech Your Math is offline  
Old
10-29-2010, 05:50 AM
  #15
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
Quote:
Originally Posted by Rhiessan71 View Post
As I said in the other thread, Adjusted stats are all well and good for most players and will give you a ROUGH idea through the era's.

However, when you try and apply what is essentially an averaging formula to anomalies like Gretzky, Orr and Lemieux who are so far above even just the average of the top 5% it fails badly.

For example, it's already been presented in this thread that Gretzky's highest assist total would of been only 127 instead of 163....I'm sorry but that's ridiculous.

Not only can you not rely on the rest of the league to dictate their numbers but you would be better off just eliminating those guys from the equations before hand and you'll get much more accurate and believable results.

Take Gretzky's 183 point year for example, the next highest scorer had only 108 and after adjusted stats are applied, all those guys hovering around 100 points lose like 4-6 total each but somehow Gretzky loses like 30-40 points.
In what world does that seem remotely accurate or right?

Seriously, anyone that thinks Gretzky in his prime playing today couldn't at least get within ear shot of 200 points is fooling themselves imo.

Lemieux less than a decade ago after recovering from cancer, major back problems, at the height of the DPE at the age of 36 was still able to score at almost a 150 point pace.
Have adjusted stats explain that one.
That hasn't been my experience with adjusted points. If it's working the way it's described they should all lose the same percentage. Gretzky should be losing about 30 and they should be losing about 15 - 20.

Seeing as he barely passed it in the 80s, I don't think he's be very close to 200. There's a lot of room between today's top scorers (110 - 120) and 200. I imagine he'd be somewhere between 140 - 160, which is still well ahead of today's stars.

What year are you referring to that 150 point pace and over how many games?

Derick* is offline  
Old
10-29-2010, 06:30 AM
  #16
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 11,326
vCash: 500
Quote:
Originally Posted by Cognition View Post
That hasn't been my experience with adjusted points. If it's working the way it's described they should all lose the same percentage. Gretzky should be losing about 30 and they should be losing about 15 - 20.

Seeing as he barely passed it in the 80s, I don't think he's be very close to 200. There's a lot of room between today's top scorers (110 - 120) and 200. I imagine he'd be somewhere between 140 - 160, which is still well ahead of today's stars.

What year are you referring to that 150 point pace and over how many games?
2000/2001 43GP 35G 41A 76Points
Debate over whether he would of kept that pace for the whole season all you want but the fact still remains that no one in the game today comes remotely close to those numbers, let alone doing it in the middle of the Dead Puck Era at the age of 36 to boot.


Back to the Gretzky totals there.
By saying everyone else should lose about 15-20 points, you are saying that not a single player would of even broke 90 points other than Gretzky that year.
C'mon, at some point common sense has to take over here.
Adjusted stats are a rough tool, nothing more and I for one will never take anyone using them at verbatim seriously.

Not that long ago someone tried to assert that Forsberg would of been a 50+ goals 100+ assist player in the 80's because of adjusted stats.
Long story short, that assertion has its' ass handed to it.

Tool to support an opinion sure but they should not at any time be used AS the opinion.
At least not on the History boards

Rhiessan71 is offline  
Old
10-29-2010, 06:42 AM
  #17
Derick*
 
Join Date: Nov 2009
Location: Toronto
Country: Canada
Posts: 9,624
vCash: 500
Send a message via MSN to Derick*
Let me stipulate that I definitely think that's very impressive, but: that's an 145 point pace, not a 150 point pace; it's barely over half a season, so it's very unlikely he would have kept it; 2001 wasn't the middle of the DPE. That was the last year before it got really bad, the three following it were the real climax. For instance, in 2003 iirc no one had more than 90-95 points, but in 2001 Sakic managed 118.

All that being said, I imagine if he'd played the full season he would have been around first in the league (Jagr had 121, Sakic had 118) and considering his age that's pretty fantastic.

Was there really a year Gretzky led 183 to 105? That doesn't sound right to me.

Anyway, this is a big long debate so let me focus on the one part I took exception to "anyone that thinks Gretzky in his prime playing today couldn't at least get within ear shot of 200 points is fooling themselves imo."

Throw all our algorithms and adjusted whatever out the window, I still don't think he would have much of a chance of hitting 200. I don't know what your definition of "earshot" is but considering that he only hit 215, you're basically saying anyone who thinks his scoring would go down more than 7% is kidding themselves.

Derick* is offline  
Old
10-29-2010, 07:05 AM
  #18
Rhiessan71
Just a Fool
 
Rhiessan71's Avatar
 
Join Date: Feb 2003
Location: Guelph, Ont
Country: Canada
Posts: 11,326
vCash: 500
Quote:
Originally Posted by Cognition View Post
Let me stipulate that I definitely think that's very impressive, but: that's an 145 point pace, not a 150 point pace; it's barely over half a season, so it's very unlikely he would have kept it; 2001 wasn't the middle of the DPE. That was the last year before it got really bad, the three following it were the real climax. For instance, in 2003 iirc no one had more than 90-95 points, but in 2001 Sakic managed 118.

All that being said, I imagine if he'd played the full season he would have been around first in the league (Jagr had 121, Sakic had 118) and considering his age that's pretty fantastic.

Was there really a year Gretzky led 183 to 105? That doesn't sound right to me.

Anyway, this is a big long debate so let me focus on the one part I took exception to "anyone that thinks Gretzky in his prime playing today couldn't at least get within ear shot of 200 points is fooling themselves imo."

Throw all our algorithms and adjusted whatever out the window, I still don't think he would have much of a chance of hitting 200. I don't know what your definition of "earshot" is but considering that he only hit 215, you're basically saying anyone who thinks his scoring would go down more than 7% is kidding themselves.
It was 86/87
Gretzky 183
Kurri 108
Lemiuex 107
Messier 107
Gilmour 105
Ciccarelli 103
Hawerchuck 100
Goulet 96
Kerr 95
Bourque 95


As far as what Gretzky would drop down by, I don't think it would be more than 10%, 15% at the most.
Still closer to 190/200 than to the 160/170 adjusted stats suggests.

I just find it ridiculous that in trying to adjust Gretzky's 215 point season that he is somehow docked 50-60 points while the other top scorers only lose 10-20.
It just doesn't make sense, he loses more points than the next 3-4 top scorers combined...silly.


Last edited by Rhiessan71: 10-29-2010 at 08:01 AM.
Rhiessan71 is offline  
Old
10-29-2010, 08:07 AM
  #19
BraveCanadian
Registered User
 
BraveCanadian's Avatar
 
Join Date: Jun 2010
Country: Canada
Posts: 10,770
vCash: 500
In my opinion the problem with adjusted stats is the same it has always been:


1 - The top players in the league (which are normally the ones we are concerned with) are not average yet we use an average to describe their play.

This is why you see crazy results when looking at some of the adjusted stats. Things that are obviously ridiculous or don't mesh with what your eyes see or saw for real.

2 - Secondly, no factor for relative team strength or even very strong teammates is considered.


As TDMM mentioned, Matnor put some work into a system using the top scorers to predict the top scorers and obviously there are limitations to that as well but it at least looked reasonable at first glance to me.

There just plain won't be a system that doesn't have problems because there are too many variables at work over time to boil the issue down into a simple calculation.

BraveCanadian is offline  
Old
10-29-2010, 08:22 AM
  #20
Hardyvan123
tweet@HardyintheWack
 
Join Date: Jul 2010
Location: Vancouver
Country: Canada
Posts: 17,558
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
The "basic" formula is:

Adj. P = P * (82 / Games in season) * (X / League GPG in season)

Where X is the GPG level to which you are adjusting. X could be an historical average over many seasons, or any number really (often 6.0 or similar number is used). X doesn't matter in this case, because all players will be affected proportionately.

A more advanced formula adjusts goals and assists separately and then adds them to yield adjusted points. Goals follow the formula above, while league APG is often calculate by multiplying league GPG and league Assists per Goal (which has drifted up steadily over the past 50+ years from ~1.60 to as high as 1.75).

Hockeyreference factors in roster size and deducts an individual player's stats from league totals before calculating GPG and APG. I find the latter especially flawed, as it highly favors players who played when there were fewer teams and also distorts the differences between higher and lower scoring players.
Your above points are true and it's not a perfect system but it's a pretty decent starting point or baseline when comparing players and their seasons from different eras or seasons.

At the end of the day relying on any one system has it's obvious flaws and we should take into account as much information and data when evaluating any players IMO.

Hardyvan123 is offline  
Old
10-29-2010, 08:28 AM
  #21
Infinite Vision*
 
Join Date: Apr 2008
Location: Ontario
Country: Canada
Posts: 2,861
vCash: 500
Quote:
Originally Posted by Rhiessan71 View Post
It was 86/87
Gretzky 183
Kurri 108
Lemiuex 107
Messier 107
Gilmour 105
Ciccarelli 103
Hawerchuck 100
Goulet 96
Kerr 95
Bourque 95


As far as what Gretzky would drop down by, I don't think it would be more than 10%, 15% at the most.
Still closer to 190/200 than to the 160/170 adjusted stats suggests.

I just find it ridiculous that in trying to adjust Gretzky's 215 point season that he is somehow docked 50-60 points while the other top scorers only lose 10-20.
It just doesn't make sense, he loses more points than the next 3-4 top scorers combined...silly.
So wait a minute, if you disagree with how much Gretzky goes down by and think it's absolutely ******** if people agree to this, why is it so crazy to believe Forsberg's adjusted stats. The adjusted stats are made to compare talent across era's, not deduct from what previous greats of the game did. When will people understand this? Your the guy that think's hockey doesn't change overtime and Gretzky could jump exactly as he was from the 80's and dominate today. You sir, haven't got a clue I can tell you that.

Infinite Vision* is offline  
Old
10-29-2010, 08:29 AM
  #22
Infinite Vision*
 
Join Date: Apr 2008
Location: Ontario
Country: Canada
Posts: 2,861
vCash: 500
Why doesn't everyone just be honest. Their problem's with adjusted stats are player's from older/higher scoring eras lose points. That's it.

Infinite Vision* is offline  
Old
10-29-2010, 08:53 AM
  #23
BraveCanadian
Registered User
 
BraveCanadian's Avatar
 
Join Date: Jun 2010
Country: Canada
Posts: 10,770
vCash: 500
Quote:
Originally Posted by Infinite Vision View Post
Why doesn't everyone just be honest. Their problem's with adjusted stats are player's from older/higher scoring eras lose points. That's it.
Lets get this out of the way then.. how many points would your avatar have in 82?

BraveCanadian is offline  
Old
10-29-2010, 10:40 AM
  #24
Infinite Vision*
 
Join Date: Apr 2008
Location: Ontario
Country: Canada
Posts: 2,861
vCash: 500
Quote:
Originally Posted by BraveCanadian View Post
Lets get this out of the way then.. how many points would your avatar have in 82?
You guys still don't get it. It's about comparing talent across eras, not who would have scored this much in whatever year or whatever era. It seems like a lot of you think that adjusted stat backers are trying to steal points from 80's players. That's not the case. Hockeyreference.com uses the historical average, for comparing all players in the history of the game. Adjusting it to any particular season would only change the numbers not the order of which a player would place above another. I'm already aware that 80's players get a bit slighted compared to 90's and 70's players, but not by too much so it certainly doesn't stop me from believing anyone from the 80's besides Gretzky and Lemieux would be outscored by a healthy Forsberg more often than not.


Last edited by Infinite Vision*: 10-29-2010 at 10:52 AM.
Infinite Vision* is offline  
Old
10-29-2010, 11:03 AM
  #25
Infinite Vision*
 
Join Date: Apr 2008
Location: Ontario
Country: Canada
Posts: 2,861
vCash: 500
Also, hockey in the 90's made a huge leap in terms of overall quality and quantity of players due to a very increased talent pool. Czech your math has very intelligently gone into detail about it in many threads so there's nothing I can say about this to convince you that he hasn't said already. So if a player from the 90's has a bit of an edge on an 80's or 70's player in adjusted stats, I have no problems thinking the 90's player is better.

Infinite Vision* is offline  
Closed Thread

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 03:55 PM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"

vBulletin Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2016 All Rights Reserved.