HFBoards

Go Back   HFBoards > NHL Eastern Conference > Atlantic Division > Toronto Maple Leafs
Notices

Advanced Stats

Reply
 
Thread Tools
Old
07-22-2013, 07:23 AM
  #326
4evaBlue
Registered User
 
4evaBlue's Avatar
 
Join Date: Jan 2011
Posts: 4,345
vCash: 500
Quote:
Originally Posted by Stephen View Post
Actually, it paints a fairly accurate and unflattering picture of the stats crowd.

A crowd of sub-intellectuals regurgitating the same distorted picture, basically, without the flexibility or open mindedness or clarity in what they're trying to say. Why can't "useless" stats like +/-'s be combined with your findings to elucidate an understanding of the game?

Why can't "possession" simply be tracked by watching footage again and recording accurate puck possession times, in what zone and when, instead of having to use this 'reasonable approximation'?
Why bother measuring what you're interested in when you can use arbitrary stats to "estimate" the measurement? I mean, a shot resulting in a goal and the quality of the shot have a strong co-relation. Maybe we could spin up a garbage metric "measuring" a teams average quality of shot generated.

4evaBlue is offline   Reply With Quote
Old
07-22-2013, 07:39 AM
  #327
7even
Moderator
Deus Ex Machina
 
7even's Avatar
 
Join Date: Feb 2012
Location: Georgia
Country: United States
Posts: 7,719
vCash: 500
Quote:
Originally Posted by 4evaBlue View Post
Why bother measuring what you're interested in when you can use arbitrary stats to "estimate" the measurement? I mean, a shot resulting in a goal and the quality of the shot have a strong co-relation. Maybe we could spin up a garbage metric "measuring" a teams average quality of shot generated.
Hah, if you want to sit in front of a computer screen with a stopwatch and record possession time for ever player on every team, be my guest.

Or you can use shots, which heavily correlates to zone time. Your choice, really.

And "shot quality" is typically defined as shot distance.

7even is online now   Reply With Quote
Old
07-22-2013, 08:31 AM
  #328
ULF_55
Global Moderator
 
ULF_55's Avatar
 
Join Date: Feb 2002
Location: Mountain Standard Ti
Posts: 55,527
vCash: 500
Quote:
Originally Posted by 7even View Post
Hah, if you want to sit in front of a computer screen with a stopwatch and record possession time for ever player on every team, be my guest.

Or you can use shots, which heavily correlates to zone time. Your choice, really.

And "shot quality" is typically defined as shot distance.
Some questions:

A shot by every player from that distance is equal?

A player with a career shooting percentage of 10 and a player with a career shooting percentage of 15 does or does not impact the results?

__________________
http://www.youtube.com/watch?v=bA3LN_8hjM8.

Vaive and Ludzik on collapse, and Phaneuf.
ULF_55 is offline   Reply With Quote
Old
07-22-2013, 08:50 AM
  #329
Leafsin2014*
 
Join Date: Jul 2013
Posts: 145
vCash: 500
Quote:
Originally Posted by ULF_55 View Post
Some questions:

A shot by every player from that distance is equal?

A player with a career shooting percentage of 10 and a player with a career shooting percentage of 15 does or does not impact the results?
Nope, shots from different distances are hugely different depending on who is taking the shot - obviously.

BTW, your signature plastered on every single post is really annoying. Do you really need to have that plastered everywhere? Forget the fact that they played on different teams, different situations, etc, but what is the point in spamming everyone with this large box on every post?

Leafsin2014* is offline   Reply With Quote
Old
07-22-2013, 08:56 AM
  #330
ULF_55
Global Moderator
 
ULF_55's Avatar
 
Join Date: Feb 2002
Location: Mountain Standard Ti
Posts: 55,527
vCash: 500
Quote:
Originally Posted by Leafsin2014 View Post
Nope, shots from different distances are hugely different depending on who is taking the shot - obviously.

BTW, your signature plastered on every single post is really annoying. Do you really need to have that plastered everywhere? Forget the fact that they played on different teams, different situations, etc, but what is the point in spamming everyone with this large box on every post?
BTW, it isn't something that I do, and it is only on one post per page. It is a software thing.

You don't have to read it especially if it bothers you. They are just stats ... kind of funny in this thread to complain about it ... ironic.

ULF_55 is offline   Reply With Quote
Old
07-22-2013, 09:10 AM
  #331
hatterson
Global Moderator
 
hatterson's Avatar
 
Join Date: Apr 2010
Location: North Tonawanda, NY
Country: United States
Posts: 11,290
vCash: 874
Send a message via Skype™ to hatterson
Quote:
Originally Posted by Stephen View Post
Actually, it paints a fairly accurate and unflattering picture of the stats crowd.

A crowd of sub-intellectuals regurgitating the same distorted picture, basically, without the flexibility or open mindedness or clarity in what they're trying to say. Why can't "useless" stats like +/-'s be combined with your findings to elucidate an understanding of the game?

Why can't "possession" simply be tracked by watching footage again and recording accurate puck possession times, in what zone and when, instead of having to use this 'reasonable approximation'?
Over the long run +/-, or other goal based metrics can be used. Shot based metrics are used when your sample size of games is smaller in order to eliminate some of the luck based variation.

The main reason that corsi and/or fenwick are used to approximate possession is because shots, misses and blocks are already tracked stats whereas player specific possession is not.

If you want to go ahead and actually track possession in each zone knock yourself out. All you should need to record is all possession changes and the exact time they occurred for each of the 720 games that were played this last year or the 1230 games that are played in a normal 82 game season. Don't worry, it shouldn't take you that long, it's only 30+ days (50+ in a normal season) of hockey.


Quote:
Originally Posted by ULF_55 View Post
If a shot is a shot, save percentage is save percentage, whether it is even strength, short handed, or power play.
A shot isn't a shot. Corsi an even strength metric. Power play and penalty kill numbers are wildly variable from season to season.

Take Lunqvist as an example. Over the last 6 years his best ES save percentage is this past year at 93.61% and his worst was 08-09 at 92.21%. His best PK save percentage was 08-09 at 92.07% and his worst was last year at 85.12%

As far as save percentage while your team has a man advantage, there's just so few shots against that it's fairly meaningless. The most "tested" goalies only get something on the order of 60-70 shots against in an entire 82 game season. Many are in the 40-50 range.

__________________
Come join us on the By The Numbers forum. Take a look at our introduction post if you're new. If you have any questions, feel free to PM me.
hatterson is offline   Reply With Quote
Old
07-22-2013, 10:09 AM
  #332
7even
Moderator
Deus Ex Machina
 
7even's Avatar
 
Join Date: Feb 2012
Location: Georgia
Country: United States
Posts: 7,719
vCash: 500
Quote:
Originally Posted by ULF_55 View Post
Some questions:

A shot by every player from that distance is equal?

A player with a career shooting percentage of 10 and a player with a career shooting percentage of 15 does or does not impact the results?
You've got some extremely high expectations for shot analysis. The intuition is,in general, which is all we're trying to do here, shots from close in have a better chance of going in. I'm hesitant to say I agree with that, but that's the common thinking.

If you're talking about good shooters vs bad shooters, there's what, a 10% difference between the very best and the very worst shooters in the league? If you're looking at shot distance, you're using it as a team metric, i.e. where did most of our shots for/shots against come from? Since most every team has a mix of shooters with high SH% and low SH%, these normalize.

Shot distance weighted for SH% for every players sounds both difficult to isolate and of limited usefulness. Seeing players' SH% at various distances would be...interesting? I guess?

7even is online now   Reply With Quote
Old
07-22-2013, 10:43 AM
  #333
Stephen
Registered User
 
Stephen's Avatar
 
Join Date: Feb 2002
Posts: 31,254
vCash: 500
Quote:
Originally Posted by hatterson View Post
If you want to go ahead and actually track possession in each zone knock yourself out. All you should need to record is all possession changes and the exact time they occurred for each of the 720 games that were played this last year or the 1230 games that are played in a normal 82 game season. Don't worry, it shouldn't take you that long, it's only 30+ days (50+ in a normal season) of hockey.
The NHL tracks ice time already, which I'm not sure how they do, but it doesn't seem to be impossible for them to also track possession.

Stephen is offline   Reply With Quote
Old
07-22-2013, 10:48 AM
  #334
wulfio*
 
Join Date: Jan 2011
Location: Burlington
Country: Canada
Posts: 1,520
vCash: 500
There's too many variables to an individual shot. what kind of shot is the player taking, what are the angles, distance, shooter, who's on the ice against him. Who are his linemates, is the D in position so the goalie doesn't have to worry about back door, or is the goalie cheating, etc..

From a team perspective, what's the system, what's the oppositions defensive scheme.

There's so much flow to this game. I just don't see how you're ever going to develop these advanced stats to the same effect as baseball. Every persons stats will hinge completely on the system he's in, or what role he's put in to. You can't generalize in hockey, it's just not that kind of sport. It's why in hockey we focus on physical/mental abilities, and intangibles over shooting %'s and whatever else. Everything is highly subjective from game to game, season to season, coach to coach. But things that never change are skills and intangibles.

Like, literally everything stats tells you about the New York rangers you can throw out the window. You can't even make an assumption what this team will do because you have nothing to base it on with the new coach. Because this will be a completely different team under AV than it was with the greatest media scrum guy ever.


Last edited by wulfio*: 07-22-2013 at 10:54 AM.
wulfio* is offline   Reply With Quote
Old
07-22-2013, 11:46 AM
  #335
hatterson
Global Moderator
 
hatterson's Avatar
 
Join Date: Apr 2010
Location: North Tonawanda, NY
Country: United States
Posts: 11,290
vCash: 874
Send a message via Skype™ to hatterson
Quote:
Originally Posted by Stephen View Post
The NHL tracks ice time already, which I'm not sure how they do, but it doesn't seem to be impossible for them to also track possession.
It's certainly not impossible, but if they do it now, it's not published anywhere, hence why corsi/fenwick is used as an approximation.

When you have freely available public data that has virtually identical predictive power, it seems silly to waste hundreds/thousands of man-hours compiling actual possession data. I'm sure that in the not so distant future tracking systems will be at the level where it's available, but for now corsi/fenwick is good enough

hatterson 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 06:03 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.