News Article: The Peak and Decline of Elite NHL Forwards
- Thread starter Appleyard
- Start date
Appleyard
Registered User
Appleyard
Registered User
The website and therefore the article are back up now and available to read.
Hope you enjoy it!
I got the idea looking at previous studies done on this are. However, when I was looking through the studies that had been done, I noticed that:
A. There had been relatively few that focus on forwards.
B. There had been few of these that only focus on very high tier forwards.
C. These studies had generally looked at Raw points instead of PPG.
D. None had seemingly taken era into account, so for example Jagr's 2005-06 season ranks similarly to 2000-01.
E. There had not been much in terms of dividing goals, assists and points.
So I figured I would change that!
Hope you enjoy it!
I got the idea looking at previous studies done on this are. However, when I was looking through the studies that had been done, I noticed that:
A. There had been relatively few that focus on forwards.
B. There had been few of these that only focus on very high tier forwards.
C. These studies had generally looked at Raw points instead of PPG.
D. None had seemingly taken era into account, so for example Jagr's 2005-06 season ranks similarly to 2000-01.
E. There had not been much in terms of dividing goals, assists and points.
So I figured I would change that!
- Sep 28, 2014
- 76,697
- 123,242
That was a fantastic read! Thank you for the hard work you put into it. It should calm down the folks who think the Flyers should just trade Giroux now since by the team they are contenders, he will be 30-32 and "completely washed up" lol
Appleyard
Registered User
He is pretty much the only exception in terms of dropping of that far (not counting retirements!)
His age 33 season is the 4th worst of the 45 players in the sample who played age 33. (Lindros, Selanne (hey there's hope! ;p) and Heatley are worse...)
His age 34 season is the worst of the 34 players in the sample who played age 34. (Owen Nolan and Brad Richards were 2nd and 3rd worst... with 44 adjusted points to Vinnys 32.)
Of all 50 guys in the sample who played both their age 33 and 34 (32 players, discrepancy to age 34 is due to lockout, guys sitting out years etc) seasons his adjusted points in that time are the combined worst.
Bottom 5 Adjusted Points age 33-34:
Lecavalier: 82 points
B. Richards: 101 points
Bertuzzi: 103 points
Turgeon: 119 points
Naslund: 121 points
He is not even close to the others who played until his age! (Amonte and Nolan would also be in this list if they had not missed years to lockout etc... but they would both almost certainly have more points than Vinny going off the years around them... Nolan probably would be at ~90 and Amonte ~105)
His age 34 seasons would also be the worst age 35 season (Nolan, 37 points adjusted is worst atm) and in terms of age 36 seasons only Roenick and Amonte (both 30 points adjusted) had worse than Vinnys age 34 year. (32 points adjusted)
Last edited:
Peak will be different from player to player and organisation to organisation.
Since the cap was put in place, the peak of players (on avg) has been lower because players dont have to chase big contracts at 31 years old. Its given to them at 25 or so. Players that are not given a bridge contract will peak earlier.
Its not a coincidence that peak has been lower since UFA age was lowered 4 years.
Im not saying everyone dogs it after getting a big contract, but priorities change with maturity and guaranteed money.
Since the cap was put in place, the peak of players (on avg) has been lower because players dont have to chase big contracts at 31 years old. Its given to them at 25 or so. Players that are not given a bridge contract will peak earlier.
Its not a coincidence that peak has been lower since UFA age was lowered 4 years.
Im not saying everyone dogs it after getting a big contract, but priorities change with maturity and guaranteed money.
Beef Invictus
Revolutionary Positivity
Appleyard
Registered User
Of course it will be, hence why a median and mean is used, across the 50 man sample there were peak years from age 22 at youngest to 37 at oldest! (age 29 had the most pure peaks, with 9. age 26 was 2nd with 7. age 28 was 3rd with 6)
If we look at the guys in this sample who played in their prime years after the lockout here are their prime ages using the measures I did in this study:
H. Sedin: 29
D. Sedin: 29
Datsyuk: 29
B. Richards: 29
Hossa : 28
Briere: 28
Zetterberg: 27
Savard: 26
Lecavalier: 26
Heatley: 26
Thornton: 26
Mean: 27.55
So it does seem the adjusted PPG that the peaks of elite forwards is similar on either side of the lockout, as this study featured guys born in the late 60's, whose prime years were in the mid-90's, to guys born in the early 80's whose primes were in the few years after the lockout.
Though ofc in ten years time when elite forwards from the post lockout era are looked at it may be closer to 26 instead of 27... but I cannot imagine it dropping below 26 really.
I think one of the reasons it appears to be earlier since the lockout it the disparity between the amount of goals per game in the league... which has been declining since the lockout. The same as why Jagr's 2005-06 season when measured in raw numbers is as good as his 2000-01 season... when it really was not. I mean, Nash's season this year (age 30) was as good adjusted as his 2008-09 (age 24) season where he got 10 more raw points.
Tripod
I hate this team
Appleyard
Registered User
Also figured I would rationalise this in relation to Giroux:
Giroux era adjusted pace seasons:
Age 21: 18g|35a|53p
Age 22: 17g|32a|49p
Age 23: 28g|55a|83p
Age 24: 34g|76a|110p
Age 25: 25g|66a|91p
Age 26: 32g|64a|94p
Age 27: 28g|53a|82p
In general:
At age 34 goals are at 69% of peak
At age 34 assists are at 84% of peak
So if we assume that Giroux peak was 34g and 76a adjusted, the results for his age 34 season would be:
23g|64a|87p (adjusted) Which equates to ~20g|59a|79p in today's scoring levels.
Or if we actually assume that 110p adjusted was a lucky aberration, and that his actual peak in terms of ability was more realistically like the 32-64-96 adjusted:
22g|54a|76p (adjusted) Which equates to ~20g|48p|68p in today's scoring levels.
Even if you went off last season (with his crazy low on ice sh%) his age 34 season would be:
19g|45a|64p (adjusted) Which equates to ~17g|41a|58p in today's scoring levels.
Giroux era adjusted pace seasons:
Age 21: 18g|35a|53p
Age 22: 17g|32a|49p
Age 23: 28g|55a|83p
Age 24: 34g|76a|110p
Age 25: 25g|66a|91p
Age 26: 32g|64a|94p
Age 27: 28g|53a|82p
In general:
At age 34 goals are at 69% of peak
At age 34 assists are at 84% of peak
So if we assume that Giroux peak was 34g and 76a adjusted, the results for his age 34 season would be:
23g|64a|87p (adjusted) Which equates to ~20g|59a|79p in today's scoring levels.
Or if we actually assume that 110p adjusted was a lucky aberration, and that his actual peak in terms of ability was more realistically like the 32-64-96 adjusted:
22g|54a|76p (adjusted) Which equates to ~20g|48p|68p in today's scoring levels.
Even if you went off last season (with his crazy low on ice sh%) his age 34 season would be:
19g|45a|64p (adjusted) Which equates to ~17g|41a|58p in today's scoring levels.
- May 22, 2008
- 9,843
- 2,169
Great article. I always value your posts. It seems that players generally peak right around the time that they are becoming UFA eligible. I wonder if the fact that they tend to begin their decline soon after that point has anything to do with the fact that they have often just signed a nice new contract or reached the age where family begins to take on greater importance rather than any real physical breakdown.
cheesesteakarmor
Registered User
- Jul 18, 2009
- 2,692
- 0
Alex, are you able to account for injuries in your analysis? Not in terms of games missed, but rather salient lingering injuries (e.g., groin, knee, and concussions) and the lagged effects thereof?
More regression oriented than descriptive, but it would likely explain timing and decline?
You might also consider assessing lambda, only using goals and assists instead of crimes
Addendum, can we also account for linemates in these analyses?
More regression oriented than descriptive, but it would likely explain timing and decline?
You might also consider assessing lambda, only using goals and assists instead of crimes
Addendum, can we also account for linemates in these analyses?
Appleyard
Registered User
Alex, are you able to account for injuries in your analysis? Not in terms of games missed, but rather salient lingering injuries (e.g., groin, knee, and concussions) and the lagged effects thereof?
More regression oriented than descriptive, but it would likely explain timing and decline?
You might also consider assessing lambda, only using goals and assists instead of crimes
Addendum, can we also account for linemates in these analyses?
I did not, but I do think I could try and account for such injuries in a bigger study with more players. (I think 50 is probably too small a sample to properly look at such injuries) I think my initial thoughts on the way I would do it is try and classify players in terms of different levels of injuries... so a group that were almost injury free, a group that were plagued by injuries and one in the middle. Then attempt to examine the differences in decline between them in two ways. Firstly simply the average peak and decline of the groups compared to each other, but also their decline vs their expected decline without injury as extracted from the 'injury free' group. Not perfect but with such a variable it may be the best way possible.
I think anything in terms of the NHL from before 2007-08 is very tricky in terms of accounting for linemates... the strength of the team itself is partially addressed in the way that hockey-ref calculates their adjusted points, but without the more in depth stats that have started in the last 8 years it would be tricky. I mean, it is obvious in some players cases that linemates did increase production... but hard to determine by how much without P/60 or PPG stats with or without the individual linemates over seasons and careers.
I also think other things such as 'luck' have an influence. The individuals on ice sh% and % of points vs goals on ice for should really be examined and standardised vs the individual players averages. (for example Giroux this year had an unsustainably low on ice sh%, which cost him ~8-10 points through what was arguably bad luck.)
Appleyard made it to Yahoo! https://ca.sports.yahoo.com/blogs/nhl-puck-daddy/3-on-3-ot-preseason-test--visnovsky--crabb-get-ptos--disney-blocking-d4---puck-headlines-183405017.html
Last edited:
Appleyard
Registered User
Haha, not ashamed to say I think that is pretty cool. Thanks for letting me know! Almost as good as that time Meltzer mentioned me.
CodyTheHuman
Registered User
deadhead
Registered User
- Feb 26, 2014
- 49,215
- 21,617
Here's an idea for an update.
Look how over 30 forwards perform in the playoffs.
Reason is coaches may reduce their minutes as they age, to keep them fresh for the playoffs where their experience in playing under pressure has more value than the regular season. So does their playoff performance show the same decline as regular season performance?
Look how over 30 forwards perform in the playoffs.
Reason is coaches may reduce their minutes as they age, to keep them fresh for the playoffs where their experience in playing under pressure has more value than the regular season. So does their playoff performance show the same decline as regular season performance?
Ad
Latest posts
-
-
-
-
Post-Series Talk: Colorado Avalanche defeat the Winnipeg Jets, 4-1- Latest: Praetorian Caps
