Moneyball...The Hockey Version...Where are you?
- Thread starter Firewagon
- Start date
Hockey Outsider
Registered User
- Jan 16, 2005
- 9,144
- 14,450
Here are some good ones:
Alan Ryder's Hockey Analytics: http://hockeyanalytics.com/Research.htm
Iain Fyffe's Puckerings: http://www.puckerings.com/research.html
Daryl Shilling's The Hockey Project: http://members.shaw.ca/hbtn/index.htm
Also, get "The Hockey Compendium" by Jeff Z Klein and Karl-Eric Reif. It's $5 in most bookstores in Canada.
Alan Ryder's Hockey Analytics: http://hockeyanalytics.com/Research.htm
Iain Fyffe's Puckerings: http://www.puckerings.com/research.html
Daryl Shilling's The Hockey Project: http://members.shaw.ca/hbtn/index.htm
Also, get "The Hockey Compendium" by Jeff Z Klein and Karl-Eric Reif. It's $5 in most bookstores in Canada.
Bear of Bad News
Your Third or Fourth Favorite HFBoards Admin
- Sep 27, 2005
- 13,511
- 26,966
On Yahoo! groups, check out the Hockey Analytics Group. Alan and Iain are prominent there, as well as several other bright statistical folks.
And I'll second the nomination of Klein/Reif!
And I'll second the nomination of Klein/Reif!
ered7
Registered User
- Oct 21, 2006
- 98
- 2
Baseball/Moneyball has the slight advantage of a tremendous collection of searchable statisical data that is readily available. Hockey unfortunately is still in its early stages of compiling and making sortable the statistical data. Baseball benefitted from being America's past time, which meant a huge media following from the newspapers and eventually radio and television. (Statisical data for baseball at least in the 20th century is definitely readily available, even on the internet for free. 19th century baseball stats have some gaps but a lot has been filled in during the last 3 decades). Hockey unfortunately hasn't had the same coverage in the U.S. and that's probably hurt the availability of stats. In many markets finding old scoresheets and data requires digging through newspapers and old Hockey News issues (which in many US libraries are not as prevalent as one would think). In Canada, I'm sure the availability is probably greater for finding specific stats on a whim. Sites like hockeydb.com and SHRP Sports are great but they lack the tie in of comprehensive searchable game info that a site like retrosheet provides for baseball. The NHL has come into the light with RTSS but it's less than 10 yrs old. So what does the researcher of older and perhaps flawed data do? They must rely on people that have taken the time to compile the old score sheets and verify how accurate the data is (which may or may not be feasible in some cases). The Hockey Summary Project seems to be addressing that issue, but that takes time to accomplish. Until there's a hockey equivalent of retrosheet, hockey analysis is at a slight disadvantage. It's wonderful when members of the media that cover hockey throw out obscure stats (thanks to the elias sports bureau and stats inc.), but largely the common fan doesn't have access to these paid service sources. Bill James and his followers gave Moneyball its genesis by having done years of analysis that impacted baseball management and created a ripe environment for a Billy Beane to attempt to run a franchise the way he has. Hockey still has a lot of old guard owners and management that hasn't necessarily adopted the same style of management: decisions based largely on statistical output as opposed to player drive, fortitude, tenacity, etc...The next 5-10 years should be interesting as RTSS will probably play a larger role. GMs will be able to look more closely at how players have perfomed game in and game out, rather than merely reading who scored or who comitted a penalty during a specified game. It should be interesting.
Last edited:
Lets also be very clear here. Baseball and hockey are not comparable in this way. Baseball is an individual sport played in teams while hockey is a team game.
Yep. Moneyball could have been summarized on one page. The only thing Billy Beane seemed to care about is if a player walks and gets on base. He only cares about 2 particular stats: OBP (on base percentage ) and SLG (slugging percentage). He doesn't care what you're batting average is, how many home runs you hit or how big, strong, or the potential that you may have. He just wants you to get on base. That's all. Because statistically it leads to runs, which win games.
Like you said hockey is a team game and is less stats crazy than baseball. Beane wouldn't care about goals, assists etc. I can't really think of a comparable stat in hockey that would be similar to an OBP or SLG. How about a players ability to draw a penalty and put the other team at a disadvantage? But that will turn the NHL into a total diving league. I don't think we want that.
mullin, you managed to summarize, to a certain extent, in two sentences what I could rant on and on about for 2,000 words. Hockey is a game that transcends statistics. While the league has started to keep track of some intangible statistics - faceoffs, shots blocked, hits, takeaways, etc. - (and much to my chagrin, might I add), there are so many you can't keep track of. You can't keep track of how good a player is at drawing penalties, winning those all-too important battles in the corners, or just all-out second effort. As for goalies, they still haven't found away to keep track of what matters most - the ability to make the saves when the game matters most.
I wouldn't call baseball an individual sport. I love the game and I love the strategy. I think coaching is more important in baseball than any other sport in North America. And I love that every single pitch counts.
But part of what makes hockey great is that only those who don't know the game would try to capture a player's value on stats alone.
OHLArenaGuide
it's dot com
You couldn't possibly be more wrong.
Wait, yes you could, you could be Joe Morgan. But you're still completely and fundamentally misunderstanding what Moneyball is about.
Moneyball is about the inefficiencies in the baseball player market and how a savvy GM could exploit those inefficiencies to win games cheaply.
Oakland was 10th in team OBP in 2006, and - get this - 27th in SLG. Yet they made the ALCS. Why? Because Billy's one step ahead of the game, and he's using new stats and analytical tools that weren't developed at the time of Moneyball's writing.
I must also throw my hat solidly behind the work of Klein and Reif; I've met both personally and can attest to the fact that in addition to being able statisticians and the closest thing hockey has to Bill James, they are both very nice men.
As OHLArenaGuide pointed out, Oakland wasn't necessarily strong in these categories. Unfortunately Moneyball became synonymous with OBP/SLG. Moneyball is more or less taking advantage of what attributes are being undervalued in the current marketplace.
At the time that the book was written, OBP was underrated versus batting average. After a couple years, a lot of other teams caught onto valuing OBP over BA and OBP was no longer cheap to come by. Next, they determined that defense was an undervalued commodity which prompted things like them acquiring Mark Kotsay.
Great thread, I loved that book and I'm sure this sort of system can be applied to hockey. I don't know how but if you think of it like Baseball it's unlikely to work. You have to develop your stats and so on.
There are stats alreay out there that give a good idea on what a guy does like shots and hits and so on. The philosophy of Moneyball is to judge a player on what he does rather the result itself. To me shots and hits are somewhat similar to this. I've never tried to build this sort of system but I'm sure it can be done.
There are stats alreay out there that give a good idea on what a guy does like shots and hits and so on. The philosophy of Moneyball is to judge a player on what he does rather the result itself. To me shots and hits are somewhat similar to this. I've never tried to build this sort of system but I'm sure it can be done.
The point of moneyball is that the A's care most about whichever player skills are undervalued in the market. At the time the book was written the most undervalued talent was patience at the plate, drawing walks, and having a high OBP. Thus, that's what Beane exploited.
Since then their strategy has changed a little, and it will keep changing based on which skills are under/over valued in the market at any one time.
Recently, the A's have placed more emphasis on a players defensive ability because they've found that to be undervalued.
Anyway, it's tough to draw comparisons over 2 completely different sports. Baseball is game of numbers while sports like hockey and soccer are completely different. Don't think there will ever be a "moneyball" team in hockey.
Since then their strategy has changed a little, and it will keep changing based on which skills are under/over valued in the market at any one time.
Recently, the A's have placed more emphasis on a players defensive ability because they've found that to be undervalued.
Anyway, it's tough to draw comparisons over 2 completely different sports. Baseball is game of numbers while sports like hockey and soccer are completely different. Don't think there will ever be a "moneyball" team in hockey.
RSBPC
Registered User
- Jan 19, 2005
- 2,356
- 0
Hockey will never be able to be statisitcally analyzed to the extent that baseball is because the two games are so fundamentally different.
In baseball, every single event can be statistically documented, measured and defined. Every event can be attributed to a specific player or reason. Everything is black and white.
In hockey, you can't do that. Every person on the ice is involved in every play and it is difficult if not impossible to define when one play begins and another one ends. There are a million gray areas that can not be defined.
I'm not saying that hockey stats won't become a more useful tool over time as people come up with better and more efficient ways to measure what goes on, but it will never catch up to baseball.
In baseball, every single event can be statistically documented, measured and defined. Every event can be attributed to a specific player or reason. Everything is black and white.
In hockey, you can't do that. Every person on the ice is involved in every play and it is difficult if not impossible to define when one play begins and another one ends. There are a million gray areas that can not be defined.
I'm not saying that hockey stats won't become a more useful tool over time as people come up with better and more efficient ways to measure what goes on, but it will never catch up to baseball.
