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.

Is there an equivalent of a "Moneyball" for the NHL?

Reply
 
Thread Tools
Old
09-10-2011, 07:54 PM
  #101
dire wolf
be cool
 
dire wolf's Avatar
 
Join Date: May 2006
Location: Cali
Country: United States
Posts: 4,539
vCash: 500
Moneypuck?

With the new Brad Pitt movie Moneyball coming out, I thought this would be a timely subject. If any of you have read the book Moneyball by Michael Lewis, you'll know what this is about. If you haven't, I highly recommend it. Even if you aren't a baseball fan, the philosophy espoused by A's GM Billy Beane has some applicability to hockey.

To summarize the premise, Billy Beane is/was the GM of the Oakland A's, which were being run on a shoestring budget. And the question was how could they compete against teams like the Yankees, who spent 4x more on their payroll. The answer turned out to be that they didn't buy into the old way of thinking, and looked for players who were undervalued in the marketplace because they didn't excel in traditional stats or didn't look like great athletes. For example, Beane and his team figured out that batting average was the most overrated statistic and that teams were over-paying players who had high batting averages. They figured out that the way to win games is to score runs, and the best way to score runs is to not get outs. The statistic that correlates best to not getting out is OBA (on base average). Therefore, they found guys that had great OBAs who were being overlooked in the draft and undervalued in the MLB.

So, the question is, are there similar theories and statistics that would apply to hockey. This is especially relevant in the salary cap era (although, obviously, hockey is less of a statistical game than baseball).

Thoughts?

dire wolf is offline   Reply With Quote
Old
09-10-2011, 07:58 PM
  #102
Xamboni
BoVat Horseman
 
Xamboni's Avatar
 
Join Date: May 2009
Location: Cascadia
Country: Canada
Posts: 7,417
vCash: 500
Ignore the +/- stat

Xamboni is offline   Reply With Quote
Old
09-10-2011, 08:00 PM
  #103
Jeffrey
Registered User
 
Jeffrey's Avatar
 
Join Date: Feb 2003
Location: Montreal
Country: Canada
Posts: 9,491
vCash: 50
Send a message via MSN to Jeffrey
The most overrated stats are the +/- and hits.

The most underrated are face-off win, and even-strenght goals

Jeffrey is offline   Reply With Quote
Old
09-10-2011, 08:02 PM
  #104
dire wolf
be cool
 
dire wolf's Avatar
 
Join Date: May 2006
Location: Cali
Country: United States
Posts: 4,539
vCash: 500
What about some sort of relative +/- stat that compare a player only to his teammates?

dire wolf is offline   Reply With Quote
Old
09-10-2011, 08:15 PM
  #105
eldiablo17
Registered User
 
eldiablo17's Avatar
 
Join Date: Aug 2007
Posts: 2,925
vCash: 500
Mike Gillis talked about using a similar approach to moneyball when he took office, what that exactly entailed no-one knows but he did refer to looking at a different set of criteria that may be overlooked.

He then went after a player like Samuelsson who had a career year in both points and assists, Ehrhoff who did the same, Malhotra's role far exceeds his price tag according to most Canuck fans, and most of his major acquisitions have been great ones outside of Ballard so far, so perhaps he's on to something perhaps not.

eldiablo17 is offline   Reply With Quote
Old
09-10-2011, 08:38 PM
  #106
Rhaego
Registered User
 
Rhaego's Avatar
 
Join Date: May 2011
Country: Norfolk Island
Posts: 6,646
vCash: 500
Watch out there going to make a Greg Sherman movie in a few years after the Avs make a new dynasty with all their re-aclamation projects

Rhaego is offline   Reply With Quote
Old
09-10-2011, 08:55 PM
  #107
sabresEH
Registered User
 
sabresEH's Avatar
 
Join Date: May 2009
Location: Kelowna, BC
Country: Canada
Posts: 1,242
vCash: 500
Look to Detroit. I swear they don't draft BPA they draft SPA(smartest player available).

sabresEH is offline   Reply With Quote
Old
09-10-2011, 09:04 PM
  #108
backdoorpass
 
Join Date: Mar 2007
Country: Canada
Posts: 979
vCash: 500
a team like nashville could almost be like how oakland was, also maybe takeaway to giveaway %

backdoorpass is offline   Reply With Quote
Old
09-10-2011, 09:17 PM
  #109
ObeyMashinter
Registered User
 
ObeyMashinter's Avatar
 
Join Date: Sep 2011
Location: Chicago, IL
Posts: 1,398
vCash: 500
Quote:
Originally Posted by dire wolf View Post
With the new Brad Pitt movie Moneyball coming out, I thought this would be a timely subject. If any of you have read the book Moneyball by Michael Lewis, you'll know what this is about. If you haven't, I highly recommend it. Even if you aren't a baseball fan, the philosophy espoused by A's GM Billy Beane has some applicability to hockey.

To summarize the premise, Billy Beane is/was the GM of the Oakland A's, which were being run on a shoestring budget. And the question was how could they compete against teams like the Yankees, who spent 4x more on their payroll. The answer turned out to be that they didn't buy into the old way of thinking, and looked for players who were undervalued in the marketplace because they didn't excel in traditional stats or didn't look like great athletes. For example, Beane and his team figured out that batting average was the most overrated statistic and that teams were over-paying players who had high batting averages. They figured out that the way to win games is to score runs, and the best way to score runs is to not get outs. The statistic that correlates best to not getting out is OBA (on base average). Therefore, they found guys that had great OBAs who were being overlooked in the draft and undervalued in the MLB.

So, the question is, are there similar theories and statistics that would apply to hockey. This is especially relevant in the salary cap era (although, obviously, hockey is less of a statistical game than baseball).

Thoughts?
Meh. Beane's teams never made it to the big dance and only came close when they had 3-4 shut down pitchers. Most teams with that kind of pitching talent could have easily finished where the A's did during those years, especially in 5 and 7 game series.

It didn't work anyway considering they lost Mulder,Zito,Hudson,Harden all to trades/free agency. It proved that in the end they really couldn't compete with the big market payrolls. Obviously Sabremetrics are nice tools to have, but to depend on them like Beane did could be a stretch as it never got them to the big dance.

Same thing can probably be said for hockey. Qualcomp,Qualteam,Corsi etc. are nice tools but if cap floor teams depend on them solely, they are probably going to be disappointed. My $.02. Thanks for reading.

ObeyMashinter is offline   Reply With Quote
Old
09-10-2011, 09:24 PM
  #110
hatterson
Global Moderator
 
hatterson's Avatar
 
Join Date: Apr 2010
Location: North Tonawanda, NY
Country: United States
Posts: 12,335
vCash: 50
Send a message via Skype™ to hatterson
Quote:
Originally Posted by ObeyJohnScott View Post
Meh. Beane's teams never made it to the big dance and only came close when they had 3-4 shut down pitchers. Most teams with that kind of pitching talent could have easily finished where the A's did during those years, especially in 5 and 7 game series.

It didn't work anyway considering they lost Mulder,Zito,Hudson,Harden all to trades/free agency. It proved that in the end they really couldn't compete with the big market payrolls. Obviously Sabremetrics are nice tools to have, but to depend on them like Beane did could be a stretch as it never got them to the big dance.

Same thing can probably be said for hockey. Qualcomp,Qualteam,Corsi etc. are nice tools but if cap floor teams depend on them solely, they are probably going to be disappointed. My $.02. Thanks for reading.
The point wasn't to win the world series per se. It was to do the best they could with what they had to work with.

Teams that use(d) Sabremetrics had significantly lower cost/win than teams that didn't use them.

It may not mean a world series, but it might mean the difference between 80 wins and 60 wins and 80 wins puts more butts in the seats than 60 wins does.

hatterson is offline   Reply With Quote
Old
09-10-2011, 09:28 PM
  #111
mpleafan
Registered User
 
mpleafan's Avatar
 
Join Date: Jul 2010
Posts: 909
vCash: 500
isnt that what The Love Guru is about???

mpleafan is offline   Reply With Quote
Old
09-10-2011, 09:44 PM
  #112
Rhaego
Registered User
 
Rhaego's Avatar
 
Join Date: May 2011
Country: Norfolk Island
Posts: 6,646
vCash: 500
Quote:
Originally Posted by mpleafan View Post
isnt that what The Love Guru is about???
What do you mean? Rob Blake doesn't take face-offs in real life? La-Quaq Grande isn't a real goalie? That's def a based off real life

Rhaego is offline   Reply With Quote
Old
09-10-2011, 10:20 PM
  #113
Nullus Reverentia
Registered User
 
Nullus Reverentia's Avatar
 
Join Date: May 2010
Location: The Periphery
Country: Lord Howe Island
Posts: 17,222
vCash: 500
Quote:
Originally Posted by ObeyJohnScott View Post
Meh. Beane's teams never made it to the big dance and only came close when they had 3-4 shut down pitchers. Most teams with that kind of pitching talent could have easily finished where the A's did during those years, especially in 5 and 7 game series.

It didn't work anyway considering they lost Mulder,Zito,Hudson,Harden all to trades/free agency. It proved that in the end they really couldn't compete with the big market payrolls. Obviously Sabremetrics are nice tools to have, but to depend on them like Beane did could be a stretch as it never got them to the big dance.

Same thing can probably be said for hockey. Qualcomp,Qualteam,Corsi etc. are nice tools but if cap floor teams depend on them solely, they are probably going to be disappointed. My $.02. Thanks for reading.
He never had success with it but I've seen it been said that his methods have been used in many other teams since and many of those teams have been succesful. It's more of the ideologie then his own individual success.

On the threads note, I have no idea. The NHL is diffrent since no team can spend more then twice what another does, so money can't buy a team.

Nullus Reverentia is offline   Reply With Quote
Old
09-10-2011, 10:21 PM
  #114
Mayor Bee
\/me_____you\/
 
Mayor Bee's Avatar
 
Join Date: Dec 2008
Posts: 15,730
vCash: 500
Quote:
Originally Posted by dire wolf View Post
With the new Brad Pitt movie Moneyball coming out, I thought this would be a timely subject. If any of you have read the book Moneyball by Michael Lewis, you'll know what this is about. If you haven't, I highly recommend it. Even if you aren't a baseball fan, the philosophy espoused by A's GM Billy Beane has some applicability to hockey.

To summarize the premise, Billy Beane is/was the GM of the Oakland A's, which were being run on a shoestring budget. And the question was how could they compete against teams like the Yankees, who spent 4x more on their payroll. The answer turned out to be that they didn't buy into the old way of thinking, and looked for players who were undervalued in the marketplace because they didn't excel in traditional stats or didn't look like great athletes. For example, Beane and his team figured out that batting average was the most overrated statistic and that teams were over-paying players who had high batting averages. They figured out that the way to win games is to score runs, and the best way to score runs is to not get outs. The statistic that correlates best to not getting out is OBA (on base average). Therefore, they found guys that had great OBAs who were being overlooked in the draft and undervalued in the MLB.

So, the question is, are there similar theories and statistics that would apply to hockey. This is especially relevant in the salary cap era (although, obviously, hockey is less of a statistical game than baseball).

Thoughts?
Beane had two problems. The first is that he lacked a manager who had a cohesive idea of how to put the admittedly lesser pieces together. Earl Weaver did a stellar job combining the efforts of marginal players into excellent results, but Beane overlooked entirely the role of a manager.

The second is that, by speaking out, he forever destroyed whatever competitive edge he may have had. When larger market teams started using similar practices, it drove the price of high OBP players through the roof, which put Oakland out of the running for their services right off the bat.

A third thing is that Michael Lewis is a tool and probably did more damage to the idea of meshing Sabremetric study with traditional scouting than anyone in the last 50 years.

Quote:
Originally Posted by eldiablo17 View Post
Mike Gillis talked about using a similar approach to moneyball when he took office, what that exactly entailed no-one knows but he did refer to looking at a different set of criteria that may be overlooked.

He then went after a player like Samuelsson who had a career year in both points and assists, Ehrhoff who did the same, Malhotra's role far exceeds his price tag according to most Canuck fans, and most of his major acquisitions have been great ones outside of Ballard so far, so perhaps he's on to something perhaps not.
Ehrhoff was hotly pursued by several teams, Malhotra is an exceptional defensive forward who has always been recognized as such, and Samuelsson was always able to get opportunities but not able to convert on them (as reflected in his low shooting percentages with Detroit).

Mayor Bee is offline   Reply With Quote
Old
09-10-2011, 10:29 PM
  #115
Ogopogo*
 
Ogopogo*'s Avatar
 
Join Date: Apr 2005
Location: Edmonton
Country: Canada
Posts: 14,214
vCash: 500
Quote:
Originally Posted by Xavier Ouellet View Post
Ignore the +/- stat


Find a GM that leans heavily on +/- when evaluating talent and he will get ripped off in trades regularly.

Ogopogo* is offline   Reply With Quote
Old
09-10-2011, 10:32 PM
  #116
Mr Wentworth
Arch Duke of Raleigh
 
Mr Wentworth's Avatar
 
Join Date: Aug 2008
Country: United States
Posts: 5,936
vCash: 500
Quote:
Originally Posted by Rockies94 View Post
Watch out there going to make a Greg Sherman movie in a few years after the Avs make a new dynasty with all their re-aclamation projects
Or Garth Snow and the Isles.

Mr Wentworth is offline   Reply With Quote
Old
09-10-2011, 10:37 PM
  #117
Hero
Buffy is Back
 
Hero's Avatar
 
Join Date: Jul 2009
Posts: 20,144
vCash: 500
Hits is SO objective

Hero is offline   Reply With Quote
Old
09-10-2011, 10:37 PM
  #118
eldiablo17
Registered User
 
eldiablo17's Avatar
 
Join Date: Aug 2007
Posts: 2,925
vCash: 500
Quote:
Originally Posted by Mayor Bee View Post
Ehrhoff was hotly pursued by several teams, Malhotra is an exceptional defensive forward who has always been recognized as such, and Samuelsson was always able to get opportunities but not able to convert on them (as reflected in his low shooting percentages with Detroit).
The point was, with those three players in particular, Gillis looked more at what they could do than what they did. Samuelsson was given a greater opportunity (and a bigger paycheck), as a legit top-6 forward and responded with 30 goals and 50+ points. Ehrhoff was a hot commodity (although given what he was acquired for...) yet he was behind a guy like Boyle, he was given that opportunity as the #1 puck mover and thrived. People considered Malhotra as slightly overpaid at 2.5 at the time, the highest Doug Wilson was going to go was 2.2 I believe although they really liked him there, nevertheless I don't think there is a Canuck fan who would say Malhotra is overpaid now (despite what they thought at the time of signing). Now I'm not saying Gillis is some sort of genius, he did win GM of the year though, but he obviously saw something more in these players then when he acquired them and it paid off, same thing can be said about Hamhuis I believe. He was behind Suter and Weber and although an extremely sought after player he performed above what most expected here.

Now how much is from the "Moneypuck" concept and how much is just filling a need is up for debate but there is no debating that in the past few years players have been better than expected coming to Vancouver even when coming from top-ranked teams like SJ and Detroit.

eldiablo17 is offline   Reply With Quote
Old
09-10-2011, 10:41 PM
  #119
Sensfanman
Registered User
 
Sensfanman's Avatar
 
Join Date: Jan 2006
Location: Los Angeles, CA
Country: Canada
Posts: 10,182
vCash: 500
I know Buffalo has an Analytics department; a guy there told me 6 teams do. I think Blackhawks, Kings and Lightning do but I can't be sure.

Personally, I have one academic paper published which uses a K-means clustering algorithm to classify players into groups based on basic nhl.com stats then uses regression to see how many points in the standings each type is worth in conjunction with TOI. I have potentially 2 more papers that explore extensions of this idea.

Sensfanman is offline   Reply With Quote
Old
09-10-2011, 10:43 PM
  #120
hatterson
Global Moderator
 
hatterson's Avatar
 
Join Date: Apr 2010
Location: North Tonawanda, NY
Country: United States
Posts: 12,335
vCash: 50
Send a message via Skype™ to hatterson
Quote:
Originally Posted by Sensfanman View Post
I know Buffalo has an Analytics department; a guy there told me 6 teams do. I think Blackhawks, Kings and Lightning do but I can't be sure.

Personally, I have one academic paper published which uses a K-means clustering algorithm to classify players into groups based on basic nhl.com stats then uses regression to see how many points in the standings each type is worth in conjunction with TOI. I have potentially 2 more papers that explore extensions of this idea.
I'd be very interested in this/these papers. Is there an online (free) copy available or are they journal only?

hatterson is offline   Reply With Quote
Old
09-10-2011, 10:49 PM
  #121
Sensfanman
Registered User
 
Sensfanman's Avatar
 
Join Date: Jan 2006
Location: Los Angeles, CA
Country: Canada
Posts: 10,182
vCash: 500
Quote:
Originally Posted by hatterson View Post
I'd be very interested in this/these papers. Is there an online (free) copy available or are they journal only?
Well, technically, it's been accepted and will be published soon

But for papers on the hockey analytics in academic journals, the one I submit to is Interfaces journal, they are doing a call for Operation Research in sports.

Other ones I can tell you about are:
http://www.bepress.com/jqas/
http://www.hockeyanalytics.com/Resea...ot_Quality.pdf (there are more at this site)


Here's a video from a conference on this:
http://www.sloansportsconference.com/?p=704 (you can see me in this video lol)

Sensfanman is offline   Reply With Quote
Old
09-10-2011, 10:53 PM
  #122
Mayor Bee
\/me_____you\/
 
Mayor Bee's Avatar
 
Join Date: Dec 2008
Posts: 15,730
vCash: 500
Quote:
Originally Posted by eldiablo17 View Post
The point was, with those three players in particular, Gillis looked more at what they could do than what they did. Samuelsson was given a greater opportunity (and a bigger paycheck), as a legit top-6 forward and responded with 30 goals and 50+ points.
Samuelsson had already scored 23 goals with Detroit and looked to be a 20-25 goal guy (he scored 14 in a 53-game season and had another year of 19), so it's not like he was an unknown. If anything, give Detroit credit for taking someone who was looking like a skilled journeyman and benefiting from giving him minutes, linemates, and stability.

Quote:
Ehrhoff was a hot commodity (although given what he was acquired for...) yet he was behind a guy like Boyle, he was given that opportunity as the #1 puck mover and thrived.
Did being behind Boyle mean that he wasn't a #1-caliber defenseman? That'd be like giving credit to a team for picking up Ryan Suter from Nashville because he's behind Shea Weber...everyone can plainly see what he can do.

Quote:
People considered Malhotra as slightly overpaid at 2.5 at the time, the highest Doug Wilson was going to go was 2.2 I believe although they really liked him there, nevertheless I don't think there is a Canuck fan who would say Malhotra is overpaid now (despite what they thought at the time of signing).
Malhotra made $1.5 million in Columbus, left after not coming to terms on a contract extension, and the Jackets replaced him with Pahlsson at $2.65 million. It stands to reason that the Jackets were probably offering Malhotra at least $2.5 million per year, which would then put his moving on under the ol' "JR got some bad advice".

Malhotra being an exceptional defensive forward is nothing new; it was plain as day during his five years in Columbus. That no one else seems to have been aware of it isn't out fault, and it's not to Gillis' or Doug Wilson's credit. I'll also point out that his career high in points came in Columbus as well.

Mayor Bee is offline   Reply With Quote
Old
09-10-2011, 10:54 PM
  #123
Tavaresmagicalplay*
 
Join Date: May 2009
Posts: 19,306
vCash: 500
Advanced statistics don't really work for hockey. They're probably better for goalies but skaters not so much. I'm not telling anybody what they don't know when I say baseball values statistics more because it's the same situations over and over and over and over and over *takes deep breath* and over and over and over and over. It's one versus one. The only real stat quality of teammates has an effect on is RBI and I guess something like walks since they can pitch around you more on a bad team. In hockey the players around you for the most part are everything.

I'm not a big baseball guy but just saying. By the way Corsi is the most garbage statistic ever invented.

Tavaresmagicalplay* is offline   Reply With Quote
Old
09-10-2011, 11:18 PM
  #124
hototogisu
Global Moderator
Poked the bear!!!!!
 
hototogisu's Avatar
 
Join Date: Jun 2006
Location: Montreal, QC
Country: Canada
Posts: 35,171
vCash: 500
Awards:
Quote:
Originally Posted by hatterson View Post
The point wasn't to win the world series per se. It was to do the best they could with what they had to work with.

Teams that use(d) Sabremetrics had significantly lower cost/win than teams that didn't use them.

It may not mean a world series, but it might mean the difference between 80 wins and 60 wins and 80 wins puts more butts in the seats than 60 wins does.
Indeed. People who say "Sabremetrics is a load of crap, Billy Beane's teams never even made it to the World Series" miss the point so spectacularly it's actually kind of amazing. Somewhere along the way people confused Sabremetrics and Beane's techniques with "a surefire way to dominate baseball using stats". It was not nor was it ever supposed to be anything at all like that.

And if you ask my opinion, no, advanced statistics could never work in such a way for hockey. Are they useful when used in the proper context? Sure. But you can never define players and situations based on stats in hockey like you could in baseball.

hototogisu is offline   Reply With Quote
Old
09-10-2011, 11:45 PM
  #125
Do Make Say Think
& Yet & Yet
 
Do Make Say Think's Avatar
 
Join Date: Jun 2007
Posts: 25,843
vCash: 50
Baseball is a static sport, that's why statistics are so useful.

In a game like hockey they usually don't mean that much: they're good indicators (even +/-) but that's about it, they're not the cornerstone of player evaluation and never will be.

Do Make Say Think 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 02:37 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.