HFBoards

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

Advanced Stats: Corsi, QoC, etc

Reply
 
Thread Tools
Old
02-03-2012, 06:01 PM
  #101
IBleedNYRBlue
Registered User
 
IBleedNYRBlue's Avatar
 
Join Date: May 2011
Location: New Jersey
Country: United States
Posts: 2,780
vCash: 500
Advanced stats in hockey are not very useful like in baseball, but I do wish there was some sort of "Quality SV%" stat. There might be, but never seen it. Stat for quality shots a G faces within the slot, crease, deflections, rebounds, etc. Would be too subjective, though.

IBleedNYRBlue is offline   Reply With Quote
Old
02-03-2012, 06:30 PM
  #102
Jason MacIsaac
MARS Officer
 
Jason MacIsaac's Avatar
 
Join Date: Jan 2004
Location: Victoria, BC
Country: Canada
Posts: 16,462
vCash: 500
Send a message via MSN to Jason MacIsaac Send a message via Yahoo to Jason MacIsaac
Quote:
Originally Posted by RL605 View Post
Advanced stats in hockey are not very useful like in baseball, but I do wish there was some sort of "Quality SV%" stat. There might be, but never seen it. Stat for quality shots a G faces within the slot, crease, deflections, rebounds, etc. Would be too subjective, though.
There is

Jason MacIsaac is offline   Reply With Quote
Old
02-03-2012, 06:32 PM
  #103
Jason MacIsaac
MARS Officer
 
Jason MacIsaac's Avatar
 
Join Date: Jan 2004
Location: Victoria, BC
Country: Canada
Posts: 16,462
vCash: 500
Send a message via MSN to Jason MacIsaac Send a message via Yahoo to Jason MacIsaac
Quote:
Originally Posted by Sivek View Post
Love how any posts concerning sabremetrics in hockey is quickly filled with posts decrying how useless they are. inb4 how using stats=living in mom's basement and never watched or played a game before and so on. People can enjoy stats and the actual game of hockey, it's true.

Anyways, the Hockey Prospectus annual is worth buying if you're interested in sabremetrics. Whether you agree with their opinions or not, having a wider breath of information can only help in evaluating players for your own amusement and gain a better understanding of what certain indicators in a players play may lead to. As noted already, hockey sabremetrics is really behind the ball when compared to other sports since their widespread discussion hasn't been really relevant until recent years, which in my opinion, makes it something interesting to follow since as a fan you can follow how sabremetrics are being crafted at the beginning rather than already having a nearly finished formula, as in baseball.

Also, some stats that really seem to have come on strong in the last couple of years are "Sheltered ice-time" for skaters and even strength save% for goalies if you're looking at some stats to start to with. Sheltered ice time would be extremely useful if there were only a quick & easy way to figure it out.
cough cough Malkin cough. It's amazing how great a match up he gets most nights.

Jason MacIsaac is offline   Reply With Quote
Old
02-03-2012, 06:36 PM
  #104
Erick
Registered User
 
Erick's Avatar
 
Join Date: Nov 2009
Location: Broward, FL
Country: United States
Posts: 11,592
vCash: 500
Quote:
Originally Posted by RL605 View Post
Advanced stats in hockey are not very useful like in baseball, but I do wish there was some sort of "Quality SV%" stat. There might be, but never seen it. Stat for quality shots a G faces within the slot, crease, deflections, rebounds, etc. Would be too subjective, though.
Pretty sure that's what DIGR is for.

Erick is offline   Reply With Quote
Old
02-03-2012, 07:10 PM
  #105
number72
Registered User
 
Join Date: Oct 2011
Posts: 5,883
vCash: 500
Quote:
Originally Posted by Sivek View Post
Love how any posts concerning sabremetrics in hockey is quickly filled with posts decrying how useless they are. inb4 how using stats=living in mom's basement and never watched or played a game before and so on. People can enjoy stats and the actual game of hockey, it's true.

Anyways, the Hockey Prospectus annual is worth buying if you're interested in sabremetrics. Whether you agree with their opinions or not, having a wider breath of information can only help in evaluating players for your own amusement and gain a better understanding of what certain indicators in a players play may lead to. As noted already, hockey sabremetrics is really behind the ball when compared to other sports since their widespread discussion hasn't been really relevant until recent years, which in my opinion, makes it something interesting to follow since as a fan you can follow how sabremetrics are being crafted at the beginning rather than already having a nearly finished formula, as in baseball.

Also, some stats that really seem to have come on strong in the last couple of years are "Sheltered ice-time" for skaters and even strength save% for goalies if you're looking at some stats to start to with. Sheltered ice time would be extremely useful if there were only a quick & easy way to figure it out.
I found problems with every stat +-, +-/60, corsi or fenwick etc.
The advanced stats do work better but there are interesting exceptions.
They are better then say +- but are prone to similar errors (just less often and to a much lesser degree).

In fact, the correlation is really not that high - in that it turns out luck is a more important factor when player skill levels are comparable.
It is the best there is but it is not perfect. Just a number that is descriptive play and skill 40% of the time and the other 60% at the mercy of "luck" or random variation or the hockey gods.

number72 is offline   Reply With Quote
Old
02-03-2012, 07:19 PM
  #106
SMantzas
Church of Corsi
 
SMantzas's Avatar
 
Join Date: Mar 2011
Location: Mount Laurel, NJ
Country: United States
Posts: 3,100
vCash: 500
Im not too knowledgeable in this area, but it is really interesting; especially QUALCOMP. For example, it shows that the Sedins and Malkin get protected, while guys like Elias and Kopitar face the toughest competition and excel while doing it

SMantzas is online now   Reply With Quote
Old
02-03-2012, 07:24 PM
  #107
Big McLargehuge
Global Moderator
Buff Drinklots
 
Big McLargehuge's Avatar
 
Join Date: May 2002
Location: Los Angeles
Country: Germany
Posts: 54,266
vCash: 422
We just had a discussion about this stuff on the Pens board.

I'm not a fan of them for hockey because, unlike baseball, it's not a zero-sum game. Even the best of hockey stats are open to a bevy of outside influences that just can't be accounted for like they can in baseball.

I love stats, it's part of the reason why I still love baseball despite the fact that my favorite team has never had a .500 season since I was old enough to understand the concept of sports...but stats just don't work as well in hockey as in many other sports.

__________________
“The most terrifying fact about the universe is not that it is hostile, but that it is indifferent. If we can come to terms with this indifference and accept the challenges of life within the boundaries of death, our existence as a species can have genuine meaning and fulfillment. However vast the darkness, we must supply our own light.” - Stanley Kubrick
http://sprites.pokecheck.org/i/054.gif
Big McLargehuge is offline   Reply With Quote
Old
02-03-2012, 07:38 PM
  #108
Talks to Goalposts
Registered User
 
Join Date: Apr 2011
Location: Edmonton
Country: Canada
Posts: 3,610
vCash: 500
Quote:
Originally Posted by Big McLargehuge View Post
We just had a discussion about this stuff on the Pens board.

I'm not a fan of them for hockey because, unlike baseball, it's not a zero-sum game. Even the best of hockey stats are open to a bevy of outside influences that just can't be accounted for like they can in baseball.

I love stats, it's part of the reason why I still love baseball despite the fact that my favorite team has never had a .500 season since I was old enough to understand the concept of sports...but stats just don't work as well in hockey as in many other sports.
What gets lost in a lot of these discussions is that "not as good as baseball stats" does not equal "worthless."

Also the a priori assumption that they are not good so you can dismiss them without even looking at them or making an attempt to understand what they are when they are brought up.

Talks to Goalposts is offline   Reply With Quote
Old
02-03-2012, 08:21 PM
  #109
Brent Burns
TROTZTROTZTROTZTROTZ
 
Brent Burns's Avatar
 
Join Date: Feb 2007
Location: Coming in hot
Posts: 6,564
vCash: 500
Love how every kid on HF comes into these threads crying foul when someone mentions stats. Don't lump this stuff in with the "Ovechkin scored x more goals than Toews therefore he is better" crowd

Brent Burns is offline   Reply With Quote
Old
02-03-2012, 08:24 PM
  #110
Elever
Hth
 
Elever's Avatar
 
Join Date: Jan 2010
Country: Canada
Posts: 10,459
vCash: 500
I'm not a fan either now although I used to trust them. Was reading one article for example that was looking at the value of defensive dmen and Jan Hejda was number one on the list based on this stat they came up with. I know from watching Jan Hejda play a lot as an Avs fan that he can be pretty brutal.

Elever is offline   Reply With Quote
Old
02-03-2012, 08:30 PM
  #111
Erick
Registered User
 
Erick's Avatar
 
Join Date: Nov 2009
Location: Broward, FL
Country: United States
Posts: 11,592
vCash: 500
Quote:
Originally Posted by Brent Burns View Post
Love how every kid on HF comes into these threads crying foul when someone mentions stats. Don't lump this stuff in with the "Ovechkin scored x more goals than Toews therefore he is better" crowd
If anything, I'd say the old-timers have a tougher time accepting advanced statistics because they've been watching the game for so long.

Erick is offline   Reply With Quote
Old
02-03-2012, 08:37 PM
  #112
Big McLargehuge
Global Moderator
Buff Drinklots
 
Big McLargehuge's Avatar
 
Join Date: May 2002
Location: Los Angeles
Country: Germany
Posts: 54,266
vCash: 422
Quote:
Originally Posted by Talks to Goalposts View Post
What gets lost in a lot of these discussions is that "not as good as baseball stats" does not equal "worthless."

Also the a priori assumption that they are not good so you can dismiss them without even looking at them or making an attempt to understand what they are when they are brought up.
I didn't say they're useless, I just said I'm not a fan of them because they're not as perfect as their advocates seem to claim they are.

There's two sides to every story in hockey...whereas in baseball stats can be fairly black and white math whenever you can adjust for things (player A plays in a hitter friendly park and player B plays in a pitcher friendly park...we have the stats to adjust for them and neutralize them...in hockey you can't really adjust for linemates/goaltenders faced/whatever nearly as well...and even on certain stats that some fans seem to love and others hate, such as faceoff zones, they generally have simple answers. When you have Jordan Staal on your roster he's going to take the bulk of the defensive zone faceoffs. That doesn't mean the other guys on the roster aren't good defensively or that Staal is bad offensively, it just says Staal plays a shutdown role...something that I think we all know by this point.).

I mean, baseball stats aren't perfect either...but as a zero-sum game they're closer to telling the whole story if you dig into things...that and adjusting for outside influences in hockey is next to impossible, but is just assumed in just about every baseball sabremetric stat. When a hockey equivalent of FIP or WAR comes around I will cry it's praises from the top of Mt. Sentinel. As it stands none of those exist...but grasping to the old standards isn't a great idea either. I mean, for what it's worth I think +/- is the worst stat in sports...and I'm including pitcher W-L record in there. Just about any sabremetric defensive stat is more credible than that.


Last edited by Big McLargehuge: 02-03-2012 at 08:46 PM.
Big McLargehuge is offline   Reply With Quote
Old
02-03-2012, 08:39 PM
  #113
Brent Burns
TROTZTROTZTROTZTROTZ
 
Brent Burns's Avatar
 
Join Date: Feb 2007
Location: Coming in hot
Posts: 6,564
vCash: 500
Quote:
Originally Posted by Erick View Post
If anything, I'd say the old-timers have a tougher time accepting advanced statistics because they've been watching the game for so long.
HFBoards has a trickle down effect. When one older member posts something that they think makes sense, all the kids jump on the bandwagon and swear by it until the next sensical poster comes along

Brent Burns is offline   Reply With Quote
Old
02-04-2012, 01:25 AM
  #114
Erick
Registered User
 
Erick's Avatar
 
Join Date: Nov 2009
Location: Broward, FL
Country: United States
Posts: 11,592
vCash: 500
Quote:
Originally Posted by Big McLargehuge View Post
I didn't say they're useless, I just said I'm not a fan of them because they're not as perfect as their advocates seem to claim they are.

There's two sides to every story in hockey...whereas in baseball stats can be fairly black and white math whenever you can adjust for things (player A plays in a hitter friendly park and player B plays in a pitcher friendly park...we have the stats to adjust for them and neutralize them...in hockey you can't really adjust for linemates/goaltenders faced/whatever nearly as well...and even on certain stats that some fans seem to love and others hate, such as faceoff zones, they generally have simple answers. When you have Jordan Staal on your roster he's going to take the bulk of the defensive zone faceoffs. That doesn't mean the other guys on the roster aren't good defensively or that Staal is bad offensively, it just says Staal plays a shutdown role...something that I think we all know by this point.).

I mean, baseball stats aren't perfect either...but as a zero-sum game they're closer to telling the whole story if you dig into things...that and adjusting for outside influences in hockey is next to impossible, but is just assumed in just about every baseball sabremetric stat. When a hockey equivalent of FIP or WAR comes around I will cry it's praises from the top of Mt. Sentinel. As it stands none of those exist...but grasping to the old standards isn't a great idea either. I mean, for what it's worth I think +/- is the worst stat in sports...and I'm including pitcher W-L record in there. Just about any sabremetric defensive stat is more credible than that.

GVT seems rather similar to WAR in baseball.

http://hockeyprospectus.com/article.php?articleid=233

Erick is offline   Reply With Quote
Old
02-04-2012, 03:33 AM
  #115
rye&ginger
Registered User
 
rye&ginger's Avatar
 
Join Date: Jan 2004
Location: Vancouver
Country: Canada
Posts: 6,156
vCash: 500
Like pretty much any stat, you can take them out of context of other related stats. And they are statistics; They dont predict the future, but tell you about the past.

I like the sabremetric stuff but it should not be a major discussion point in hockey. Understanding how plays develop and how a player is supposed to make a decision is more interesting and useful to understand. That is scouting and coaching mostly. Stats are quantifications of that.

rye&ginger is offline   Reply With Quote
Old
02-04-2012, 04:17 AM
  #116
Malkin4Top6Wingerz
Can you like, shutup
 
Malkin4Top6Wingerz's Avatar
 
Join Date: Mar 2009
Country: United States
Posts: 4,950
vCash: 500
Quote:
Originally Posted by rye&ginger View Post
And they are statistics; They dont predict the future, but tell you about the past.
Hockey sabremetrics are actually designed to project future performance, and while it's hardly a flawless approach, I would say that they do a much better job of this than any fan or analyst who relies on watching games.

Quote:
I like the sabremetric stuff but it should not be a major discussion point in hockey. Understanding how plays develop and how a player is supposed to make a decision is more interesting and useful to understand. That is scouting and coaching mostly. Stats are quantifications of that.
What me and many others like about advanced statistics in hockey is the complete lack of user bias within the data. You don't get that viewpoint from watching games, particularly from your own team. Even somebody who tries to follow the entire league is only going to get a microscopic picture of a player or team because of the limited sample size that they're disecting. The other issue is that we don't know how valuable any random internet poster's subjective opinion is.

What I find interesting is that many of the people who disregard sabremetrics in hockey do use statistics to form opinions and make arguments, they just use the least useful ones.

Malkin4Top6Wingerz is offline   Reply With Quote
Old
02-04-2012, 04:44 AM
  #117
rye&ginger
Registered User
 
rye&ginger's Avatar
 
Join Date: Jan 2004
Location: Vancouver
Country: Canada
Posts: 6,156
vCash: 500
I actually do refer to stats often but to say they can 'predict the future' is a bit of a stretch. Nothing can really do that, you can take stats and other qualitative info like 'player X is battling hard lately and can turn the tide of the game'.

And I agree, I look at the more advanced stats before much else. GF and GA are the best team stats out of the basic ones, not points/wins/losses. The advanced stats take it even further of course. We need longer term data to really see how valuable it is though. A few seasons is really not long enough just yet.

This stuff is over the casual fans head so we cant expect it to be discussed on TV. Its made for online.

rye&ginger is offline   Reply With Quote
Old
02-04-2012, 05:06 AM
  #118
Jason MacIsaac
MARS Officer
 
Jason MacIsaac's Avatar
 
Join Date: Jan 2004
Location: Victoria, BC
Country: Canada
Posts: 16,462
vCash: 500
Send a message via MSN to Jason MacIsaac Send a message via Yahoo to Jason MacIsaac
Quote:
Originally Posted by rye&ginger View Post
I actually do refer to stats often but to say they can 'predict the future' is a bit of a stretch. Nothing can really do that, you can take stats and other qualitative info like 'player X is battling hard lately and can turn the tide of the game'.

And I agree, I look at the more advanced stats before much else. GF and GA are the best team stats out of the basic ones, not points/wins/losses. The advanced stats take it even further of course. We need longer term data to really see how valuable it is though. A few seasons is really not long enough just yet.

This stuff is over the casual fans head so we cant expect it to be discussed on TV. Its made for online.
Fenwick has a .62 correlation with future scoring at the team level. In terms of predicting that has a strong correlation.

Jason MacIsaac is offline   Reply With Quote
Old
02-04-2012, 08:55 AM
  #119
number72
Registered User
 
Join Date: Oct 2011
Posts: 5,883
vCash: 500
Quote:
Originally Posted by Jason MacIsaac View Post
Fenwick has a .62 correlation with future scoring at the team level. In terms of predicting that has a strong correlation.
The studies I have seen shown an
r = .62
rsquared correlation = .62x.62 = 0.384

Fenwick Road Tied is the best measure that exists to describe team skill but even so it fails to explain over 60% why teams win.

number72 is offline   Reply With Quote
Old
02-04-2012, 10:05 AM
  #120
almostawake
Registered User
 
Join Date: Jan 2006
Posts: 3,187
vCash: 500
Attempts at quantitative analysis in Hockey generally suffer from two problems. First it is much more difficult to decouple the effect of one player from the other players on the ice. Second, the statistics that have been traditionally recorded (Goals, Assists, etc) cannot be used to generate statistics that are as predictive as those that have been found for Baseball.

The second problem is pretty easy to solve, hire people to watch every game and record different statistics that are more meaningful.

The first problem is a bit more difficult but many scientific disciplines have over come much much more complicated (and complex) problems. With a background physics/biology/economics I can think of at least a half dozen techniques off the top of my head that have the potential to work.

Despite their popularity today, Sabremetrics still struggle to say much about pitching, one of the main elements of the baseball. Furthermore, 15+ years ago there was a lot of debate in the Sabremetrics community as to whether or not a useful statistic could ever be arrived at for fielding. People tried everything with the traditionally collected statistics and got nothing (range is garbage IMO). But then some people started to collect new positioning data and they basically broke the problem open.

IMO, high level quantitative analysis may or may not work in Hockey. There are a lot of challenges and there are a lot of techniques that may or may not be able to meet those challenges. What I can say with absolute certainty is that any person that tells you that they definitely can/cannot work is either full of **** or the most intelligent human being on earth.

almostawake is online now   Reply With Quote
Old
02-04-2012, 10:35 AM
  #121
Mystlyfe
We're Touched
 
Join Date: Oct 2010
Location: Philadelphia
Posts: 13,008
vCash: 500
Sabremetrics in hockey are still in their infant state compared to baseball. We're reaching the point where data collection is becoming the new bottleneck, as we've done a lot of analysis and reconstructing of existing stats into something more useful, but are hitting the point where the next level of stats are going to need more detailed data. Tracking scoring chances and defensive errors are some of the new trends among stat-related team bloggers, and will help lead to some more statistics to utilize.

Mystlyfe is online now   Reply With Quote
Old
02-04-2012, 11:09 AM
  #122
The Fear Boners
Plz stop pucks
 
The Fear Boners's Avatar
 
Join Date: Jun 2007
Location: Scrip Club
Country: United States
Posts: 20,057
vCash: 500
Quote:
Originally Posted by Big McLargehuge View Post
I didn't say they're useless, I just said I'm not a fan of them because they're not as perfect as their advocates seem to claim they are.

There's two sides to every story in hockey...whereas in baseball stats can be fairly black and white math whenever you can adjust for things (player A plays in a hitter friendly park and player B plays in a pitcher friendly park...we have the stats to adjust for them and neutralize them...in hockey you can't really adjust for linemates/goaltenders faced/whatever nearly as well...and even on certain stats that some fans seem to love and others hate, such as faceoff zones, they generally have simple answers. When you have Jordan Staal on your roster he's going to take the bulk of the defensive zone faceoffs. That doesn't mean the other guys on the roster aren't good defensively or that Staal is bad offensively, it just says Staal plays a shutdown role...something that I think we all know by this point.).

I mean, baseball stats aren't perfect either...but as a zero-sum game they're closer to telling the whole story if you dig into things...that and adjusting for outside influences in hockey is next to impossible, but is just assumed in just about every baseball sabremetric stat. When a hockey equivalent of FIP or WAR comes around I will cry it's praises from the top of Mt. Sentinel. As it stands none of those exist...but grasping to the old standards isn't a great idea either. I mean, for what it's worth I think +/- is the worst stat in sports...and I'm including pitcher W-L record in there. Just about any sabremetric defensive stat is more credible than that.
A very accurate and good assessment.

At least with the stats being somewhat flawed at this point, we have a good starting point to build on different ideas.

Baseball has it so easy...

The Fear Boners is offline   Reply With Quote
Old
02-04-2012, 11:15 AM
  #123
The Fear Boners
Plz stop pucks
 
The Fear Boners's Avatar
 
Join Date: Jun 2007
Location: Scrip Club
Country: United States
Posts: 20,057
vCash: 500
Quote:
Originally Posted by Malkin4Top6Wingerz View Post
Hockey sabremetrics are actually designed to project future performance, and while it's hardly a flawless approach, I would say that they do a much better job of this than any fan or analyst who relies on watching games.
Absolutely.
Quote:
What me and many others like about advanced statistics in hockey is the complete lack of user bias within the data. You don't get that viewpoint from watching games, particularly from your own team. Even somebody who tries to follow the entire league is only going to get a microscopic picture of a player or team because of the limited sample size that they're disecting. The other issue is that we don't know how valuable any random internet poster's subjective opinion is.

What I find interesting is that many of the people who disregard sabremetrics in hockey do use statistics to form opinions and make arguments, they just use the least useful ones.
I do enjoy the fact that the stats don't lie, they do tell you of past performance and project future performance but that analysis of a player offensive or defensively will still be ever so slightly flawed until we figure out how to better judge team defensive and offensive systems.

The Fear Boners is offline   Reply With Quote
Old
02-04-2012, 12:00 PM
  #124
cutchemist42
Registered User
 
cutchemist42's Avatar
 
Join Date: Apr 2011
Location: Winnipeg
Country: Canada
Posts: 2,987
vCash: 500
Quote:
Originally Posted by Erick View Post
Anyone interested in hockey analytics?

What's the best site?

I've seen hockeyprospectus but it's not updated, and you gotta pay $20 apparently for that.

Behind the Net seems pretty popular for the CORSI stat, but I was looking for other things (maybe I'm missing them).

Anyone recommend a better site if there is one? Thanks.
hockeyprospectus' GVT is free to view.

cutchemist42 is offline   Reply With Quote
Old
02-04-2012, 12:13 PM
  #125
cutchemist42
Registered User
 
cutchemist42's Avatar
 
Join Date: Apr 2011
Location: Winnipeg
Country: Canada
Posts: 2,987
vCash: 500
Quote:
Originally Posted by Erick View Post
If anything, I'd say the old-timers have a tougher time accepting advanced statistics because they've been watching the game for so long.
Baseball still has this problem with old time baseball writers.

cutchemist42 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 10:30 AM.

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.