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.

Brian Burke is a Dinosaur

Reply
 
Thread Tools
Old
03-06-2013, 01:03 PM
  #151
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,387
vCash: 500
Quote:
Originally Posted by Man Bear Pig View Post
The post below answers your question. The RedSox bailed on the moneyball approach and went the way of their rivals, the Yankees. Spend tons of cash and pray for the best.
This is so wrong, I don't know where to begin.

Theo was hired in 2003. AFTER Manny. AFTER Pedro.

The Legend is offline   Reply With Quote
Old
03-06-2013, 02:06 PM
  #152
Kyle Doobas*
#groovystats
 
Join Date: Oct 2009
Location: Toronto, ON
Country: Canada
Posts: 8,689
vCash: 159
Quote:
Originally Posted by TOML View Post
Um, wouldn't a lamppost still be useful for illumination as well as support, drunk or not?


TOML
.....really?

The idea is that you can't see anyway.

Kyle Doobas* is offline   Reply With Quote
Old
03-06-2013, 03:39 PM
  #153
Mowerman
Registered User
 
Mowerman's Avatar
 
Join Date: Oct 2010
Location: Toronto
Country: Canada
Posts: 1,571
vCash: 500
Hockey doesn't lend itself too well to advanced statistics. Maybe if you're on a hockey forum and you're trying to justify an argument with more than just your eye quickly. Not really if you are analyzing players in the show. Trained eyes are what counts there. Don't really see what's so controversial about any of his quotes if you limit them to the realm of hockey (which I'm assuming was the context... cherrypicked quotes like this are worthless).

Mowerman is offline   Reply With Quote
Old
03-06-2013, 04:06 PM
  #154
Darylman
Registered User
 
Darylman's Avatar
 
Join Date: Aug 2009
Posts: 1,646
vCash: 500
Quote:
Originally Posted by FrontalLombardomy View Post
Hockey doesn't lend itself too well to advanced statistics. Maybe if you're on a hockey forum and you're trying to justify an argument with more than just your eye quickly. Not really if you are analyzing players in the show. Trained eyes are what counts there. Don't really see what's so controversial about any of his quotes if you limit them to the realm of hockey (which I'm assuming was the context... cherrypicked quotes like this are worthless).
I agree. Burke is making a very black and white statement, he's saying stats are useless if you aren't already carefully watching a player.

Take some random prospect overseas, if you go SIMPLY by his statline, however complex you make it out to be, this is no substitute for actual eyewitness scouting. I don't see how this is so controversial. Advanced stats give nuance to observations, but they don't really make them on their own.

Darylman is offline   Reply With Quote
Old
03-06-2013, 04:26 PM
  #155
GuineaPig
Registered User
 
GuineaPig's Avatar
 
Join Date: Jul 2011
Location: Montréal
Posts: 2,143
vCash: 500
I get why people are hesitant to say it, but I will anyways: over a large sample (let's call it 100+ games at least) I absolutely think using stats is better than watching a given player in the NHL.

GuineaPig is offline   Reply With Quote
Old
03-06-2013, 11:53 PM
  #156
Noldo
Registered User
 
Join Date: May 2007
Posts: 1,077
vCash: 500
One thing were advance stats will excel (once they have been sufficiently refined) is eliminating any style bias still existing in hufe amounts in hockey. Someone brought up that stat can not measure the effect of Stevens' of Pronger's physicality, but quite contrary the stats won't care how a player does the job if his style is effective. Most likely result being that the GMs investing in advance stats will be able to identify unorthodox/un-flashy players that contribute more than their style would suggest while avoiding players that seem good based on traditional views. Further, the stats isn't really needed to identify the very best, they stand out anyway. But stats will most likely help GMs to identify the most effective support players, allowing teams to gather more depth (and the depth wins championships) or leave more space to be allocated to stars.

Noldo is offline   Reply With Quote
Old
03-07-2013, 12:59 AM
  #157
Man Bear Pig
Registered User
 
Man Bear Pig's Avatar
 
Join Date: Aug 2008
Location: Earth
Posts: 10,676
vCash: 500
Quote:
Originally Posted by The Legend View Post
This is so wrong, I don't know where to begin.

Theo was hired in 2003. AFTER Manny. AFTER Pedro.
Not wrong. Look at what Theo did. He went after big priced players. He certainly wasn't in a hurry to start cutting payroll was he? He signed Drew,Dice-k,Crawford,Lackey,Cameron,Gagne,Lugo,Renteria etc. Not exactly a Billy Beane kind of approach.

Man Bear Pig is offline   Reply With Quote
Old
03-07-2013, 09:45 AM
  #158
Joe Canadian
Registered User
 
Join Date: May 2004
Location: St. John's, NL, CAN
Country: Canada
Posts: 162
vCash: 500
Quote:
Originally Posted by Man Bear Pig View Post
Not wrong. Look at what Theo did. He went after big priced players. He certainly wasn't in a hurry to start cutting payroll was he? He signed Drew,Dice-k,Crawford,Lackey,Cameron,Gagne,Lugo,Renteria etc. Not exactly a Billy Beane kind of approach.
Of course none of those players had anything to do with the 2004 Red Sox (except for Renteria who was the Cardinal batter who grounded out to end the World Series that year). And Epstein wouldn't be the first GM to stray away from one approach in favour of throwing money around, nor was he the last. Your only evidence that the team wasn't a Moneyball team seems to be the payroll totals. That ignores the decision making process that went into filling out the other roster spots. Moneyball isn't a synonym for low-payroll team, high payroll teams can deploy it. Especially when said teams don't have unlimited funds and need to fill out a roster beyond those high priced players.

Joe Canadian is offline   Reply With Quote
Old
03-07-2013, 11:53 AM
  #159
Bad Will Hunting
It's my fault
 
Bad Will Hunting's Avatar
 
Join Date: Oct 2007
Location: Southie
Posts: 26,520
vCash: 69
Quote:
Originally Posted by Man Bear Pig View Post
Not wrong. Look at what Theo did. He went after big priced players. He certainly wasn't in a hurry to start cutting payroll was he? He signed Drew,Dice-k,Crawford,Lackey,Cameron,Gagne,Lugo,Renteria etc. Not exactly a Billy Beane kind of approach.
As pointed out ...those were all after the 2004 win when ownership put pressure on him to win again and spend what was needed to do it


Last edited by Bad Will Hunting: 03-07-2013 at 12:34 PM.
Bad Will Hunting is online now   Reply With Quote
Old
03-07-2013, 12:21 PM
  #160
Jeri
Registered User
 
Jeri's Avatar
 
Join Date: Dec 2009
Posts: 376
vCash: 500
It will be interesting to see if hockey comes up with an equivalent WAR stat to baseball in the next couple of years.

Jeri is offline   Reply With Quote
Old
03-07-2013, 12:31 PM
  #161
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,387
vCash: 500
Quote:
Originally Posted by Jeri View Post
It will be interesting to see if hockey comes up with an equivalent WAR stat to baseball in the next couple of years.
As a gigantic stathead, even if I have to admit it's almost impossible. Baseball is a individual sport masquerading as a team sport - hockey has actions that are much more interrelated.

I would like someone to debunk the "momentum"myth.

The Legend is offline   Reply With Quote
Old
03-07-2013, 01:07 PM
  #162
Jeri
Registered User
 
Jeri's Avatar
 
Join Date: Dec 2009
Posts: 376
vCash: 500
Quote:
Originally Posted by The Legend View Post
As a gigantic stathead, even if I have to admit it's almost impossible. Baseball is a individual sport masquerading as a team sport - hockey has actions that are much more interrelated.

I would like someone to debunk the "momentum"myth.
Oh definitely, it is near impossible to come up with, because there is such a limited number of variables in baseball compared to hockey, but even a flawed statistic such as WAR (although still useful) would be a decent all encompassing indicator.

How exactly is "momentum" a myth? It's one of those more intangible "forces" wouldn't you say? A fight or two quick goals have the effect of getting a team pumped the same way a diving catch or a 2 out steal of 3rd base would have in a baseball game?

Jeri is offline   Reply With Quote
Old
03-07-2013, 03:01 PM
  #163
Mathletic
Registered User
 
Mathletic's Avatar
 
Join Date: Feb 2002
Location: St-Augustin, Québec
Country: Canada
Posts: 13,224
vCash: 500
Quote:
Originally Posted by Jeri View Post
Oh definitely, it is near impossible to come up with, because there is such a limited number of variables in baseball compared to hockey, but even a flawed statistic such as WAR (although still useful) would be a decent all encompassing indicator.

How exactly is "momentum" a myth? It's one of those more intangible "forces" wouldn't you say? A fight or two quick goals have the effect of getting a team pumped the same way a diving catch or a 2 out steal of 3rd base would have in a baseball game?
there have been studies on fights. I know one guy looked for the rate of shots after a fight. Another looked for goal scoring rate vs expected goal scoring rate. Neither could find momentum effect due to a fight. Sure, there's times when a team scores after a fight. But it doesn't happen more often than when there's no fight.

Mathletic is offline   Reply With Quote
Old
03-07-2013, 11:55 PM
  #164
Mathletic
Registered User
 
Mathletic's Avatar
 
Join Date: Feb 2002
Location: St-Augustin, Québec
Country: Canada
Posts: 13,224
vCash: 500
couple interesting articles on hockey analytics

http://nhlnumbers.com/2013/3/7/why-s...ant-item-45604

and about the MIT conference

http://www.cbc.ca/sports/hockey/opin...tatistics.html

Mathletic is offline   Reply With Quote
Old
03-08-2013, 01:46 AM
  #165
GuineaPig
Registered User
 
GuineaPig's Avatar
 
Join Date: Jul 2011
Location: Montréal
Posts: 2,143
vCash: 500
Quote:
Originally Posted by Jeri View Post
It will be interesting to see if hockey comes up with an equivalent WAR stat to baseball in the next couple of years.
http://puckprospectus.com/sortable/

GuineaPig is offline   Reply With Quote
Old
03-08-2013, 06:09 AM
  #166
Ozamataz Buckshank
Registered User
 
Ozamataz Buckshank's Avatar
 
Join Date: Oct 2010
Location: Massachusetts
Country: United States
Posts: 4,456
vCash: 500
Quote:
Originally Posted by Joe Canadian View Post
Of course none of those players had anything to do with the 2004 Red Sox (except for Renteria who was the Cardinal batter who grounded out to end the World Series that year). And Epstein wouldn't be the first GM to stray away from one approach in favour of throwing money around, nor was he the last. Your only evidence that the team wasn't a Moneyball team seems to be the payroll totals. That ignores the decision making process that went into filling out the other roster spots. Moneyball isn't a synonym for low-payroll team, high payroll teams can deploy it. Especially when said teams don't have unlimited funds and need to fill out a roster beyond those high priced players.
Actually...it is. Moneyball is not about just using advanced statistics, it's about using them to find value in players to save money while still contending.

Ozamataz Buckshank is online now   Reply With Quote
Old
03-08-2013, 07:12 AM
  #167
Man Bear Pig
Registered User
 
Man Bear Pig's Avatar
 
Join Date: Aug 2008
Location: Earth
Posts: 10,676
vCash: 500
Quote:
Originally Posted by Joe Canadian View Post
Of course none of those players had anything to do with the 2004 Red Sox (except for Renteria who was the Cardinal batter who grounded out to end the World Series that year). And Epstein wouldn't be the first GM to stray away from one approach in favour of throwing money around, nor was he the last. Your only evidence that the team wasn't a Moneyball team seems to be the payroll totals. That ignores the decision making process that went into filling out the other roster spots. Moneyball isn't a synonym for low-payroll team, high payroll teams can deploy it. Especially when said teams don't have unlimited funds and need to fill out a roster beyond those high priced players.
Moneyball applies more towards low-budget teams, it kind of defeats the purpose of great value when you sign guys for $20 million a season. The point is to use the advanced stats to find cheap players as the alternative. Guys like Beane are geniuses because they're great at searching the bargain bin by using advanced statistics and maximizing the asset. Boston just signs and trades for everybody in sight. That's not moneyball my friend.

Man Bear Pig is offline   Reply With Quote
Old
03-08-2013, 09:14 AM
  #168
oilersrule14
Registered User
 
Join Date: May 2003
Posts: 1,464
vCash: 500
Quote:
Originally Posted by Noldo View Post
One thing were advance stats will excel (once they have been sufficiently refined) is eliminating any style bias still existing in hufe amounts in hockey. Someone brought up that stat can not measure the effect of Stevens' of Pronger's physicality, but quite contrary the stats won't care how a player does the job if his style is effective. Most likely result being that the GMs investing in advance stats will be able to identify unorthodox/un-flashy players that contribute more than their style would suggest while avoiding players that seem good based on traditional views. Further, the stats isn't really needed to identify the very best, they stand out anyway. But stats will most likely help GMs to identify the most effective support players, allowing teams to gather more depth (and the depth wins championships) or leave more space to be allocated to stars.
But that still hasn't addressed whether stats can measure the effect of physicality.

You mentioned an alternative point to support the stats, and I think it's very valid. However, there's no stat that can tell me if the physicality of a Pronger has a positive effect on his teammates, in terms of opening up room for them or allowing them to play at ease.

Any stats attempting to do so are going to suffer from the fact that you can't separate Pronger's effectiveness of play from his physical play, nor can you separate his teammates's regular effectiveness of play from any increased effectiveness of play due to physicality.

I think that's an overlooked aspect of the game by the talking heads of the statisticians groups, and an overemphasized aspect of the game by the talking heads of the no-stats group. Usually, it's the quieter guys not causing a stir that put fair emphasis on all aspects.

oilersrule14 is offline   Reply With Quote
Old
03-08-2013, 09:47 AM
  #169
kmad
Riot Survivor
 
kmad's Avatar
 
Join Date: Jun 2003
Location: Vancouver
Country: Canada
Posts: 33,034
vCash: 500
Quote:
Originally Posted by KesselForSelke View Post
Actually...it is. Moneyball is not about just using advanced statistics, it's about using them to find value in players to save money while still contending.
Sabremetrics is using advanced statistics in sports, moneyball is using those statistics to find undervalued players.

kmad is offline   Reply With Quote
Old
03-08-2013, 10:30 AM
  #170
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,387
vCash: 500
Quote:
Originally Posted by KesselForSelke View Post
Actually...it is. Moneyball is not about just using advanced statistics, it's about using them to find value in players to save money while still contending.
Moneyball is about market inefficiencies.

The Legend is offline   Reply With Quote
Old
03-08-2013, 11:07 AM
  #171
Mathletic
Registered User
 
Mathletic's Avatar
 
Join Date: Feb 2002
Location: St-Augustin, Québec
Country: Canada
Posts: 13,224
vCash: 500
Quote:
Originally Posted by oilersrule14 View Post
But that still hasn't addressed whether stats can measure the effect of physicality.

You mentioned an alternative point to support the stats, and I think it's very valid. However, there's no stat that can tell me if the physicality of a Pronger has a positive effect on his teammates, in terms of opening up room for them or allowing them to play at ease.

Any stats attempting to do so are going to suffer from the fact that you can't separate Pronger's effectiveness of play from his physical play, nor can you separate his teammates's regular effectiveness of play from any increased effectiveness of play due to physicality.

I think that's an overlooked aspect of the game by the talking heads of the statisticians groups, and an overemphasized aspect of the game by the talking heads of the no-stats group. Usually, it's the quieter guys not causing a stir that put fair emphasis on all aspects.
You're asking for the wrong questions. There's no point in measuring physicality. It's like asking how much batters were intimidated when facing Randy Johnson. If the guy is afraid after getting pitched inside or simply facing RJ then he'd a less effective batter or RJ a more effective pitcher, period. If Pronger's physicality helps him being a more effective defender, then it will show in his stats.

Randy Johnson gave up less HR, had more strike outs, better DICE/DIPS, better Pitch FX and all you want. No one ever attemped measuring fear in the batter's eye.

Also, if you read the first article I mentionned, the author touches a bit on those subjects.

Quote:
Originally Posted by Man Bear Pig View Post
Moneyball applies more towards low-budget teams, it kind of defeats the purpose of great value when you sign guys for $20 million a season. The point is to use the advanced stats to find cheap players as the alternative. Guys like Beane are geniuses because they're great at searching the bargain bin by using advanced statistics and maximizing the asset. Boston just signs and trades for everybody in sight. That's not moneyball my friend.
The value a player brings to a team is relative to each market. The value of Manny Ramirez was not the same for Boston than it would have been to Oakland for example. Those are things that have been studied quite extensively. The value of the 90th win for a team like Boston is worth about 3 million IIRC. That's quite a bit. A player's performance will impact on team wins, which then in turn have an impact on attendance/tv ratings, which lead to various revenues. The revenues generated by Manny Ramirez in Boston were well beyond what they paid for him. Analytics will help you figuring out how much that is worth and put a number on value for each dollar invested. It's not a black or white thing where stats have value for low budget teams and don't for high budget teams.


Last edited by Mathletic: 03-08-2013 at 12:30 PM.
Mathletic is offline   Reply With Quote
Old
03-08-2013, 04:33 PM
  #172
Joe Canadian
Registered User
 
Join Date: May 2004
Location: St. John's, NL, CAN
Country: Canada
Posts: 162
vCash: 500
Quote:
Originally Posted by KesselForSelke View Post
Actually...it is. Moneyball is not about just using advanced statistics, it's about using them to find value in players to save money while still contending.
No. It's not. And the reasoning you stated doesn't back up your point, though it is correct. Moneyball is about finding market ineffcienies. Teams that have $140 Million budget can have a high payroll, with large contracts, and can still deploy Moneyball techniques to fill out their roster. Unless a team has an unlimited payroll, which the 2004 Red Sox did not, there are always going to be roster spots that GMs need to fill with limited funds.

Joe Canadian is offline   Reply With Quote
Old
03-08-2013, 11:02 PM
  #173
Hero
Buffy is Back
 
Hero's Avatar
 
Join Date: Jul 2009
Posts: 20,148
vCash: 500
Baseball is a 1 vs 1 scenario repeated many times. very easy to keep stats and compare them, not so with hockey.

Hero is offline   Reply With Quote
Old
03-09-2013, 05:55 AM
  #174
Noldo
Registered User
 
Join Date: May 2007
Posts: 1,077
vCash: 500
Quote:
Originally Posted by oilersrule14 View Post
But that still hasn't addressed whether stats can measure the effect of physicality.

You mentioned an alternative point to support the stats, and I think it's very valid. However, there's no stat that can tell me if the physicality of a Pronger has a positive effect on his teammates, in terms of opening up room for them or allowing them to play at ease.

Any stats attempting to do so are going to suffer from the fact that you can't separate Pronger's effectiveness of play from his physical play, nor can you separate his teammates's regular effectiveness of play from any increased effectiveness of play due to physicality.

I think that's an overlooked aspect of the game by the talking heads of the statisticians groups, and an overemphasized aspect of the game by the talking heads of the no-stats group. Usually, it's the quieter guys not causing a stir that put fair emphasis on all aspects.
But in the end you don't really need to separate Pronger's physicality from his overall effectiveness. The effect of the physical play is included in the player's overall effectiveness, assuming that the physicality actually leads to better results.

Having said that, I think that the largest obstacle for development of feasile advance statistics for hockey is the effect of other teammates. After all, even the best effort can result in no positive development if the linemate manages to put the puck to the stands from a perfect opportunity while sometimes piss poor pass might result in big positive if the linemate manages to anticipate the situation.

Due to its nature as fast paced team game development and collection of advance stats will probably require enormous amount of watching of the games. And in a one sense it might be that statistics can never show us something that we could not have seen just by watching the games. But statistics might give us a way to better compare the findings of different watchers because in the end there is limited time that we can spend watching the game.

Noldo is offline   Reply With Quote
Old
03-09-2013, 07:36 AM
  #175
Ozamataz Buckshank
Registered User
 
Ozamataz Buckshank's Avatar
 
Join Date: Oct 2010
Location: Massachusetts
Country: United States
Posts: 4,456
vCash: 500
Quote:
Originally Posted by Joe Canadian View Post
No. It's not. And the reasoning you stated doesn't back up your point, though it is correct. Moneyball is about finding market ineffcienies. Teams that have $140 Million budget can have a high payroll, with large contracts, and can still deploy Moneyball techniques to fill out their roster. Unless a team has an unlimited payroll, which the 2004 Red Sox did not, there are always going to be roster spots that GMs need to fill with limited funds.
I just reread the post I quoted. For some reason I thought I was replying to something completely different to what I thought I was. I'll blame that on posting at 7AM before I have my coffee.

Ozamataz Buckshank is online now   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 10:02 AM.

monitoring_string = "e4251c93e2ba248d29da988d93bf5144"
Contact Us - HFBoards - Archive - Privacy Statement - Terms of Use - Advertise - Top - AdChoices

vBulletin Copyright ©2000 - 2015, Jelsoft Enterprises Ltd.
HFBoards.com is a property of CraveOnline Media, LLC, an Evolve Media, LLC company. ©2015 All Rights Reserved.