HFBoards

Go Back   HFBoards > NHL Eastern Conference > Metropolitan Division > Carolina Hurricanes
Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
Notices

A Sabremetric Case for Zach Boychuk

Closed Thread
 
Thread Tools
Old
07-11-2012, 12:26 PM
  #1
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
A Sabremetric Case for Zach Boychuk

A lot of this board seems to have given up on one Mr. Zachary Boychuk.

Today I am here to tell you why you shouldn't be so quick to give up on him, and hopefully restore some of your faith in the kid.

A lot of people do not like advanced statistics, and that is fine. You may disregard a lot of these, and that's your opinion, but personally I LOVE advanced stats. I think they tell you a lot more than what the basic statistics can.

So here goes:

-This season, Boychuk did not take a single penalty, which leaves him at 0.0 penalties taken per 60 minutes. On the other hand, Boychuk drew 1.9 penalties per 60 minutes, second only to Skinner (2.5).

-Per 60 minutes, Boychuk had more primary assists than Tlusty, Brent, Nodl, LaRose, and Dwyer. Also per 60 minutes, he had more secondary assists than Staal, Jokinen, Skinner, Ruutu, Tlusty, LaRose, Dwyer, Sutter, or anyone who played more than 15 games.

-He was also rather unlucky. His on ice save percentage was the worst amongst any forward, and his on ice shooting percentage was 13th. This leaves him having easily the worst PDO on the team. (PDO is those two stats added together, which is basically a statistic that quantifies luck.) Oddly enough, LaRose had the second lowest PDO, and Staal had the fourth lowest.

-Boychuk lead the team in Corsi Relative. Which is very impressive in my opinion. What that means is when Boychuk was on the ice, the Canes had more shots for per 60 than against per 60 than when any other player was on the ice.

Essentially what these stats tell us is that Boychuk had ridiculously unsustainable bad luck with goalies stopping pucks and he and his teammates putting pucks in the net while he was on the ice. Independent of that, Boychuk had more of a positive impact in terms of shooting attempts (and the next logical step would be puck possession) than any other forward on the team.

Now going back to 2010-2011:

-Boychuk was 4th on the team in ES goals scored per 60 minutes, ahead of Jokinen and Ruutu. While being 13th on the team in primary assists, he was 5th on the team in secondary assists.

-Those left him at 7th in ES points per 60 minutes, behind Stillman, Staal, Cole, Jokinen, Ruutu, and Skinner.

-He was third on the team in penalties drawn per 60 behind Skinner and Samson, while taking .5 penalties per 60 which was slightly below average.

-His Corsi relative was nowhere near as good as 11-12, but he had the highest jump on the team in terms of offensive zone starts to offensive zone finishes, which better quantifies actual puck possession.

I fully realize I'm probably going to get called a stats nerd or whatever and get flamed for this, but I thought it was interesting. Also I'm bored.

Flame away.

Stephen Goalbert is offline  
Old
07-11-2012, 12:36 PM
  #2
w e l o s t b o y s
Drawing Frog 8
 
w e l o s t b o y s's Avatar
 
Join Date: Nov 2009
Location: London
Country: Canada
Posts: 2,938
vCash: 500
These stats don't really say "bad luck" to me, the quality of the shot can explain both sv% against and his shooting %.

w e l o s t b o y s is offline  
Old
07-11-2012, 12:46 PM
  #3
Boom Boom Anton
Registered User
 
Boom Boom Anton's Avatar
 
Join Date: Sep 2006
Posts: 12,180
vCash: 500
The problem is that you are trying to make a statistical argument on an insufficient statistical sample size. This past season, Boychuk played 16 games and only 8 min / game on average so a total of 129min total. Comparing stats for him (even trying to adjust it to "per 60 minutes") to guys who played over 70 games and over 1000 minutes is flawed as there's way to much margin of error.

Boom Boom Anton is offline  
Old
07-11-2012, 12:47 PM
  #4
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by Stahl2Stall View Post
These stats don't really say "bad luck" to me, the quality of the shot can explain both sv% against and his shooting %.
PDO is the sum of "On-Ice Shooting Percentage" and "On-Ice Save Percentage" while a player was on the ice. It regresses very heavily to the mean in the long-run: a team or player well above 1000 has generally played in good luck and should expect to drop going forward and vice-versa.

Also, it's not his shooting percentage. It's the entire team while he's on the ice.

As a winger, save percentage is almost entirely luck in his case as wingers are rarely responsible for allowing high quality shots, and soft goals are a reality and a major flaw in +/-.

Stephen Goalbert is offline  
Old
07-11-2012, 12:49 PM
  #5
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by Boom Boom Anton View Post
The problem is that you are trying to make a statistical argument on an insufficient statistical sample size. This past season, Boychuk played 16 games and only 8 min / game on average so a total of 129min total. Comparing stats for him (even trying to adjust it to "per 60 minutes") to guys who played over 70 games and over 1000 minutes is flawed as there's way to much margin of error.
I do realize this, but there's nothing else to base anything on. These stats mean more than calling him a bust because he doesn't put up a lot of points in those 129 minutes, or really any other argument that labels him a bust.

Stephen Goalbert is offline  
Old
07-11-2012, 12:50 PM
  #6
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
I also just remembered that "sa-ber" gets auto corrected to "sabre". So the title says "sabremetric".

Stephen Goalbert is offline  
Old
07-11-2012, 12:51 PM
  #7
Boom Boom Anton
Registered User
 
Boom Boom Anton's Avatar
 
Join Date: Sep 2006
Posts: 12,180
vCash: 500
My issue with Boychuk is that he just seemed so unsure at times when he was with the Canes. Almost like he was afraid to make a mistake and thus didn't play to his strengths and natural talents.

I'm not sure if it's questionable hockey sense (which was an assessment by others), being asked to play a role in the past (grinding style) that doesn't suit him, a confidence issue, or something else. I still haven't thrown in the towel on him and hopefully under Muller he sees an opportunity to have a fresh start.

Boom Boom Anton is offline  
Old
07-11-2012, 12:52 PM
  #8
Boom Boom Anton
Registered User
 
Boom Boom Anton's Avatar
 
Join Date: Sep 2006
Posts: 12,180
vCash: 500
Quote:
Originally Posted by Shutdown Sutter View Post
I do realize this, but there's nothing else to base anything on. These stats mean more than calling him a bust because he doesn't put up a lot of points in those 129 minutes, or really any other argument that labels him a bust.
Not really. Conclusions based off insufficient (thus invalid) data are no more meaningful than conclusions based off no data.

Boom Boom Anton is offline  
Old
07-11-2012, 01:04 PM
  #9
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by Boom Boom Anton View Post
Not really. Conclusions based off insufficient (thus invalid) data are no more meaningful than conclusions based off no data.
His lack of playing doesn't mean these stats are invalid. It means you have to take them with a grain of salt. Something is more useful than nothing.

Stephen Goalbert is offline  
Old
07-11-2012, 01:13 PM
  #10
Boom Boom Anton
Registered User
 
Boom Boom Anton's Avatar
 
Join Date: Sep 2006
Posts: 12,180
vCash: 500
Quote:
Originally Posted by Shutdown Sutter View Post
His lack of playing doesn't mean these stats are invalid. It means you have to take them with a grain of salt. Something is more useful than nothing.
Call it what you want. It means you can't draw a valid conclusion from stats with insufficient sample size. To me that means invalid and the data is meaningless to me.

It's common statistical principles. If you don't have sufficient sample size, the error bars around the data are huge and any conclusions drawn from such data suspect.

Anyhow, I'm not one calling him a bust and I haven't given up on him, but I am not basing my assessment of Boychuk off of "nothing". I'm basing it off of the games I've watch him over the past 2-3 seasons.

Boom Boom Anton is offline  
Old
07-13-2012, 09:33 AM
  #11
TheBigKahuna
Registered User
 
TheBigKahuna's Avatar
 
Join Date: Dec 2010
Location: Canes Country
Posts: 2,861
vCash: 500
If Zach put in HALF the effort Skinner does over the summers, he'd have a good chance at playing in the NHL already. THIS is what separates the "haves from the have nots".

http://www.youtube.com/watch?v=q4uvG...ature=youtu.be

TheBigKahuna is offline  
Old
07-13-2012, 11:51 AM
  #12
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
Quote:
Originally Posted by Shutdown Sutter View Post

I fully realize I'm probably going to get called a stats nerd or whatever and get flamed for this, but I thought it was interesting. Also I'm bored.

Flame away.
0 goals in 16 games. What does that extrapolate out to over an 82 game season? Or that the other 4 guys in the bottom 5 of on ice shooting % were Nodl/Dwyer/Joslin/Samson.

And the PDO stuff varying from 1000 being luck is such half-ass analysis.

The idea that players should usually expect 1000 and if it varies, that's sometimes luck, sounds like something a statistician would write briefly at the top of page to make it an accessible idea to the general public.

The easy and relevant to our team counter-example is Alexander Semin, who shoots at 14.1% for his career and has been surrounded by skilled players. His expected PDO is going to be way above 1000. And his PDO history reflects that.

wallym is offline  
Old
07-13-2012, 02:13 PM
  #13
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by wallym View Post
0 goals in 16 games. What does that extrapolate out to over an 82 game season? Or that the other 4 guys in the bottom 5 of on ice shooting % were Nodl/Dwyer/Joslin/Samson.
Boychuk's game is as a playmaker. I'm not overly worried about him going 16 games without a goal playing on the 4th line under Maurice in a grinding role.

In 2010-2011 he scored at a 15 goal pace with over half of his time in the bottom six, and less than half with Staal and Cole (44% and 41% respectively)

Quote:
And the PDO stuff varying from 1000 being luck is such half-ass analysis.

The idea that players should usually expect 1000 and if it varies, that's sometimes luck, sounds like something a statistician would write briefly at the top of page to make it an accessible idea to the general public.

The easy and relevant to our team counter-example is Alexander Semin, who shoots at 14.1% for his career and has been surrounded by skilled players. His expected PDO is going to be way above 1000. And his PDO history reflects that.
Semin's PDO the last three years has gone from 1,051 to 1,044 to 1,011. So yes, it has regressed back to the mean.

PDO does have a lot to do with luck, I don't see how it couldn't. Half of it is determined by whether or not the goalie makes a save, which has almost nothing to do with a winger like Boychuk, but still negatively affects his +/-.

Stephen Goalbert is offline  
Old
07-13-2012, 03:10 PM
  #14
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
I don't think you know what you're talking about. Do you? (think that you understand how to analyze advanced statistics) Or are you just kind of guessing based on your intuitive understanding?

wallym is offline  
Old
07-13-2012, 03:15 PM
  #15
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by wallym View Post
I don't think you know what you're talking about. Do you? (think that you understand how to analyze advanced statistics) Or are you just kind of guessing based on your intuitive understanding?
I'd be better able to understand what you're asking if I knew what it was you're referring to.

Stephen Goalbert is offline  
Old
07-13-2012, 03:47 PM
  #16
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
I don't think you know what regressing towards the mean means.

Regressing towards the mean says that if the mean is 1000, every season his expectation is to have a PDO of 1000. So after his 1051 season, regressing to the mean would be the same player put in the same situation would expect 1000 the next season. The more seasons over 1000 in a row he puts up, the more it supports the case that the mean isn't in fact 1000.

A 3rd season (even if we pretend it was the same situation as the 3 previous seasons) of a value above 1000 supports his mean being higher than 1000, not that it is actually 1000.

An example.

I tell you a coin is fair and bet heads.

I flip the coin 1000 times. It comes up heads 700 times. I flip it again, it comes up heads 670 times. I flip it a 3rd time. It comes up 550 times.

Then I say "See, you can tell it's fair, because the number of times it's coming up heads has decreased 3 times in a row!"

wallym is offline  
Old
07-13-2012, 04:24 PM
  #17
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
I
Quote:
Originally Posted by wallym View Post
I don't think you know what regressing towards the mean means.

Regressing towards the mean says that if the mean is 1000, every season his expectation is to have a PDO of 1000. So after his 1051 season, regressing to the mean would be the same player put in the same situation would expect 1000 the next season. The more seasons over 1000 in a row he puts up, the more it supports the case that the mean isn't in fact 1000.

A 3rd season (even if we pretend it was the same situation as the 3 previous seasons) of a value above 1000 supports his mean being higher than 1000, not that it is actually 1000.

An example.

I tell you a coin is fair and bet heads.

I flip the coin 1000 times. It comes up heads 700 times. I flip it again, it comes up heads 670 times. I flip it a 3rd time. It comes up 550 times.

Then I say "See, you can tell it's fair, because the number of times it's coming up heads has decreased 3 times in a row!"
I don't see how a coin flip equates to all the variables that occur on the ice in a hockey game. Also, heads/tails are two outcomes. PDO doesn't mean that 1,000 is expected and anything else is lucky/unlucky.

Read these: http://www.pensburgh.com/2010/6/23/1...-what-it-means

http://www.arcticicehockey.com/2011/...-nhl-statistic

Stephen Goalbert is offline  
Old
07-13-2012, 04:48 PM
  #18
cptjeff
[insert joke here]
 
cptjeff's Avatar
 
Join Date: Sep 2008
Location: Washington, DC.
Country: United States
Posts: 9,039
vCash: 500
The goalie making a save or not sure as hell depends on the player. Some shots are better than others. If Boychuck's shot is weaker, or he doesn't pick spots well, has a slow release, lacks the deceptive release other players have, only manages to get shots from bad locations...

That's not bad luck, that's a lack of skill. Whether a goalie makes a save or not is dependent on much more than the goalie. If I scoot the puck along the ice at 20 MPH and the goalie stops it, it's counted as a save just like the glove save on the 90 MPH wrister top corner, but while half of those top corner shots might be stopped, about 100% of the weak little sliders along the ice from, say, the top of the slot, will be. The goalie stopping my shot while letting in the sniper's shot isn't his good luck and my bad luck- it's him performing better than I am.

cptjeff is offline  
Old
07-13-2012, 04:54 PM
  #19
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
You misused the concept of regression towards the mean. I explained why it is a misuse with the coinflip example. There was no other purpose for the coinflip example other than to show what regression towards the mean doesn't mean.

It was relevant, because I said Alexander Semin has a higher expected PDO than 1000 because of an assortment of hockey factors, in particular his consistently high shooting % and good overall team.

Then you said "Semin has 1051-1046-1011, which is regressing towards the mean."

Let's toss out the misuse of the term. What do you think the expected PDO for Semin during Washington's 09-10 season is for example. With luck being a non-factor.

Do you think other players on the Washington Capitals from the 09-10 season, with luck being removed, would have the same expected PDO

If you don't believe that other players on the Washington Capitals have the same expected PDO as Semin, and probably had expected PDO's 30 or 40 points lower, then how are you coming to the conclusion that Boychuk was unlucky, because he had the lowest PDO on the Hurricanes?

Which goes back to the original question of "Do you think you know what you're talking about, or are you just guessing?"

wallym is offline  
Old
07-13-2012, 04:59 PM
  #20
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by cptjeff View Post
The goalie making a save or not sure as hell depends on the player. Some shots are better than others. If Boychuck's shot is weaker, or he doesn't pick spots well, has a slow release, lacks the deceptive release other players have, only manages to get shots from bad locations...

That's not bad luck, that's a lack of skill. Whether a goalie makes a save or not is dependent on much more than the goalie. If I scoot the puck along the ice at 20 MPH and the goalie stops it, it's counted as a save just like the glove save on the 90 MPH wrister top corner, but while half of those top corner shots might be stopped, about 100% of the weak little sliders along the ice from, say, the top of the slot, will be. The goalie stopping my shot while letting in the sniper's shot isn't his good luck and my bad luck- it's him performing better than I am.
Save percentage means save percentage of his own goalie while he's on the ice.

Stephen Goalbert is offline  
Old
07-13-2012, 05:16 PM
  #21
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by wallym View Post
Let's toss out the misuse of the term. What do you think the expected PDO for Semin during Washington's 09-10 season is for example. With luck being a non-factor.
It's impossible for luck to be a non-factor in PDO. The stat is almost entirely luck. That's like saying "Tell me what Cam Ward's save percentage was this year with saves being a non-factor."

Quote:
Do you think other players on the Washington Capitals from the 09-10 season, with luck being removed, would have the same expected PDO

If you don't believe that other players on the Washington Capitals have the same expected PDO as Semin, and probably had expected PDO's 30 or 40 points lower, then how are you coming to the conclusion that Boychuk was unlucky, because he had the lowest PDO on the Hurricanes?
PDO has little to no bearing upon the quality of the player.

Our ten highest PDO's were Nash, Brett Sutter, Stewart, Ruutu, Brent, Tlusty, Jokinen, Bowman, Welsh, Dwyer.

Does that list look anything like how you would rank our top ten forwards?

Jordan Staal and James Neal were 13th and 15th respectively amongst Penguin forwards.

D. Sedin and Kesler were 11th and 12th on Vancouver. Parise and Kovalchuk were 8th and 10th on New Jersey.

I don't see your point here. It's not a stat to measure the quality of a given player. It's a stat to measure the luck of a given player.

Semin has had high PDO's in the past, sure, but 3 out of his 5 seasons that PDO is available for, he's been 102% or lower. PDO's as high above 100% as he was in 09-10 and 10-11 are unsustainable over the course of a career. It's a fact.

Stephen Goalbert is offline  
Old
07-13-2012, 05:35 PM
  #22
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
Quote:
Originally Posted by Shutdown Sutter View Post
It's impossible for luck to be a non-factor in PDO.
I dispute almost every other conclusion you've drawn, but until we get past this, we're not going anywhere.

You claim the PDO is almost entirely a measure of luck. Which insinuates there is a 'mean' value with which to compare it to. An 'expected' value. Some number that if greater than, a person is lucky. If smaller than, the person is unlucky.

In a hypothetical world, where the 09-10 Capitals played a trillion games, you would find the true "mean" of Alexander Semin's PDO. That is the PDO he would have if luck were completely neutral. (his average PDO over a trillion games)

What do you think this hypothetical luck neutral value of Alexander Semin's PDO was in 09-10? (The value that we would say he was neither lucky or unlucky)


Last edited by wallym: 07-13-2012 at 05:50 PM.
wallym is offline  
Old
07-13-2012, 05:52 PM
  #23
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by wallym View Post
I dispute almost every other conclusion you've drawn, but until we get past this, we're not going anywhere.

You claim the PDO is almost entirely a measure of luck. Which insinuates there is a 'mean' value with which to compare it to. An 'expected' value. Some number that if greater than, a person is lucky. If smaller than, the person is unlucky.

In a hypothetical world, where the 09-10 Capitals played a trillion games, you would find the true "mean" of Alexander Semin's PDO. That is the PDO he would have if luck were completely neutral.

What do you think this hypothetical luck neutral value of Alexander Semin's PDO was in 09-10? (The value that we would say he was neither lucky or unlucky)
Based on the data available to us, it would be 1000.

Stephen Goalbert is offline  
Old
07-13-2012, 05:54 PM
  #24
wallym
Registered User
 
Join Date: May 2009
Posts: 2,728
vCash: 500
Follow up. What do you think is Alexander Semin's shooting %? In a luck neutral world?

And for added info, what's the average NHL shooting %. (general guess is fine)

Even simpler if you want, would you agree Alexander Semin has a higher Shooting % based on skill (not because of luck) than the average NHL hockey player?


Last edited by wallym: 07-13-2012 at 06:17 PM.
wallym is offline  
Old
07-13-2012, 06:22 PM
  #25
Stephen Goalbert
Puck Don't Lie
 
Stephen Goalbert's Avatar
 
Join Date: Oct 2010
Location: Bronx, NY
Country: United States
Posts: 17,438
vCash: 500
Quote:
Originally Posted by wallym View Post
Follow up. What do you think is Alexander Semin's shooting %? In a luck neutral world?

And for added info, what's the average NHL shooting %. (general guess is fine)

Even simpler if you want, would you agree Alexander Semin has a higher Shooting % based on skill (not because of luck) than the average NHL hockey player?
Yes, but I think you're missing the fact that it isn't just Semin's shooting percentage that influences PDO. It's everyone's collective shooting percentage while Semin is on the ice.

Stephen Goalbert is offline  
Closed Thread

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 04:45 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.