HFBoards

Go Back   HFBoards > General Hockey Discussion > National Hockey League Talk
National Hockey League Talk Discuss NHL players, teams, games, and the Stanley Cup Playoffs.

The success of Anaheim shows that Boudreau is an excellent coach.

Reply
 
Thread Tools
Old
03-21-2013, 12:09 PM
  #301
duxfever
Registered User
 
duxfever's Avatar
 
Join Date: Mar 2009
Location: Orange, CA
Country: United States
Posts: 798
vCash: 500
Send a message via AIM to duxfever
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Why does it bother you so much that people who are stats-minded like to dig into a greater level of depth than simple wins-losses?

Enjoy your recent success.
Because stats can be skewed and misinterpreted, as well as be misrepresented. They do not always tell the whole story, and more times than not, they tell such a small percentage of the story.

TBH, I am a numbers' guy myself. For Ducks' fans, it has just become annoying with all of these "stat" guys telling us that we are eventually going to hit a wall and struggle. Where is the stat for that? Is the only stat that explains hitting the wall the generality that we will regress towards the mean? Stats are great, don't get me wrong, but many are flawed and do not represent what is being portrayed on the ice night in and night out.

Many of these "stat" guys come in here as if they are of some higher intellectual prowess and as if their findings are 100% accurate predictors of what is going to happen. That is just simply not true. There is value in watching games as well as looking into stats. I think both sides need to do a little more of both.

duxfever is offline   Reply With Quote
Old
03-21-2013, 12:15 PM
  #302
GuineaPig
Registered User
 
GuineaPig's Avatar
 
Join Date: Jul 2011
Location: Montréal
Posts: 2,103
vCash: 500
By the way, you can look at past PDO graphs over the last 5 seasons at Behind the Net.

Here are a couple interesting ones from last year:



This matches almost perfectly with the media portrayal of the Bruins last season. You can see them go from "Cup hangover!" to "Best team ever!" to "Tim Thomas has destroyed the team by not going to the White House!"



This is Minnesota's. They were in Anaheim's shoes last year, and their fans said the same things. I suppose the principal differences being the length of the season allowed the Wild to drop out of the playoffs, and that Anaheim is a better team.



A happier story (well, for me at least): the media portrayed this as the scrappy Sens gelling under MacLean and fighting back for a place in the playoffs, but the reality was that Ottawa was good all year. Just their luck started off terrible and finished good.

GuineaPig is offline   Reply With Quote
Old
03-21-2013, 12:18 PM
  #303
MMonarchs
Registered User
 
Join Date: Apr 2011
Posts: 550
vCash: 500
Quote:
Originally Posted by GuineaPig View Post
This is Minnesota's. They were in Anaheim's shoes last year, and their fans said the same things. I suppose the principal differences being the length of the season allowed the Wild to drop out of the playoffs, and that Anaheim is a better team.
This may have worked a few weeks ago, but Anaheim is well past being the Minnesota of last year.

MMonarchs is offline   Reply With Quote
Old
03-21-2013, 12:27 PM
  #304
Fallenity
Registered User
 
Fallenity's Avatar
 
Join Date: Apr 2011
Posts: 1,182
vCash: 950
Quote:
Originally Posted by GuineaPig View Post
All the graphs
Those are really interesting, thanks. The only thing I don't really like is the fact that those graphs have a massive variance at the beginning that evens out as the season progresses due to increasing sample size. This makes it harder to compare different performances with each other visually (like how the team performed in the beginning of the season vs how they performed at the end).

It'd be interesting to see the same graph, but with the average of 10 last games always shown. That way you can better see the development. It could be superimposed on those graphs so you get to see the overall development.

Fallenity is offline   Reply With Quote
Old
03-21-2013, 12:31 PM
  #305
GuineaPig
Registered User
 
GuineaPig's Avatar
 
Join Date: Jul 2011
Location: Montréal
Posts: 2,103
vCash: 500
Quote:
Originally Posted by Fallenity View Post
It'd be interesting to see the same graph, but with the average of 10 last games always shown. That way you can better see the development. It could be superimposed on those graphs so you get to see the overall development.
Yeah, I think that would definitely be an improvement. It's much harder to see the swings in percentages and the little streaks that happen (except in the beginning, when they're over-exaggerated).

GuineaPig is offline   Reply With Quote
Old
03-21-2013, 12:32 PM
  #306
bminucci
Registered User
 
bminucci's Avatar
 
Join Date: Nov 2010
Location: Valley Village, CA
Country: United States
Posts: 251
vCash: 500
Quote:
Originally Posted by Sojourn View Post
I'd like to know when winning games stopped being a benchmark for a successful team. Now we're applying an asterisk next to a team because they aren't winning the "right" way?
When the Ducks won too many for the naysayers to explain.

bminucci is offline   Reply With Quote
Old
03-21-2013, 12:38 PM
  #307
bminucci
Registered User
 
bminucci's Avatar
 
Join Date: Nov 2010
Location: Valley Village, CA
Country: United States
Posts: 251
vCash: 500
I would love to see the Ducks Fenwick graph for 2002-2003.

bminucci is offline   Reply With Quote
Old
03-21-2013, 01:06 PM
  #308
Feed Me A Stray Cat
Registered User
 
Feed Me A Stray Cat's Avatar
 
Join Date: Mar 2005
Location: Boston, MA
Country: United States
Posts: 11,123
vCash: 500
Send a message via AIM to Feed Me A Stray Cat
Quote:
Originally Posted by MMonarchs View Post
This may have worked a few weeks ago, but Anaheim is well past being the Minnesota of last year.
Why and how? The ducks have played 29 games this season.

Feed Me A Stray Cat is offline   Reply With Quote
Old
03-21-2013, 01:07 PM
  #309
Feed Me A Stray Cat
Registered User
 
Feed Me A Stray Cat's Avatar
 
Join Date: Mar 2005
Location: Boston, MA
Country: United States
Posts: 11,123
vCash: 500
Send a message via AIM to Feed Me A Stray Cat
Quote:
Originally Posted by GuineaPig View Post
By the way, you can look at past PDO graphs over the last 5 seasons at Behind the Net.

Here are a couple interesting ones from last year:



This matches almost perfectly with the media portrayal of the Bruins last season. You can see them go from "Cup hangover!" to "Best team ever!" to "Tim Thomas has destroyed the team by not going to the White House!"



This is Minnesota's. They were in Anaheim's shoes last year, and their fans said the same things. I suppose the principal differences being the length of the season allowed the Wild to drop out of the playoffs, and that Anaheim is a better team.



A happier story (well, for me at least): the media portrayed this as the scrappy Sens gelling under MacLean and fighting back for a place in the playoffs, but the reality was that Ottawa was good all year. Just their luck started off terrible and finished good.
Good post.

The interesting scenario will be the reasons that are given for Anaheim's eventual malaise. No one will attribute it to luck reversion - it will be something like "the team got too confident and slacked off", or "Boudreau tinkered with things too much."

Feed Me A Stray Cat is offline   Reply With Quote
Old
03-21-2013, 01:10 PM
  #310
WarriorOfGandhi
Was saying Boo-urns
 
WarriorOfGandhi's Avatar
 
Join Date: Jul 2007
Location: Cincinnati, Ohio
Country: Scotland
Posts: 12,649
vCash: 142
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Good post.

The interesting scenario will be the reasons that are given for Anaheim's eventual malaise. No one will attribute it to luck reversion - it will be something like "the team got too confident and slacked off", or "Boudreau tinkered with things too much."

WarriorOfGandhi is offline   Reply With Quote
Old
03-21-2013, 01:21 PM
  #311
mmbt
Cheeky Monkey
 
mmbt's Avatar
 
Join Date: Feb 2002
Location: California
Country: United States
Posts: 9,205
vCash: 500
Quote:
Originally Posted by GuineaPig View Post
No, they can't. If you flip a coin one hundred times in a row, it's extremely unlikely that you'll guess the exact number of heads and tails.

That doesn't mean that guessing 50/50 isn't the most likely to be accurate play.
True, but in that situation if you assume an unbiased coin, you know that the "true" percentage should be 50% heads 50% tails. Problem is with hockey teams, we don't know what the "true" perecentage is. Rosters are in flux even within a season, players get healthy or get hurt, young guys may be improving greatly from previous averages, etc.. Even an 82 game sample isn't necessarily definitive ... individual players have good and bad years.

It seems unlikely that if the Ducks are a "true" .500 team that they'd be capable of a 22-3-4 start. At this point I think it's pretty safe to say, they're very likely a good team. How good, who knows for sure?

It's like people talking about shooting percentages ... well how do you know what the "real" shooting percentage for this particular team should be if they played an infinite number of games against league average competition? Is 11% "sustainable? Probably not. But there are plenty of teams that shoot in the 10% range for a full season. Heck, just 7 years ago 18 teams finished over 10% shooting.

And if you're going to claim that they "should" be shooting 9% or something as some here do ... well then basically aren't they saying that there's 23 goals in 29 games that the Ducks somehow should not have scored? I'd like to see them point out all of those lucky goals, and then do the same for other teams to demonstrate that Anaheim has benefited from unusual puck luck.

Look, I've worked in science, and I've done statistical analysis, and have been published in journals. Indirect measurements and such are useful ... but they also need to match up with the experimental results. If something isn't coming out like you predicted, then first you need to sift through your experiments and see if you did something wrong (in this case, finding an inordinate number of lucky plays that would explain the contrary results), or you need to review your predictions because you may be missing something (other factors that contribute to winning which you had disregarded).

Essentially what we have here is experimental result A (wins) that people claim are a consequence of B (puck possession). However, there's no real measurement of B ... so they use C (shots) which is suggested by many closely correlates to B. But if you're getting a positive result on A, with a negative result on C, then the FIRST thing you need to do is question whether C and B really correlate as well as you think. Not question whether your positives on A are truly positive, or whether they're fraudulent.

And if as someone suggested, C with larger sample sizes becomes less and less predictive of A, then I'd really start to look hard at just how well C truly measures B. Or I'd question how well B measures A. Or whether there's other factors D, E, and F that combine with B in some way to better measure A. I wouldn't cling to the idea that C is the definitive measuring tool and that there must be an error with the experimental result A.

mmbt is offline   Reply With Quote
Old
03-21-2013, 01:34 PM
  #312
Ducksgo
#EtemUp
 
Join Date: Jan 2006
Location: Lakewood CA
Country: United States
Posts: 6,502
vCash: 500
Send a message via AIM to Ducksgo
Quote:
Originally Posted by Ishdul View Post
Anaheim has gotten absurd luck, with the highest shooting percentage post lockout and getting very fluky good goaltending from Fasth (unless you actually think he's a Vezina calibre goaltender), which covers up for the fact that they're a terrible possession team with the 6th worst Fenwick and 3rd worst Corsi in the league. They are this year's Minnesota but with only 20 games to collapse instead of 50. They will regress heavily. Boudreau is also a bad coach but mostly for other reasons.
Wow just wow..... Luck? Really?

Haters ARE always going to hate....

Ducksgo is online now   Reply With Quote
Old
03-21-2013, 01:35 PM
  #313
Ducks DVM
Moderator
There is no grunion
 
Ducks DVM's Avatar
 
Join Date: Jun 2010
Location: Long Beach, CA
Country: United States
Posts: 14,061
vCash: 500
How does Minnesota's "regression" match up with their injury issues last year? How much was due to the magic of averages and how much was due to having no second line when that crash started and an AHL defense for a good part of the year? They were one of the leaders in man games lost to injuries, and they were significant players, no?

Ducks DVM is offline   Reply With Quote
Old
03-21-2013, 01:36 PM
  #314
Diggy
Registered User
 
Diggy's Avatar
 
Join Date: Oct 2007
Location: Orange County
Country: United States
Posts: 750
vCash: 500
Quote:
Originally Posted by GuineaPig View Post
No, they can't. If you flip a coin one hundred times in a row, it's extremely unlikely that you'll guess the exact number of heads and tails.

That doesn't mean that guessing 50/50 isn't the most likely to be accurate play.
The probability of a coin flip is 50/50.

This is not the same thing as the Corsi or Fenwick "advanced statistics". Anyone who thinks that Corsi or Fenwick, which rely on judgement calls by humans and only count part of game among many other problems, is nearly as accurate is a coin flip is just being plain ignorant. Really, how many good and solid statistics do you know that get less accurate with more information like the Corsi and Fenwick do?

Now I am not saying the Corsi and Fenwick are useless. They do highlight and indicate certain things, but I don't think one can use them alone to decide if a team is good or just lucky.

Diggy is offline   Reply With Quote
Old
03-21-2013, 01:40 PM
  #315
Feed Me A Stray Cat
Registered User
 
Feed Me A Stray Cat's Avatar
 
Join Date: Mar 2005
Location: Boston, MA
Country: United States
Posts: 11,123
vCash: 500
Send a message via AIM to Feed Me A Stray Cat
Quote:
Originally Posted by mmbt View Post
True, but in that situation if you assume an unbiased coin, you know that the "true" percentage should be 50% heads 50% tails. Problem is with hockey teams, we don't know what the "true" perecentage is. Rosters are in flux even within a season, players get healthy or get hurt, young guys may be improving greatly from previous averages, etc.. Even an 82 game sample isn't necessarily definitive ... individual players have good and bad years.

It seems unlikely that if the Ducks are a "true" .500 team that they'd be capable of a 22-3-4 start. At this point I think it's pretty safe to say, they're very likely a good team. How good, who knows for sure?

It's like people talking about shooting percentages ... well how do you know what the "real" shooting percentage for this particular team should be if they played an infinite number of games against league average competition? Is 11% "sustainable? Probably not. But there are plenty of teams that shoot in the 10% range for a full season. Heck, just 7 years ago 18 teams finished over 10% shooting.

And if you're going to claim that they "should" be shooting 9% or something as some here do ... well then basically aren't they saying that there's 23 goals in 29 games that the Ducks somehow should not have scored? I'd like to see them point out all of those lucky goals, and then do the same for other teams to demonstrate that Anaheim has benefited from unusual puck luck.

Look, I've worked in science, and I've done statistical analysis, and have been published in journals. Indirect measurements and such are useful ... but they also need to match up with the experimental results. If something isn't coming out like you predicted, then first you need to sift through your experiments and see if you did something wrong (in this case, finding an inordinate number of lucky plays that would explain the contrary results), or you need to review your predictions because you may be missing something (other factors that contribute to winning which you had disregarded).

Essentially what we have here is experimental result A (wins) that people claim are a consequence of B (puck possession). However, there's no real measurement of B ... so they use C (shots) which is suggested by many closely correlates to B. But if you're getting a positive result on A, with a negative result on C, then the FIRST thing you need to do is question whether C and B really correlate as well as you think. Not question whether your positives on A are truly positive, or whether they're fraudulent.

And if as someone suggested, C with larger sample sizes becomes less and less predictive of A, then I'd really start to look hard at just how well C truly measures B. Or I'd question how well B measures A. Or whether there's other factors D, E, and F that combine with B in some way to better measure A. I wouldn't cling to the idea that C is the definitive measuring tool and that there must be an error with the experimental result A.
Seven years ago is completely irrelevant right now. The median shooting percentage this season is 9.1%.

The problem is that a 29 game sample is not nearly enough to come to legitimate conclusions about a team's real shooting ability. Until the Ducks demonstrate multiple seasons of shooting 2.7% above the league average, I'm going to be inclined to believe they will regress to the mean, just like every other team has in recent memory.

Additionally, you won't necessarily be able to point to every lucky goal. Some goals might not even look lucky by themselves. But maybe there were subtle bounces that occurred in the ten seconds before the goal was scored to shift the luck in Anaheim's favor? This is not a science of simply watching every goal scored and saying "lucky" or "not lucky".

Feed Me A Stray Cat is offline   Reply With Quote
Old
03-21-2013, 01:42 PM
  #316
GuineaPig
Registered User
 
GuineaPig's Avatar
 
Join Date: Jul 2011
Location: Montréal
Posts: 2,103
vCash: 500
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Good post.

The interesting scenario will be the reasons that are given for Anaheim's eventual malaise. No one will attribute it to luck reversion - it will be something like "the team got too confident and slacked off", or "Boudreau tinkered with things too much."
I imagine it'll be similar to what people are saying about Hitchcock. The only difference between St. Louis last year and St. Louis this year is that they're getting unsustainably bad goaltending instead of unsustainably good goaltending. And so you have people theorizing that Hitchcock's style "only works for one year, and then the players tune out."

GuineaPig is offline   Reply With Quote
Old
03-21-2013, 01:44 PM
  #317
Feed Me A Stray Cat
Registered User
 
Feed Me A Stray Cat's Avatar
 
Join Date: Mar 2005
Location: Boston, MA
Country: United States
Posts: 11,123
vCash: 500
Send a message via AIM to Feed Me A Stray Cat
Quote:
Originally Posted by Diggy View Post
The probability of a coin flip is 50/50.

This is not the same thing as the Corsi or Fenwick "advanced statistics". Anyone who thinks that Corsi or Fenwick, which rely on judgement calls by humans and only count part of game among many other problems, is nearly as accurate is a coin flip is just being plain ignorant. Really, how many good and solid statistics do you know that get less accurate with more information like the Corsi and Fenwick do?

Now I am not saying the Corsi and Fenwick are useless. They do highlight and indicate certain things, but I don't think one can use them alone to decide if a team is good or just lucky.
I'm not using Fenwick to say Anaheim has been lucky. I'm looking at Anaheim's unsustainable shooting percentage to say that they've been getting lucky. Fenwick and shooting percentage are largely independent of each other.

Feed Me A Stray Cat is offline   Reply With Quote
Old
03-21-2013, 02:03 PM
  #318
duxfever
Registered User
 
duxfever's Avatar
 
Join Date: Mar 2009
Location: Orange, CA
Country: United States
Posts: 798
vCash: 500
Send a message via AIM to duxfever
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Seven years ago is completely irrelevant right now. The median shooting percentage this season is 9.1%.

The problem is that a 29 game sample is not nearly enough to come to legitimate conclusions about a team's real shooting ability. Until the Ducks demonstrate multiple seasons of shooting 2.7% above the league average, I'm going to be inclined to believe they will regress to the mean, just like every other team has in recent memory.

Additionally, you won't necessarily be able to point to every lucky goal. Some goals might not even look lucky by themselves. But maybe there were subtle bounces that occurred in the ten seconds before the goal was scored to shift the luck in Anaheim's favor? This is not a science of simply watching every goal scored and saying "lucky" or "not lucky".
This is ****ing laughable. Define "luck".

duxfever is offline   Reply With Quote
Old
03-21-2013, 03:19 PM
  #319
KEEROLE Vatanen
Failures Of Fenwick
 
KEEROLE Vatanen's Avatar
 
Join Date: Oct 2011
Posts: 14,267
vCash: 500
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Why does it bother you so much that people who are stats-minded like to dig into a greater level of depth than simple wins-losses?

Enjoy your recent success.
Keep digging, i'll continue to enjoy being 19 games over .500

KEEROLE Vatanen is offline   Reply With Quote
Old
03-21-2013, 03:38 PM
  #320
bminucci
Registered User
 
bminucci's Avatar
 
Join Date: Nov 2010
Location: Valley Village, CA
Country: United States
Posts: 251
vCash: 500
Quote:
Originally Posted by Feed Me A Stray Cat View Post
Most of this discussion has been about Anaheim's shooting percentage, and thus their offense, so let's keep it there for now. I think their goaltending is unsustainable too, but that's another argument.

You're essentially assuming that Anaheim is better than other teams at creating high quality chances, thus leading to their high shooting percentage. You can certainly believe that, just know that Minnesota Wild fans thought the same thing in December 2011, Colorado Avalanche fans thought the same thing after 2010, and Washington fans thought the same thing after 2009.

Teams that live on the extremes of the percentages will always come back to the middle.
OK, next year they can come back down.

bminucci is offline   Reply With Quote
Old
03-21-2013, 03:40 PM
  #321
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 mmbt View Post

It's like people talking about shooting percentages ... well how do you know what the "real" shooting percentage for this particular team should be if they played an infinite number of games against league average competition? Is 11% "sustainable? Probably not. But there are plenty of teams that shoot in the 10% range for a full season. Heck, just 7 years ago 18 teams finished over 10% shooting.
Well, I believe the 11% figure is for 5-on-5 play only. When you say that 18 teams finished over 10%, I believe that you're referring to overall shooting percentage, which is higher on average - about 9% these days, as opposed to about 8% at EV.

In any event, it's a good question.

A true-talent 5-on-5 shooting percentage of 11% is attainable in theory. The problem is that the probability of observing such a team is extremely remote.

The talent standard deviation among teams in 5-on-5 shooting percentage is about 0.5%. So given that the league average is 8%, 11% would be six standard deviations above the mean in talent.

You're right in that some teams have shot 10% or even better at EV over the last few years - 06-07 BUF, 06-07 NSH and 09-10 WSH come to mind. 09-10 COL may have even done it, too.

The problem is that all of those teams regressed hard in the following seasons. Bayesian analysis held that 06-07 BUF (who I believe shot 11.4% at EV), were actually a true talent 9% team. And I think they shot 9% at EV in the following season. WSH and NSH experienced even harder regressions, IIRC.

To be fair, teams change from one year to the next, and that surely accounts for some of the observed change. But regression can be observed and measured within seasons as well. Theory dictates that a team's EV shooting percentage should regress 79% to the mean from one half of the year to the other.

And if you look empirically at what actually happened - and I have - the observed regression is 77%. So the theory matches the results quite well.

Anyway, sorry for the rambling response, but I hope it's contributed to the discussion here on some level.

Master_Of_Districts is offline   Reply With Quote
Old
03-21-2013, 03:41 PM
  #322
bminucci
Registered User
 
bminucci's Avatar
 
Join Date: Nov 2010
Location: Valley Village, CA
Country: United States
Posts: 251
vCash: 500
Who is to say that the Ducks shooting percent is going to go down by less goals? The number of shots could go up to make the % fall closer in line.

bminucci is offline   Reply With Quote
Old
03-21-2013, 03:46 PM
  #323
duxfever
Registered User
 
duxfever's Avatar
 
Join Date: Mar 2009
Location: Orange, CA
Country: United States
Posts: 798
vCash: 500
Send a message via AIM to duxfever
Quote:
Originally Posted by Master_Of_Districts View Post
Well, I believe the 11% figure is for 5-on-5 play only. When you say that 18 teams finished over 10%, I believe that you're referring to overall shooting percentage, which is higher on average - about 9% these days, as opposed to about 8% at EV.

In any event, it's a good question.

A true-talent 5-on-5 shooting percentage of 11% is attainable in theory. The problem is that the probability of observing such a team is extremely remote.

The talent standard deviation among teams in 5-on-5 shooting percentage is about 0.5%. So given that the league average is 8%, 11% would be six standard deviations above the mean in talent.

You're right in that some teams have shot 10% or even better at EV over the last few years - 06-07 BUF, 06-07 NSH and 09-10 WSH come to mind. 09-10 COL may have even done it, too.

The problem is that all of those teams regressed hard in the following seasons. Bayesian analysis held that 06-07 BUF (who I believe shot 11.4% at EV), were actually a true talent 9% team. And I think they shot 9% at EV in the following season. WSH and NSH experienced even harder regressions, IIRC.

To be fair, teams change from one year to the next, and that surely accounts for some of the observed change. But regression can be observed and measured within seasons as well. Theory dictates that a team's EV shooting percentage should regress 79% to the mean from one half of the year to the other.

And if you look empirically at what actually happened - and I have - the observed regression is 77%. So the theory matches the results quite well.

Anyway, sorry for the rambling response, but I hope it's contributed to the discussion here on some level.
And personnel, coaching, better competition, etc. had nothing to do with these teams year to year? It was just a return to the mean just because of the law of averages?

Each year is unique. How can both years be one in the same?

duxfever is offline   Reply With Quote
Old
03-21-2013, 03:49 PM
  #324
duxfever
Registered User
 
duxfever's Avatar
 
Join Date: Mar 2009
Location: Orange, CA
Country: United States
Posts: 798
vCash: 500
Send a message via AIM to duxfever
Quote:
Originally Posted by bminucci View Post
Who is to say that the Ducks shooting percent is going to go down by less goals? The number of shots could go up to make the % fall closer in line.
That's too simple!

This was my point in another thread. So if the Ducks were to shoot 50 times, but a goalie stood on his head and made, let's say, 49 saves. Since the shooting % would revert towards the "mean", in theory, this would make the Ducks a more legitimate team statistics wise and less likely to see a regression coming?

duxfever is offline   Reply With Quote
Old
03-21-2013, 03:51 PM
  #325
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 duxfever View Post
And personnel, coaching, better competition, etc. had nothing to do with these teams year to year? It was just a return to the mean just because of the law of averages?

Each year is unique. How can both years be one in the same?
I believe I acknowledged that. Perhaps you should re-read my post - I also discussed how the strong regression is observed within seasons as well.

In any event, some metrics repeat quite well from year to year at the team level. Like fenwick, for example. In fact, most possession-based statistics have high reliability across seasons. Even strength shooting percentage does not.

What does that tell you? Just think about it.

Master_Of_Districts 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 07:06 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.