HFBoards

Go Back   HFBoards > NHL Eastern Conference > Atlantic Division > Toronto Maple Leafs
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
Notices

Advanced Stats

Reply
 
Thread Tools
Old
07-08-2013, 11:01 AM
  #26
hatterson
Global Moderator
 
hatterson's Avatar
 
Join Date: Apr 2010
Location: North Tonawanda, NY
Country: United States
Posts: 12,343
vCash: 50
Send a message via Skype™ to hatterson
Quote:
Originally Posted by mydnyte View Post
they dont work there either ...player B always strikes out on a curveball ...player B has never hit a home run on a curveball ...next at bat player B hits a home run on a curveball.

nothing in sports is can be determined as an absolute, stats help, but, often-time, they also end up costing you the game.
You don't seem to understand how advanced stats work, or at least how they're applied.

No set of stats can ever tell you "X is impossible" or "Y is guaranteed to happen" There will always be a level of luck/determination/heart/intangibles/whatever that go into specific plays.

Do you distrust using the infield shift against certain batters? That's application of an 'advanced' stat. If you see a batter hits a significant percentage of his ground balls to the left side of the infield you're likely to shift your second baseman over that way. Does that mean he can't knock a ground right up the first base line? No absolutely not, but it's saying he's much more likely to hit it between 2nd and 3rd so adjusting your players give you a higher percentage chance to make a play on the ball.

There's nothing inherently different between 'advanced' stats and 'traditional' stats. If I'm playing basketball I would use a fairly traditional stat (3 point shooting percentage) to tell me that I probably shouldn't leave Ray Allen unguarded at the 3 point line, but that I can likely get away with leaving Dwight Howard standing out there unopposed.

Now, stats will never tell us everything. Watching players, taking in games, is still extremely valuable, however the simple fact is with 1,230+OT hours of hockey played each regular season (51.25 days) it is basically impossible to watch everything every player does, even if it's your full time job. So it's valuable to watch games to get a representation of what a player does and then use stats (advanced or otherwise) to help fill in some of the gaps.

Some of the more popular advanced stats (quality of competition, zone starts, etc.) can help us tell what situations a player frequently plays in. It's easy to see that the Sedin's see extremely offensive related minutes or that a player like Nikolai Kulemin plays extremely hard defensive minutes. That can be hard to tell if you only see a player 3-4 times a year.

__________________
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
07-08-2013, 11:03 AM
  #27
weems
Registered User
 
weems's Avatar
 
Join Date: Jul 2008
Country: Canada
Posts: 10,935
vCash: 500
Advanced stats are valueble if taken in the right context.

The problem with advanced stats in hockey is the sport is so fluid and so many little situational plays will never show up on a score sheet. A stat like Corsi is also heavily dependant on your other 2 linemates and things like aggressive vs passive gameplanning never get taken into account. Many teams employ a defensive system to play everything to the perimeter and are more then willing to give up alot of shots as long as there low percentage shots. It would be akin to a basketball team playing zone all the time vs a man to man defense. Player A could make a good play in the defensive zone, pass the puck up to this center only to have him turn it over and the opposing team come down and get a shot on net......BOOM player A although made a strong play gets docked negatively by Corsi.

I'm of the opinion that although stats will help in alot of situations a skilled scouts eyes will always be much more beneficial. Advanded stats and SCOUTING should be used in unison to come to conclusions.

weems is offline   Reply With Quote
Old
07-08-2013, 11:04 AM
  #28
TieClark
Registered User
 
Join Date: Jun 2011
Posts: 4,115
vCash: 500
They really need to be taken with a grain of salt. Far too many people (James Mirtle from the Globe for example) take them as definite judgements of teams or players. That's not what stats have ever been used for, even advanced ones. They're used as a tool to see relationships between what you're watching and what the stats say. A good example is the people who try to use the Leafs lack of puck possession as a reason they overachieved. While it looks that way if you look at the advanced stats, if you actually watched the games you'd clearly see the Leafs play a style that worked for them yet would directly impact their puck possession numbers in a negative way.

TieClark is offline   Reply With Quote
Old
07-08-2013, 11:13 AM
  #29
weems
Registered User
 
weems's Avatar
 
Join Date: Jul 2008
Country: Canada
Posts: 10,935
vCash: 500
Quote:
Originally Posted by TieClark View Post
They really need to be taken with a grain of salt. Far too many people (James Mirtle from the Globe for example) take them as definite judgements of teams or players. That's not what stats have ever been used for, even advanced ones. They're used as a tool to see relationships between what you're watching and what the stats say. A good example is the people who try to use the Leafs lack of puck possession as a reason they overachieved. While it looks that way if you look at the advanced stats, if you actually watched the games you'd clearly see the Leafs play a style that worked for them yet would directly impact their puck possession numbers in a negative way.
Exactly. Style of play is different from one team to the next and thus greatly would impact what a player or teams advanced stats say.

For example many teams play a far more aggressive forechecking style then others. For these teams there far more likely to pin a team deep in there zone then a more passive forchecking team. These aggressive forechecking teams usually dont get outshot as much as the passive ones but not every roster has the personel to be a consistant heavy forechecking team.


Last edited by weems: 07-08-2013 at 11:27 AM.
weems is offline   Reply With Quote
Old
07-08-2013, 11:21 AM
  #30
Leafsfan1028
Registered User
 
Join Date: Mar 2013
Location: Guelph
Country: Canada
Posts: 43
vCash: 500
Outright dismissing advanced statistics is a poor decision but so is putting all your faith in them. The big thing about advanced statistics like fenwick close and corsi is that they are used to predict future success not so much evaluate how a player has played. Teams that shoot the puck more and give up less shots against tend to win games more frequently then those that don't. The majority of the teams that make the playoffs have either positive shot differentials or close to break even shot differentials. This would seem to indicate that shot differentials can be a predictor of success which is why so many people use them now to evaluate how a team is playing.

As it relates to the leafs, they made the playoffs last season with the lowest fenwick since the Montreal Canadiens in 2002 when Jose Theodore won the Hart. This is the great equalizer with shot differentials, if you have very good goaltending you can overcome them. That was the case this year, Reimer's sv% was 7th in the league, where as in contrast the NJD had a very good shot differential but terrible goaltending which was the downfall of that team. So while I think shot differentials are important they are just one piece of the puzzle.

If the Leafs are to make the playoffs next year they are going to have to get good goaltending again (This is probably a good bet with both Reimer and Bernier on the team) and they will have to either continue to shoot at a high % (the leafs led the league in shooting % last season) or they will have to improve their shot differential to compensate for a lower shooting %.

Leafsfan1028 is offline   Reply With Quote
Old
07-08-2013, 11:24 AM
  #31
achtungbaby
Registered User
 
achtungbaby's Avatar
 
Join Date: Oct 2006
Country: Canada
Posts: 4,504
vCash: 500
Quote:
Originally Posted by leafspring View Post
It was before the play-offs when we were being badly outshot predicting all doom/gloom because of being statistical losers lol. We kept winning,kept getting out shot,made the play-offs,took Boston to 7 games.

Stats don't know who the coach is,they don't know the style you play,they don't know who is injured,or hung over (seguin)lol. These stats are overboard.
Stats don't tell you who will win the games. They'll tell you the odds of one team winning over the other. Do you play poker? Do you use the odds of a given hand when factoring in your decisions? Obvously A-A is great, yet it won't win all the time but the more information you have, the more power to you.

I don't get how people can "hate" a mathematical formula. If things didn't change and new ideas didn't come to light we'd all still be living in caves and wiping our ***** with pine cones.

achtungbaby is offline   Reply With Quote
Old
07-08-2013, 11:43 AM
  #32
anderson3133
Registered User
 
anderson3133's Avatar
 
Join Date: Apr 2010
Location: Kitchener
Country: Canada
Posts: 2,431
vCash: 500
Quote:
Originally Posted by mydnyte View Post
they dont work there either ...player B always strikes out on a curveball ...player B has never hit a home run on a curveball ...next at bat player B hits a home run on a curveball.

nothing in sports is can be determined as an absolute, stats help, but, often-time, they also end up costing you the game.
Stats rarely delve into absolutes, like the ones you are using in your example. There will always be instances, no matter how small the occurrences, where people beat the odds.

However, that does not remove their effectiveness within the game.

anderson3133 is offline   Reply With Quote
Old
07-08-2013, 11:56 AM
  #33
DogsFan*
GHG
 
Join Date: Jan 2009
Location: Hamilton, ON
Country: Canada
Posts: 3,036
vCash: 500
http://www.broadstreethockey.com/201...099158/kangzzz
http://www.broadstreethockey.com/201...i-intro-primer

Both are very good primers on advanced stats. Advanced stats tell much more than you think they do. I remember seeing a model someone came up with on twitter, the guy correctly predicted every series using his possession-based model, including the Blackhawks cup win.

DogsFan* is offline   Reply With Quote
Old
07-08-2013, 11:59 AM
  #34
Ari91
Registered User
 
Join Date: Nov 2010
Location: Toronto
Country: Canada
Posts: 9,570
vCash: 500
Stats are useful but I think some people treat them as being more important than the game itself. They are complimentary, they are not definitive.

When a team is producing results (which ultimately is the most important thing) and people try to use stats to diminish that, that's when I don't take people's use of stats seriously.

People criticized the Leafs and Ducks possession numbers, shot totals, etc. and a quarter into the season they said that they couldn't sustain it for the next quarter of the season. And they did. When the half way mark came, they said they couldn't sustain it any longer and would certainly regress. They didn't. When the season ended and the teams made playoffs, then people said 'well if it were a full 82 game schedule they would have certainly fell off the map'. Well you and your numbers were wrong about them being able to sustain it for 48 games, what makes you think you're right about them being unable to sustain it for 82?

As mentioned, there are so many variables to consider. Stats have a place in the industry and I'm sure teams do use it, but unlike some people, I'm sure teams don't use stats as the be-all of a player or team's abilities. I'm sure they leave that to the armchair experts.

Ari91 is offline   Reply With Quote
Old
07-08-2013, 12:02 PM
  #35
sommervr
Registered User
 
sommervr's Avatar
 
Join Date: Feb 2013
Posts: 1,358
vCash: 500
Stats analysis of hockey is here to stay. Leafs are way behind but what else is new.

If the leafs had anyone who understood hockey stat analysis then Bozak would be long gone.

Also there is nothing "advanced" about them. They are simple calculations based around shot attempts for the most part.

sommervr is offline   Reply With Quote
Old
07-08-2013, 12:05 PM
  #36
ryno23
Registered User
 
Join Date: Feb 2010
Posts: 3,141
vCash: 500
Quote:
Originally Posted by sommervr View Post
Stats analysis of hockey is here to stay. Leafs are way behind but what else is new.

If the leafs had anyone who understood hockey stat analysis then Bozak would be long gone.

Also there is nothing "advanced" about them. They are simple calculations based around shot attempts for the most part.
So that means Jason Blake should be brought back cause he was full of shots on net. This is where advance stats betray people.....tell me who you would rather have Jason Blake or Tyler Bozak cause Blake would be a stats guy go to guy with tons of shots on net.

Those who have seen both play cane say for sure that Bozak is better than Blake

ryno23 is offline   Reply With Quote
Old
07-08-2013, 12:09 PM
  #37
mydnyte
Registered User
 
mydnyte's Avatar
 
Join Date: Sep 2004
Country: Italy
Posts: 11,480
vCash: 500
Quote:
Originally Posted by sommervr View Post
Stats analysis of hockey is here to stay. Leafs are way behind but what else is new.

If the leafs had anyone who understood hockey stat analysis then Bozak would be long gone.

Also there is nothing "advanced" about them. They are simple calculations based around shot attempts for the most part.
do the Leafs have a better option for winning a faceoff than bozak?

key word is simple ...tap in to an open net, vs a just inside the post slapshot through 3 players from the blue line are both goals, and, statistically, they are the same, but, are they the same?

stats help, and can be used to advantage but, stats are also a joke if you rely on them, or misinterpret them.

mydnyte is offline   Reply With Quote
Old
07-08-2013, 12:11 PM
  #38
sommervr
Registered User
 
sommervr's Avatar
 
Join Date: Feb 2013
Posts: 1,358
vCash: 500
Quote:
Originally Posted by ryno23 View Post
So that means Jason Blake should be brought back cause he was full of shots on net. This is where advance stats betray people.....tell me who you would rather have Jason Blake or Tyler Bozak cause Blake would be a stats guy go to guy with tons of shots on net.

Those who have seen both play cane say for sure that Bozak is better than Blake
To analyse Bozak all you have to do is look at his points and who he play with:

http://hockeyanalysis.com/2013/02/16...e-tyler-bozak/

sommervr is offline   Reply With Quote
Old
07-08-2013, 12:13 PM
  #39
crump
~ ~ (ړײ) ~ ~
 
crump's Avatar
 
Join Date: Feb 2004
Location: Ontariariario
Country: Canada
Posts: 5,516
vCash: 987
Like many have said, there are too many variables to just use stats to make roster decisions. From an ownership point of view, they want to fill the seats. That could mean anything from icing a team that is fast or skilled, or a pugilistic team a combination of all those. If you went strictly by stats teams would have players named Jake Muzzin, Cam Atkinson filling their roster. Or would you really want the Sedin Twins leading your team into the playoffs?

crump is offline   Reply With Quote
Old
07-08-2013, 12:16 PM
  #40
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,383
vCash: 500
Quote:
Originally Posted by caribouPINE View Post
They can help sure, but with anything, there are limitations and pitfalls. And, like anything, they shouldn't be taken as a whole just part of the puzzle. Another piece.

If you follow James Mirtle though you would think they are essential to evaluating all trades, players, etc. In fact, I think Mirtle should go follow baseball as he's clearly a stathead and has little use for watching the game, rather just look at bar charts and draw Venn diagrams.

And yes, I know James reads this site. Hi James. Shouldn't you be reading Bill James' book on Win Shares to see how you can extrapolate that into the hockey world?
There is this ridiculous thought out there that people who prefer to find objective methods to evaluating players (or "advanced stats") are somehow not enjoying watching the game. In fact, I find that a deeper understanding of a player's value enhances the game experience for me.

I'm glad that advanced stats are finding their way into the MSM for hockey, because for far too long analysts (and many posters on here) rely on old, tired cliches to explain or justify their opinion of a player. If you find yourself using words or phrases like "heart", "will to win", "playoff performer" or "grit", all you are saying is: "I like this player and what he represents or has accomplished, but I cannot explain why".

The Legend is offline   Reply With Quote
Old
07-08-2013, 12:17 PM
  #41
James Mirtle
Registered User
 
James Mirtle's Avatar
 
Join Date: May 2006
Location: Toronto
Country: Canada
Posts: 225
vCash: 500
Quote:
Originally Posted by TieClark View Post
They really need to be taken with a grain of salt. Far too many people (James Mirtle from the Globe for example) take them as definite judgements of teams or players. That's not what stats have ever been used for, even advanced ones. They're used as a tool to see relationships between what you're watching and what the stats say. A good example is the people who try to use the Leafs lack of puck possession as a reason they overachieved. While it looks that way if you look at the advanced stats, if you actually watched the games you'd clearly see the Leafs play a style that worked for them yet would directly impact their puck possession numbers in a negative way.
Well, only if the Leafs are the first team to ever find a way to win consistently with very low puck possession and an 11.5 per cent shooting percentage.

In every other previous case, that type of "style" was unsustainable.

It's not a "definite judgement." It's about probability and regression to the mean. What the Leafs did was highly unusual and out of the norm and that means it's likely, not definite, they will fall back next season.

I guess we'll see.

James Mirtle is offline   Reply With Quote
Old
07-08-2013, 12:18 PM
  #42
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,383
vCash: 500
Quote:
Originally Posted by ryno23 View Post
This is where advance stats betray people.....tell me who you would rather have Jason Blake or Tyler Bozak cause Blake would be a stats guy go to guy with tons of shots on net.
Shots on goal is not an advanced stat.

The Legend is offline   Reply With Quote
Old
07-08-2013, 12:22 PM
  #43
The Legend
Stamkos in 2016
 
The Legend's Avatar
 
Join Date: Feb 2004
Country: Canada
Posts: 4,383
vCash: 500
Quote:
Originally Posted by Lightsol View Post
All these "new" stats, I wonder if hockey is becoming more like baseball, what with the recent insane argument that Miguel Cabrera shouldn't win MVP despite winning the Triple Crown (one of the hardest achievements in baseball) simply because Mike Trout had better Sabremetrics...
Is defense sabremetrics? Is baserunning sabremetrics? What about getting credit in a triple crown category just because a guy in front of you happened to be on third base?

The Legend is offline   Reply With Quote
Old
07-08-2013, 12:24 PM
  #44
wulfio*
 
Join Date: Jan 2011
Location: Burlington
Country: Canada
Posts: 1,520
vCash: 500
Quote:
Originally Posted by achtungbaby View Post
I find that most people who have something against advanced stats are in one of two camps. A: They don't understand them and haven't taken the time to learn about it. B: They don't like the way it makes it their favourite players look.

Like them or not, they're here to stay as more and more teams are beginning to understand how important they can be if used with good old fashioned scouting.
C) actually has played hockey before and knows you can't quantify the sport in to stats.

wulfio* is offline   Reply With Quote
Old
07-08-2013, 12:27 PM
  #45
7even
Moderator
Deus Ex Machina
 
7even's Avatar
 
Join Date: Feb 2012
Location: Georgia
Country: United States
Posts: 8,977
vCash: 500
They're a useful tool as long as you have context. I suppose that goes for everything.

Too many people freak the **** out when you bring them up.

7even is offline   Reply With Quote
Old
07-08-2013, 12:29 PM
  #46
7even
Moderator
Deus Ex Machina
 
7even's Avatar
 
Join Date: Feb 2012
Location: Georgia
Country: United States
Posts: 8,977
vCash: 500
Quote:
Originally Posted by wulfio View Post
C) actually has played hockey before and knows you can't quantify the sport in to stats.
I wouldn't say that. If you can model storm cells and other chaotic systems, you can, to a limited extent, model a hockey game.

7even is offline   Reply With Quote
Old
07-08-2013, 12:38 PM
  #47
ryno23
Registered User
 
Join Date: Feb 2010
Posts: 3,141
vCash: 500
Quote:
Originally Posted by The Legend View Post
There is this ridiculous thought out there that people who prefer to find objective methods to evaluating players (or "advanced stats") are somehow not enjoying watching the game. In fact, I find that a deeper understanding of a player's value enhances the game experience for me.

I'm glad that advanced stats are finding their way into the MSM for hockey, because for far too long analysts (and many posters on here) rely on old, tired cliches to explain or justify their opinion of a player. If you find yourself using words or phrases like "heart", "will to win", "playoff performer" or "grit", all you are saying is: "I like this player and what he represents or has accomplished, but I cannot explain why".
So why do you have to explain why a player is good? I think Mark Messier has great grit, playoff performer, heart and the best leader in hockey. Problem is for the stats guys you can't put a number on it to analyze it to fit in a nice package. Intangibles is a word that I know in baseball the stats geeks hate and laugh at but really that is what a player brings to the team inside the lockeroom and on the ice.

Sometimes you don't need to explain with stats why a player is good or bad you can watch the game as the visual image of his playing will tell you anything before you go and look at a spreadsheet. I don't need stats to tell me Brett Lebda was horrible D

For instance how did you put numbers to Scott Stevens when he played. He put the fear in god to forwards who would cross the blueline with their head down. Most players knew he was on the ice and would dump it in instead of trying to carry it over the blueline in fear of a devastating bodycheck by Stevens. I don't need fancy stats to tell me Stevens is a difference maker out there regardless of his puck possesion numbers or fenwick or corsi. Ask Paul Karyia his number took a beating after the cup final.

And that is where the divide is from the stats people who cannot accept why a player is good by just saying it without backing stats....they cannot rely on the actual watching a player and forming an view of the player.

Now I know the last statement was pretty general but it is mainly true for a good %

ryno23 is offline   Reply With Quote
Old
07-08-2013, 12:38 PM
  #48
sangreale
Registered User
 
Join Date: Feb 2008
Posts: 2,064
vCash: 500
Quote:
Originally Posted by 7even View Post
They're a useful tool as long as you have context. I suppose that goes for everything.

Too many people freak the **** out when you bring them up.
Where I stand. And I think that in all sports enlightened team management look at them the same way.

To ignore them is to put yourself at a disadvantage. To employ them at the exclusion of all else would suggest a mathematician could run a team and that is obviously not the case.

sangreale is offline   Reply With Quote
Old
07-08-2013, 12:42 PM
  #49
achtungbaby
Registered User
 
achtungbaby's Avatar
 
Join Date: Oct 2006
Country: Canada
Posts: 4,504
vCash: 500
Quote:
Originally Posted by wulfio View Post
C) actually has played hockey before and knows you can't quantify the sport in to stats.
Playing a game of hockey makes you a math expert? How does that work?? If we can model a group of pigeons movements around a park we can certainly model a game with objectives to some degree of accuracy.

achtungbaby is offline   Reply With Quote
Old
07-08-2013, 12:51 PM
  #50
Ari91
Registered User
 
Join Date: Nov 2010
Location: Toronto
Country: Canada
Posts: 9,570
vCash: 500
Quote:
Originally Posted by sangreale View Post
Where I stand. And I think that in all sports enlightened team management look at them the same way.

To ignore them is to put yourself at a disadvantage. To employ them at the exclusion of all else would suggest a mathematician could run a team and that is obviously not the case.
I think that most of the responses in this thread would suggest that most people aren't afraid or dislike them but think that they are useful when used in the appropriate context. Stats help to figure out the odds, they don't guarantee the results. Stats only provide a figurative framework to logical predictions.

Variables aside:

-Generally speaking, a team that takes more shots and allows less shots against are more likely to be successful than a team that doesn't
-Generally speaking, a team that scores more goals than they allow are more likely to be successful than a team that doesn't
-Generally speaking, a player/team that spends more time with the puck rather than chasing it more likely to be successful than a player/team that doesn't.

Ari91 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 12:05 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.