HFBoards

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

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.

Top 3 Adjusted Years

View Poll Results: How many games constitutes a 'season' rather than a 'pace'
35 0 0%
40 0 0%
45 0 0%
50 1 14.29%
55 2 28.57%
60 3 42.86%
65 1 14.29%
70 0 0%
Voters: 7. You may not vote on this poll

Closed Thread
 
Thread Tools
Old
08-12-2012, 11:22 AM
  #26
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
Quote:
Originally Posted by solidmotion View Post
Just slipped that in there eh... assuming Bernie Nicholls?
haha, i forgot about that! Way to notice, ya - that's Bernie. He can thank 99 for making the list.

tombombadil is offline  
Old
08-12-2012, 01:40 PM
  #27
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
Tom,

I don't think there is an exact point where a pace becomes a season. A season is a season and a pace is a pace. A pace becomes more reliable as the sample size increases, but it never exactly equates to a season. Also, I believe the reliability increases if the pace is similar to other seasons the player had. The more full seasons which are close to the pace for that season, the more reliable the pace would seem to be.

I voted for 55 based on this logic: At that point, the games missed become 1/2 or less of the games played, so the reliability has improved significantly. That's an intuitive "inflection point", if you will, but may be wrong. Still, I stand by the points made in the first paragraph, the two are never really the same.

Before I start working on this, any suggestions for a total game threshold over multiple seasons? Maybe a couple numbers between 200 and 400? If not, I'll pick at least one number that doesn't appear to affect either Wayne or Mario unfairly.

Czech Your Math is offline  
Old
08-12-2012, 01:44 PM
  #28
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
I don't know if 'sliding scale' is a term that applies to this, but I crunched the top 10 guys from the original list into a 5 year list with a minimum of 70 games played. However, based on extraordinary performances, some guys had better years in total adjusted points in less than 70 games than they did in 'full' seasons. Add to that that Crosby and Malkin actually don't have 5 'full' seasons. So, gave zero credit for games missed up to 70, then took the ratio 82/70 (or 80/70, 84/70 depending on year) and used that to pro rate the season. so, if a player had 98 points in 63 games, it would count as 98/70, thus 82/70 x 98 = whatever.

I feel this doubly penalizes for injury, but allows everyone to participate, and gives credit to great seasons. After crunching all these, I feel strongly that a more balanced method would be to take their point per game based on what they acheived, yet only apply it to 12 phantom games (ten in the 80's, 14 in those two annoying years) Anyways, Iam not going to re-crunch, as my wife is looking at me like I am a complete waste of time!

1. Gretzky - 828
2. Lemieux - 795
3. Jagr - 671
4. Forsberg - 583
5. Selanne - 577
6. Ovechkin - 574
7. Sakic - 570
8. Thornton - 560
9. Malkin - 553
10. Crosby - 549

Notes - Lemieux, Crosby, and Malkin get dinged the hardest here. Without turning this into a war, I'm guessing that Mario might actually be with or ahead of Wayne, had the bar been set at 60 games, or if the second method of crunching this was applied. Either way, in the context of history, I wouldn't use any of these numbers to split hairs about which one was more dominant offensively in a game of hockey in their primes. They are CLEARLY in their own ballpark, and even in this instance, that favors 99, 6 is beat by 6 points a year. They are clearly A++ players, and to further rate them one should look past the numbers, at this point. Jagr is clearly third, offensively - there really is no case to argue him into the top 2, or into the next group. I'm speaking strict production here, not intangibles, although, even then, and I am a HUGE Selke/physical edge promoter, it is very difficult to make a statement for Foppa or Joe (the two more 'intangible guys in this group).

Foppa ahead of Joe! Strange for me, as strange as Joe being ahead on the 3 season/60 game version. The guy with the gloriously long career actually has less super solid seasons, and the guy notorious for a huge peak and too many injuries, actually has less outstanding seasons. hmmm, glad i did the list, just to learn this. Again, they are so close in either list, intangibles should be what seperates them, not numbers.

Malkin, Crosby, Ovechkin all get shafted strictly on lack of seasons to choose from. Impressive that they have done so much in so little time. Ditto Stamkos on the 3 year list.

Selanne and Thornton - you know, I always sort of overlooked these guys as very high end compilers. That Selanne was elite has been beaten into my head on this site, so I have accepted it recently, and confirmed it, now. Jumbo Joe... geez, i know his playoffs stink a bit, but maybe we all overlook him a bit?

tombombadil is offline  
Old
08-12-2012, 01:49 PM
  #29
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
Quote:
Originally Posted by Czech Your Math View Post
Tom,

I don't think there is an exact point where a pace becomes a season. A season is a season and a pace is a pace. A pace becomes more reliable as the sample size increases, but it never exactly equates to a season. Also, I believe the reliability increases if the pace is similar to other seasons the player had. The more full seasons which are close to the pace for that season, the more reliable the pace would seem to be.

I voted for 55 based on this logic: At that point, the games missed become 1/2 or less of the games played, so the reliability has improved significantly. That's an intuitive "inflection point", if you will, but may be wrong. Still, I stand by the points made in the first paragraph, the two are never really the same.

Before I start working on this, any suggestions for a total game threshold over multiple seasons? Maybe a couple numbers between 200 and 400? If not, I'll pick at least one number that doesn't appear to affect either Wayne or Mario unfairly.
Pat Lafontaine would thank you.

I am very interested in a total game threshold, and also even a consecutive game prime (maybe even including playoff games...)

A total game threshold would eliminate arguments over what a season is, and would, in effect, simply be a gigantic 'pace' measurement. I like it, a lot, actually.

I like somewhere around 300 games, but ya, 200 - 400 area seems right, for sure. Maybe 250 as it is basically 3 full seasons.... i dunno, do a poll!!

tombombadil is offline  
Old
08-12-2012, 01:56 PM
  #30
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
Quote:
Originally Posted by tombombadil View Post
I don't know if 'sliding scale' is a term that applies to this, but I crunched the top 10 guys from the original list into a 5 year list with a minimum of 70 games played. However, based on extraordinary performances, some guys had better years in total adjusted points in less than 70 games than they did in 'full' seasons. Add to that that Crosby and Malkin actually don't have 5 'full' seasons. So, gave zero credit for games missed up to 70, then took the ratio 82/70 (or 80/70, 84/70 depending on year) and used that to pro rate the season. so, if a player had 98 points in 63 games, it would count as 98/70, thus 82/70 x 98 = whatever.

I feel this doubly penalizes for injury, but allows everyone to participate, and gives credit to great seasons. After crunching all these, I feel strongly that a more balanced method would be to take their point per game based on what they acheived, yet only apply it to 12 phantom games (ten in the 80's, 14 in those two annoying years) Anyways, Iam not going to re-crunch, as my wife is looking at me like I am a complete waste of time!

1. Gretzky - 828
2. Lemieux - 795
3. Jagr - 671
4. Forsberg - 583
5. Selanne - 577
6. Ovechkin - 574
7. Sakic - 570
8. Thornton - 560
9. Malkin - 553
10. Crosby - 549

Notes - Lemieux, Crosby, and Malkin get dinged the hardest here. Without turning this into a war, I'm guessing that Mario might actually be with or ahead of Wayne, had the bar been set at 60 games, or if the second method of crunching this was applied. Either way, in the context of history, I wouldn't use any of these numbers to split hairs about which one was more dominant offensively in a game of hockey in their primes. They are CLEARLY in their own ballpark, and even in this instance, that favors 99, 6 is beat by 6 points a year. They are clearly A++ players, and to further rate them one should look past the numbers, at this point. Jagr is clearly third, offensively - there really is no case to argue him into the top 2, or into the next group. I'm speaking strict production here, not intangibles, although, even then, and I am a HUGE Selke/physical edge promoter, it is very difficult to make a statement for Foppa or Joe (the two more 'intangible guys in this group).

Foppa ahead of Joe! Strange for me, as strange as Joe being ahead on the 3 season/60 game version. The guy with the gloriously long career actually has less super solid seasons, and the guy notorious for a huge peak and too many injuries, actually has less outstanding seasons. hmmm, glad i did the list, just to learn this. Again, they are so close in either list, intangibles should be what seperates them, not numbers.

Malkin, Crosby, Ovechkin all get shafted strictly on lack of seasons to choose from. Impressive that they have done so much in so little time. Ditto Stamkos on the 3 year list.

Selanne and Thornton - you know, I always sort of overlooked these guys as very high end compilers. That Selanne was elite has been beaten into my head on this site, so I have accepted it recently, and confirmed it, now. Jumbo Joe... geez, i know his playoffs stink a bit, but maybe we all overlook him a bit?
That's a good list of post-WHA talents. I still think it needs more Lindros, and I'm not a big fan of the Big E.

Selanne and Thornton may both be underrated a bit due to their playoff numbers and relative lack of team success compared to some other greats. They are similar to Dionne in this way.

Sakic's edge over Forsberg is mainly much more longevity and being a better goal-scorer. He had his own injuries as well, so he wasn't that much more durable within each season. Forsberg's biggest edges, from a data standpoint, are his even strength effectiveness (see adjusted plus-minus, ES GF/GA ratios on vs. off) and his team's record with or without him in the lineup. There are many pieces to the puzzle. What's important is the big picture. Too much is placed on one stat and/or one season, when it's multiple metrics over multiple seasons that shows the cream rising to the top.

Czech Your Math is offline  
Old
08-12-2012, 02:00 PM
  #31
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
Quote:
Originally Posted by tombombadil View Post
Pat Lafontaine would thank you.

I am very interested in a total game threshold, and also even a consecutive game prime (maybe even including playoff games...)

A total game threshold would eliminate arguments over what a season is, and would, in effect, simply be a gigantic 'pace' measurement. I like it, a lot, actually.

I like somewhere around 300 games, but ya, 200 - 400 area seems right, for sure. Maybe 250 as it is basically 3 full seasons.... i dunno, do a poll!!
I'll pick some round numbers (either nearest 50 or multiples of 80/82 games) that don't seem too biased for/against the top 3. I'm not going to include playoff games, as that would require a lot more work to separately adjust and include those numbers.

Now go do something with your wife, and expect some results late tonight or early next week.

Czech Your Math is offline  
Old
08-12-2012, 02:06 PM
  #32
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
the more i think about your total games version, the more i like it. It's 'who had the most productive individual career without hurting someone for playing too long, or for missing too much time' or, more simply, "who had the better prime"

So, what's a prime? Some guys fly right away, but the rule of thumb i know of for young guys is the 4 year rule. roughly guessing, an average forward might be hitting his prime at 24? I have no idea, but i'm going to suggest that somewhere around 32 is generally the start of a decline. i don't suggest these age parameters be used whatsoever, just using it to say there is roughly a 9 year 'prime' on average (again guessing most of this) Very few guys play all of those games, very few play half, maybe 75% would be a decent average for a player over 9 years?

720 x .75 = 540.

I'm not using these numbers for anything more than process of thought, by the way.

Maybe I am leaning more towards your threshold of 400 now, though.

I like it. Can't stand Crosby, for instance - but the guy deserves credit for producing so well over the last 2 seasons. and 400 games would balance out the crazy pace factor of simply giving him credit for close to 2 adj points a game for those seasons.

Whichever threshold you choose, i like it. More than my list, actually.

tombombadil is offline  
Old
08-12-2012, 02:07 PM
  #33
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
haha, thanks, see ya!

tombombadil is offline  
Old
08-12-2012, 02:17 PM
  #34
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
For a forward, I think prime is generally ~21/22 to ~29/30. Goalies and defensemen would generally be later.

I might start a separate thread for some of the data, once it's complete. Not sure yet.

The good thing about using a threshold of games: Players aren't receiving credit for a higher proportion of missed games during a season. Each game is weighted equally (although one could adjust for schedule... which I may do, esp. if go back to much earlier decades).

Czech Your Math is offline  
Old
08-12-2012, 04:35 PM
  #35
Bear of Bad News
Mod Supervisor
 
Bear of Bad News's Avatar
 
Join Date: Sep 2005
Location: Windsor
Posts: 4,866
vCash: 663
Quote:
Originally Posted by Czech Your Math View Post
Now go do something with your wife, and expect some results late tonight or early next week.
Wow - what sort of results are we hoping for?

Bear of Bad News is offline  
Old
08-12-2012, 04:37 PM
  #36
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
hilarious!

We already have 3 kids - that **** is over!

tombombadil is offline  
Old
08-12-2012, 05:19 PM
  #37
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
Quote:
Originally Posted by Taco MacArthur View Post
Wow - what sort of results are we hoping for?
Scoring results of course! Let's hope he remembers to pace himself... those "I was on pace for..." arguments won't work with the wife.

Czech Your Math is offline  
Old
08-12-2012, 05:24 PM
  #38
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
haha, that is way too funny. Never mind not playing a full season, I'm just an energy player.... 45 second shifts, and back on the bench!

tombombadil is offline  
Old
08-12-2012, 06:10 PM
  #39
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
nice, thanks. mollifying the Gretzky fans is funny to me. Just like soccer, the fans are way more vicious than the player(s)!

tombombadil is offline  
Old
08-12-2012, 07:03 PM
  #40
Czech Your Math
Registered User
 
Czech Your Math's Avatar
 
Join Date: Jan 2006
Location: bohemia
Country: Czech_ Republic
Posts: 4,841
vCash: 500
I looked at the big 3 (WG,ML,JJ) and the little 3 (SC,AO,EM) in terms of games in their best seasons, so that the little 3 can be fairly included and so that there's no real Wayne vs. Mario bias.

The number of games I will use are: 300, 425 and 640.

300 allows Malkin's best seasons to be included and not his worst. Most of these players have a break point between 300-330, so that seemed the fairest solution.

425 allows Malkin and Crosby's careers to be included. The only one who's a bit penalized of these 6 would be Gretzky, but he will still be at the top I would guess, so it shouldn't matter too much. That seems a fair tradeoff to be able to include all of the little 3 in this list. Ovechkin still gets to take a season off, so that seems pretty fair as well.

640 is eight 80 game seasons, which is important since Wayne and Mario had a lot of 80 game schedules. It also is near breakpoints for Lemieux (646) and Jagr (640). Gretzky again is slightly penalized, but at this point it shouldn't matter much. Also keep in mind that since actual games will be used for these 3 lists, so Jagr is penalized for the shortened lockout season (and he missed a lockout season as well).

I will eventually post tables which will include the number of actual games for each player for each period. That way, people can judge for themselves and consider that a player had more games than other players.

If you have any suggested changes to the number of games, please post them in the very near future, as once I begin the calculations, I'm not going to go back and do double work.

Czech Your Math is offline  
Old
08-12-2012, 07:06 PM
  #41
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
looks good.

tombombadil is offline  
Old
08-12-2012, 07:57 PM
  #42
seventieslord
Registered User
 
seventieslord's Avatar
 
Join Date: Mar 2006
Location: Regina, SK
Country: Canada
Posts: 29,235
vCash: 500
Quote:
Originally Posted by tombombadil View Post
haha, i forgot about that! Way to notice, ya - that's Bernie. He can thank 99 for making the list.
Oh good. I was going to ask.

seventieslord is online now  
Old
08-13-2012, 09:01 PM
  #43
tombombadil
Registered User
 
Join Date: Jan 2010
Location: West Kelowna, Canada
Country: Canada
Posts: 1,029
vCash: 500
I'd like to see this thread closed, based on that their is a new and improved version out now.


Last edited by Czech Your Math: 08-13-2012 at 10:53 PM. Reason: thy will shall be done, good sir
tombombadil 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 05:48 PM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"

vBulletin Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. 2017 All Rights Reserved.